Data Notes
Some things a user of a table should know cannot be expressed by the table itself: an answer key in the source that looks wrong, covariates the source filled with a mean, what a wave column counts, forms pooled under shared item names. IRW reproduces its sources as released (Corrections, rule 7), so these are recorded as notes rather than fixed. Each note also appears on the table’s own page.
A note is not a defect report. Tables with a known IRW error are flagged on their page until the fix is released, and released fixes are listed on Corrections. Notes about item text are on Known Issues.
1545 notes covering 1525 tables. Start typing to search.
| Table | Note | Issue | Added |
|---|---|---|---|
aanesen_2018_coastal_recreation |
Conjoint table (irw_conjoint). Discrete choice experiment, choice = “Hva foretrekker du?” among the current-regulation status quo (profile 1, fixed levels) and two stricter-regulation scenarios, 8 fixed cards; no opt-out beyond the status quo. Level text is the questionnaire’s Norwegian wording mapped from the codebook codes; the questionnaire’s example cards differ from the data’s cards. 82 tasks without a single valid choice dropped; only the with-jobs half of the split sample is deposited (N = 517). | 2026-10-09 | |
aanesen_2019_marine_tradeoffs |
Conjoint table (irw_conjoint). Discrete choice experiment, choice = “Hva foretrekker du?” among the current-regulation status quo (profile 1, fixed levels) and two stricter-regulation scenarios, 8 fixed cards; no opt-out beyond the status quo. Level text is the questionnaire’s Norwegian wording mapped from the codebook codes; the questionnaire’s example cards differ from the data’s cards. 75 tasks without a choice dropped; without-jobs half of a split sample (N = 490; the with-jobs half is aanesen_2018_coastal_recreation). | 2026-10-09 | |
aanesen_2021_cold_water_coral |
Conjoint table (irw_conjoint). Discrete choice experiment: ‘I prefer’ Alternative 1, 2 or 3 (same as today); the status quo is profile 3 with its printed levels, no further opt-out. Each participant answered the 12 cards before and again after a group discussion (tasks 13-24, trial_repeat_of). Two fixed card versions; 40 second-round tasks with an invalid cost code and 16 without exactly one choice dropped. | 2026-10-09 | |
aassve_2024_family_ideals_a |
Conjoint table (irw_conjoint_2). Six vignettes per respondent, experiment 1 of two; rating = perceived success of the family, 0-10 (higher = more successful; wording not in the deposit). Eight countries pooled (cov_country), weight kept. Level text is the authors’ short value labels, not the displayed sentences. Vignettes come from a fixed blocked set (trial_vignette_id, trial_block). N = 7985 respondents; ratings missing for a few vignettes are omitted. | 2026-10-11 | |
aassve_2024_family_ideals_b |
Conjoint table (irw_conjoint_2). Six vignettes per respondent, experiment 2 (adds child education and financial support for children) of two; rating = perceived success of the family, 0-10 (higher = more successful; wording not in the deposit). Eight countries pooled (cov_country), weight kept. Level text is the authors’ short value labels, not the displayed sentences. Vignettes come from a fixed blocked set (trial_vignette_id, trial_block). N = 12156 respondents; ratings missing for a few vignettes are omitted. | 2026-10-11 | |
aassve_2026_parenthood |
Conjoint table (irw_conjoint). Five paired scenario tasks; choice = forced choice of the scenario that would make the respondent more inclined to have a child (no opt-out; 399 tasks without a choice keep their ratings), rating = the respondent’s intention to have a child within the next three years under each scenario, 1-10 (paraphrase; wording not deposited). Levels are the authors’ short labels; income is the amount shown in local currency. Four countries pooled (cov_country); N = 8,330 (press release 8,334). | 2026-10-09 | |
abascal_2021_neighborhood_diversity |
Conjoint table (irw_conjoint). choice: “Which of these two neighborhoods do you think is more racially diverse?”; rating 1 = not racially diverse at all to 7 = very racially diverse (a perception, not a preference). The only stimulus was a composition such as ‘90% Asian, 8% Black, and 2% White’, split into a share pattern and three groups. 196 straightliners kept and flagged (cov_straightliner); the paper’s N of 1,803 excludes them. | 2026-10-07 | |
abascal_2025_racial_classification |
Conjoint table (irw_conjoint). One profile per task: respondents rated each of 9-10 profiles on six statements (“This person is White”, … “is a person of color”), 1 = strongly disagree to 5 = strongly agree. Dynata (1,719) and student (1,016) samples pooled as in the article (cov_sample); one Dynata profile per respondent was dropped by the authors for a programming error. trial_condition is the respondent-level arm (self-report on an anonymous survey vs a scholarship application). The face is stored as the authors’ image code and skin-tone version; attr_photo_gender and attr_photo_ethnicity are the authors’ coding of the face, not displayed text. 58 respondents with no data and rows with no answers dropped: 2,653 respondents. | 2026-10-07 | |
acharya_2018_democratic_peace |
Conjoint table (irw_conjoint). One text vignette per respondent, four independent binary factors (alliance, democracy, trade, motives/threat statement). rating = ‘Would you favor or oppose using the U.S. military to attack the country’s nuclear development sites?’ 1 Favor strongly to 5 Oppose strongly, stored raw (higher = more opposed). Covariates unlabelled, kept as codes. One respondent without an answer omitted (1,246 of 1,247). | 2026-10-09 | |
acharya_2018_scotus_nominees |
Conjoint table (irw_conjoint). Single nominee profiles, six per respondent; rating = support (‘On a scale from strongly oppose to strongly support…’), rating_qualified, rating_trust, each 5-point answer text coded 1-5, higher = more favourable. Between-subject arms (trial_arm): party leaning shown (partisan) or ‘(not shown)’ (nonpartisan); attribute order randomized per respondent (attrpos_). Issue-specific trust questions Q23/Q24 not kept. 1,567 respondents rated at least one profile; the article reports 1,650 (886 partisan, 764 nonpartisan). | 2026-10-09 | |
adiguzel_2025_checks_balances |
Conjoint table (irw_conjoint). Five paired candidate tasks with two Turkish-text attributes (judicial-oversight law, social-policy plan); choice = which candidate the respondent would select (paraphrase; no opt-out; 231 tasks unanswered), rating_democrat = how democratic the candidate is, 0-10, rating_like = support for the candidate, 1-5 (paraphrases; anchors not deposited). trial_version marks the arm in which the oversight laws carry a gridlock or corruption justification. 510 of 515 respondents answered; the authors’ 5 speeders are kept (cov_time_min). | 2026-10-09 | |
adman_2024_climate_refugees |
Conjoint table (irw_conjoint). Outcomes: forced choice ‘Which refugee should receive the visa?’ (choice) and ‘which refugee would be more likely to integrate successfully after arriving in the United States?’ (choice_integrate), no opt-out. The article reports 5,175 profile pairs while the deposit holds 5,585 (1,117 x 5). Unlabelled survey codes and Qualtrics IDs dropped; randomization details not documented. | 2026-10-08 | |
adsera_2023_value_of_democracy |
Conjoint table (irw_conjoint). Seven pairs of hypothetical societies; choice = the society that would be best for the respondent (forced choice, no opt-out); rating = each society on 0 (very bad) to 10 (very good) per the screen, but the deposit’s values run 1-10 (stored raw, undocumented). Attribute text is English with amounts rebuilt from the article’s design (Brazil/France saw translated text); no restrictions documented. All 6,014 respondents kept; the authors’ duration >= 600 filter (cov_duration) gives exactly the article’s 26,740 / 23,702 / 22,736 observations. | 2026-10-09 | |
aguirre_camacho_2021_shai |
Item text: item_text is NA for all 18 items, and that is correct. Each SHAI item is a set of four graded statements with no shared stem; the statements are in option_text (one per resp value). | #1649 | 2026-10-01 |
ahlen_2026_family_reunification |
Conjoint table (irw_conjoint). Six pairs of fictitious resident family members; choice = which one should be allowed family reunification (paraphrase; forced), rating = 1-7 support for the profile (paraphrase; higher = more inclusive per the authors’ code; anchors not deposited). Levels are the authors’ English recode labels; the incoming member’s origin is fixed per respondent (trial_incoming_origin); combinations restricted (e.g. native-born with native parents always citizen and fluent); Germany, Italy and Sweden pooled; N = 4,119. | 2026-10-09 | |
ahlen_2026_migrant_admission |
Conjoint table (irw_conjoint). Six pairs of migrant profiles; choice = which migrant is preferred (wording not deposited; forced). Levels are the authors’ English recode labels; religion and origin country form one combined attribute with unequal shares, and escape from war never occurs with Christians from the USA or Muslims from India, nor labour migration with inability to work; Germany, Italy and Sweden pooled; N = 4,119. | 2026-10-09 | |
ahler_2022_james_vignette |
Conjoint table (irw_conjoint). Factorial vignette (2x2x3x3), one per respondent: rating = answer to ‘Which of the following is most likely?’ coded 1 = James works in sales, 2 = … and is an active supporter of the Democratic Party, 3 = … of the Republican Party (nominal codes, not a scale; 2 and 3 are conjunction fallacies). 9 rows without condition or answer dropped; authors’ quality filter kept as cov_untrustworthy (1,507 kept). | 2026-10-09 | |
ahrens_2026_wealth_tax_spending |
Conjoint table (irw_conjoint). choice: which of two spending proposals the respondent prefers (paraphrase; no opt-out); rating: “Wie sehr sind Sie für Option 1/2?” 1 Sehr stark dafür .. 5 Sehr stark dagegen, stored raw (lower = more in favour). Design restriction: every proposal has at least one ‘Keine zusätzlichen Ausgaben’ level. Earlier framing arm in trial_frame; entropy-balancing weights kept. 4,393 tasks have ratings but no choice (choice NA). Authors’ 4,653-respondent sample. | 2026-10-08 | |
ahsanjansson_2026_tax_compliance |
Conjoint table (irw_conjoint). Factorial vignette (8 binary factors, equal probability), one per respondent. rating_pay_tax ‘Should the person pay income tax to the government?’, rating_justifiable ‘Would it be justified if the person did not report all their income to pay less tax?’, rating_trust_gov ‘Would you trust the government to spend the tax money wisely?’: 1 = Yes, 0 = No, Don’t know/Refuse missing. Levels are the authors’ short value labels. Deposit N 2,080 vs article 2,000; 26 respondents with no answers dropped. | 2026-10-09 | |
aklin_2026_climate_pledges |
Conjoint table (irw_conjoint). Experiment 1 (home-country pledges). choice: ‘Which of the following pledges would you prefer that [HOMECOUNTRY] makes?’ forced choice between two profiles, 3 tasks. Ten countries pooled (cov_country), levels in English as stored (surveys in local languages). Attribute order: first two fixed, last two randomized per respondent (not recorded). Profile order inferred from file order. Respondents who did not choose in every task were excluded by the authors. | 2026-10-09 | |
aklin_2026_climate_transfers |
Conjoint table (irw_conjoint). Experiment 3 (international transfers). choice: ‘Which country would you prefer to see financial and technical assistance given to?’ forced choice between two profiles, 3 tasks. Ten countries pooled (cov_country), levels in English as stored (surveys in local languages). Attribute order: first two fixed, last two randomized per respondent (not recorded). Profile order inferred from file order. Respondents who did not choose in every task were excluded by the authors. | 2026-10-09 | |
albert_2024_negative_partisanship |
Conjoint table (irw_conjoint_2). choice = which of the two candidates the respondent would vote for, choice_represent = which would best represent the respondent’s interests (both paraphrased from the authors’ code, forced choice A/B, no opt-out, up to 6 pairs). Attribute text is the piped fragment stored in the export; the ideology set depends on the respondent’s party block (trial_primary); independents and attention-check failures have no answers and are omitted. Study 2 (issue-position variant) is not built because only Candidate A’s attributes are recorded; N against the article not checked. | 2026-10-09 | |
aleksovska_2021_accountability_priority |
Conjoint table (irw_conjoint). Up to four paired stakeholder tasks; choice = ‘Which stakeholder’s demand would you prioritize?’ (forced choice, no opt-out). Level text from the article’s Table 1 (English instrument; Dutch sample saw Dutch text, not deposited). Tasks with identical profiles were removed by the authors before deposit; 58 unanswered tasks dropped. UK and NL pooled (cov_country); 4,637 tasks match the article. | 2026-10-09 | |
alhatmi_2026_oman_graduate_hiring |
Conjoint table (irw_conjoint_2). Single-profile text resumes (nine attributes) rated by Omani employers: rating_hire (likelihood of hiring) and rating_train (likelihood of being easy to train), both 0-100 as stored (the README says 0-10; anchors unknown). Fixed blocked design: each respondent saw the 12 vignettes of one of 12 decks (trial_deck) for one occupation (trial_job_type); no opt-out. Only a README, not the article, identifies the study. | 2026-10-14 | |
alkhaldi_2023_whoqol_autism |
cov_age is an age-band code, not age in years: 1 = “Less than 35 years”; 2 = “From 35 to 45 years”; 3 = “45 years and over”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
alrababah_2023_refugee_return |
Conjoint table (irw_conjoint). Single-profile vignette conjoint read aloud: choice = ‘Under these conditions, would you be willing to return to Syria?’ (1 yes, 0 no; opt_out yes: no is the outside option). Five vignettes per respondent; level text is the article’s English translation; jobs and services alternate within one Syria slot and one Lebanon slot. Attribute order randomized by respondent (not recorded). cov_survey_weight = sampling weight. 287 unanswered and 5 blank vignettes dropped. | 2026-10-09 | |
alsuhibani_2022_npi_s3 |
Item text: item_text is NA for all 13 items, and that is correct. The NPI-13 is forced-choice, with no item stem; each item is a pair of statements, and both are in option_text (one per resp value). | #1649 | 2026-10-01 |
alvarado_2024_tax_fairness_argentina |
Conjoint table (irw_conjoint). choice: “¿Cuál de las dos personas preferiría usted que pague una tasa de impuestos más alta?” (forced choice of one of two; skipped tasks omitted); choice = 1 marks the person chosen to pay MORE tax (unfavourable, kept as asked). Attributes fully randomized per the article; attribute order randomized per respondent (attrpos_). All 1,688 respondents with an answer kept; cov_in_authors_sample marks the author’s 1,527 (finished, consented, not speeding), matching the article. | 2026-10-07 | |
alvarado_2024_tax_fairness_australia |
Conjoint table (irw_conjoint). choice: “Which of the two individuals would you personally prefer to charge a higher tax rate to?” (forced choice of one of two; skipped tasks omitted); choice = 1 marks the person chosen to pay MORE tax (unfavourable, kept as asked). Attributes fully randomized per the article; attribute order randomized per respondent (attrpos_). All 1,633 respondents with an answer kept; cov_in_authors_sample marks the author’s 1,450 (finished, consented, not speeding), matching the article. | 2026-10-07 | |
alvarado_2024_tax_fairness_chile |
Conjoint table (irw_conjoint). choice: “¿Cuál de las dos personas preferiría usted que pague una tasa de impuestos más alta?” (forced choice of one of two; skipped tasks omitted); choice = 1 marks the person chosen to pay MORE tax (unfavourable, kept as asked). Attributes fully randomized per the article; attribute order randomized per respondent (attrpos_). All 1,584 respondents with an answer kept; cov_in_authors_sample marks the author’s 1,418 (finished, consented, not speeding), matching the article. | 2026-10-07 | |
alvarado_2024_tax_fairness_us |
Conjoint table (irw_conjoint). choice: “Which of the two individuals would you personally prefer to charge a higher tax rate to?” (forced choice of one of two; skipped tasks omitted); choice = 1 marks the person chosen to pay MORE tax (unfavourable, kept as asked). Attributes fully randomized per the article; attribute order randomized per respondent (attrpos_). All 2,081 respondents with an answer kept; cov_in_authors_sample marks the author’s 1,946 (finished, consented, not speeding), matching the article. US weights not deposited. | 2026-10-07 | |
amoah_2026_vigilantism |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (Ghana, social-media users): vigilante group’s party (NPP/NDC) x action (snatch ballot boxes in a peaceful area / protect ballot boxes in a malpractice-prone area; one bundled factor), plus a placebo arm showing neither attribute ‘(not shown)’. Six ratings in the authors’ codes (1-5 agreement for justified, undermined credibility, threat to democracy; 0-2 for willingness to participate, sign a petition, join a demonstration). 1,853 of 2,603 respondents have an arm and an outcome (late dropout). No article DOI in the deposit. | 2026-10-14 | |
amorim_2025_climej_florescimentonotrabalho |
resp runs 1 to 7. The source paper (Amorim et al. 2025) says the flourishing items were answered from 1 (discordo totalmente) to 5 (concordo totalmente), but the deposit’s own dictionary records the scale as “Intervalar, 1 a 7, Concordância” and the data use all seven points. The data are right and the paper’s sentence is wrong; do not rescale to 1-5. | #2313 | 2026-09-30 |
andersen_2026_party_masking |
Conjoint table (irw_conjoint). choice: forced choice between two candidates (wording not deposited), no opt-out; rating_guess_democrat: no-party arm, task 5 only, respondent’s guess of each candidate’s party (1 = Democrat, 0 = Republican). Between-subject arms in trial_party_info; attr_party is (not shown) in the no-party arm. Restrictions and attribute order undocumented. 1,363 ids (arms 683/662 as in the authors’ log); 37 with no answers dropped. | 2026-10-07 | |
anoll_2018_good_neighbor |
Conjoint table (irw_conjoint_2). Four hypothetical neighbours per respondent shown as an information table; rating_like and rating_respect are 1-7 ratings (wording and anchors not in the deposit; 7 = most presumed). Profile 4 is always the control ‘Annually attends neighborhood potluck dinners’ (text from the article’s corrigendum); the other three values of each attribute are used once per respondent. Three samples pooled (cov_sample); N = 2,000 + 1,201 + 500. | 2026-10-11 | |
arceneaux_2023_republican_primary |
Conjoint table (irw_conjoint). choice: “Suppose you are voting in a primary election between two Republican candidates who are running for Congress. Which candidate do you prefer?” Candidate 1 / Candidate 2, no opt-out. Shown only to Republican and Independent identifiers (cov_party_id). Randomization restrictions and level weights undocumented; level shares in the table are near uniform. All 1,883 respondents with an answered task kept; the authors use good completes (cov_good_complete = 1, 1,481). No survey weights deposited. | 2026-10-07 | |
argote_2023_independent_candidates |
Conjoint table (irw_conjoint). choice: vote for candidate 1, candidate 2 or neither (wording not deposited); neither = 0 on both profiles, opt_out = yes (6,908 of 19,825 tasks; 646 respondents chose neither in every task). Independent shown 50% of the time. Coalition levels are the authors’ labels (Center-Right, Center-Left, Left, Right, Independents), not the displayed coalition names. Priming arm in trial_priming. Panel IDs re-keyed; cell and rake weights kept, analyses unweighted. | 2026-10-08 | |
argote_2024_migrants_colombia |
choice = which of two migrants the respondent preferred (forced, no opt-out), missing for 423 rows where only the 1-7 support rating was answered; rating direction inferred (7 = most support). Outcome wording and the Spanish level text are not in the deposit, so levels use the authors’ English labels. Dropped 5 respondents with no attributes, the authors’ dummies/derived variables (incl. ipw), unlabelled items, and respondent_serial (contains personal data in the source); counts match the authors’ Table D1 (2,428 choice / 2,811 support rows). | #2887 | 2026-10-07 |
argote_2025_ideology_candidates_chile |
Conjoint table (irw_conjoint). choice = which of two presidential candidates the respondent picks (options ‘Opción 1’, ‘Opción 2’, ‘Ninguno’; question wording not deposited); OPT-OUT: ‘Ninguno’ chosen in 6,184 of 15,375 tasks, both profiles 0 (the authors drop these). Attributes randomized simultaneously; levels are the authors’ English labels of the Spanish display. Age levels are 40/50/60 in the data (the article text says 35/45/55/65). Cell (cov_survey_weight) and rake (cov_weight_rake) post-stratification weights kept; main estimates are unweighted. | 2026-10-08 | |
argote_2026_informality_immigrants |
Conjoint table (irw_conjoint). One vignette per respondent about a Venezuelan migrant (name Juan/María Pérez, previous occupation, degree recognition shown only for engineer/doctor and otherwise (not shown), salary expectation). Ratings stored as answer codes: rating = presence of Venezuelan immigrants like [name] 1 Muy positiva .. 5 Muy negativa; rating_stay = [name] should be allowed to stay 1 Muy de acuerdo .. 5 Muy en desacuerdo; rating_job_chances and rating_salary 1 decreases .. 3 improves; rating_compete 1 Muy probable .. 5 Muy improbable. Occupation levels are unequal by design (engineer/doctor about a third each). Panel ID and a free-text answer dropped; N 1,203 matches the README. | 2026-10-08 | |
arias_2022_climate_migrants_germany |
Same design and wording in German (“Ihr Land”), 7 = definitely admit; forced choice, no opt-out; the fluency/internal-origin restriction held. Same drops as the US table; 1,072 here, with the US 2,157 vs the article’s 2,160. | #2887 | 2026-10-07 |
arias_2022_climate_migrants_us |
rating = “On a scale from 1 to 7, where 1 indicates that your state should absolutely not admit the migrant and 7 indicates that your state should definitely admit the migrant, how would you rate Migrant 1/2?”; choice = which of the two the respondent would prefer be allowed to stay (forced, no opt-out). The intended restriction (no ‘None’ language fluency for migrants from another region of the country) was not enforced in the US data. Preview responses, IP/location, Qualtrics/panel IDs, city, free text and unlabelled attitude items dropped; 1,085 here vs 2,160 for both countries in the article. | #2887 | 2026-10-07 |
arias_2026_local_climate_plans |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical local climate plans shown to elected US local policymakers; choice = the plan the respondent prefers (forced, no opt-out); rating = support for each plan on a five-point scale (5 = strongly support, 4 = somewhat support; labels of 1-3 not stated), stored raw. Attribute row positions are recorded (attrpos_) and the weight used in the replication code is kept. 573 respondents (matches the article), 2,102 tasks; 18 profiles lack a rating. | 2026-10-12 | |
arista_2010_achievement_motivation |
Five items (item1, item9, item11, item14, item26) correlate about zero with the rest of the scale, consistent with negatively worded items left unreversed; the item wording is not deposited, so they are shipped as stored. | 2026-10-09 | |
armaly_2024_scotus_nominees |
Conjoint table (irw_conjoint). Four single nominee profiles per respondent, six ratings each: rating = support for the nominee, rating_trust (impartial decisions), rating_quality (qualifications), all 1 weak to 6 strong; rating_court_approval (1 strongly disapprove to 5 strongly approve of the Court if the nominee joined), rating_court_legal and rating_court_agree (1 strongly disagree to 5 strongly agree). Wordings paraphrased from the article. Attribute order randomized per respondent (attrpos_). 1,223 respondents as in the article; Asian and Hispanic nominee profiles kept. | 2026-10-09 | |
aspide_2025_debt_austerity |
Conjoint table (irw_conjoint). choice: ‘Quale tra questi due partiti voterebbe?’ (forced, no opt-out); rating: agreement with each party agenda, 0 (per niente) to 10 (totalmente). Task 9 repeats task 1 with the two columns swapped (trial_repeat_of = 1; 69% consistent). Attribute order randomized per respondent (attrpos_). Respondents without a task-1 choice dropped, as in the authors’ code; covariates are numeric codes without a codebook. | 2026-10-09 | |
atik_2025_gwb |
Out-of-set cells dropped as entry errors: C1 = 6 or 7 (10 cells, only on C1), C10 = 6 (1), and six 0s across C9-C11. Column blocks (A = SWLS, B = AI scale, C = general well-being) were identified from the paper’s Methods and its reported .74 criterion correlation; the deposit has no labels. | 2026-10-06 | |
atkeson_2020_school_board |
Outcome: which of two school board candidates the respondent would vote for, forced choice, no opt-out (exact wording in the article). Attribute levels are as stored in the deposit; how the screen labelled them is undocumented. Qualtrics IDs re-keyed; free-text “other” answers and the derived shared-party variables are dropped. | #2887 | 2026-10-07 |
avina_2026_immigrant_benefit |
Outcome is “Which of these two immigrants is most beneficial to the US?” (OSF preprint SI), a forced choice with no opt-out: a judgement of benefit, not admission. Renamed from avina_2026_immigrant_admission on 2026-10-08. Levels are the codebook labels; randomization not uniform (‘Illegal alien’ about half of profiles). The authors’ high/low education, income and ideology splits and prior-attitude classification dropped, Prolific-type IDs re-keyed; N 1,999 not checked against the paywalled article. | #2887 | 2026-10-07 |
babenko_2018_achievement_goals |
Items suffixed _RC are stored already reverse-scored (every one correlates positively with its block) and are kept as stored. | 2026-10-09 | |
babenko_2018_basic_needs |
Items suffixed _RC are stored already reverse-scored (every one correlates positively with its block) and are kept as stored. | 2026-10-09 | |
babenko_2018_olbi_engagement |
Items suffixed _RC are stored already reverse-scored (every one correlates positively with its block) and are kept as stored. | 2026-10-09 | |
babenko_2018_olbi_exhaustion |
Items suffixed _RC are stored already reverse-scored (every one correlates positively with its block) and are kept as stored. | 2026-10-09 | |
babenko_2018_scs_sf |
Items suffixed _RC are stored already reverse-scored (every one correlates positively with its block) and are kept as stored. | 2026-10-09 | |
bajka_2025_occupation_choice |
Conjoint table (irw_conjoint). choice: respondents chose between two juxtaposed occupational profiles (paraphrase; verbatim prompt not deposited), forced choice, no opt-out. Level text is the English attribute text of the article’s Table 1, matched to the deposit’s one-word labels; attribute order randomized but not recorded. 2,144 respondents, as in the article. | 2026-10-08 | |
bakker_2021_populist_appeal |
Conjoint table (irw_conjoint). choice = ‘Which candidate do you prefer?’ (Candidate A or B, forced, one per task); rating_feeling = feeling thermometer for each candidate, 0 very unfavorable to 100 very favorable. Two rounds. Level text is the displayed text from the Qualtrics export. 13 answered tasks with unsaved profiles dropped (865 respondents vs 868 in the paper). Agreeableness (cov_agree_1-10, raw) and authoritarianism items kept; demographics as unlabelled codes. | 2026-10-09 | |
ballardrosa_2017_income_tax |
Conjoint table (irw_conjoint). choice = which of two income tax plans the respondent prefers (wording not deposited; paraphrase), forced, no opt-out. Profiles give the marginal tax rate in six income brackets and the revenue the plan would raise relative to current revenue; the revenue attribute was shown to 2,000 respondents (the authors’ main sample) and not to 250 (attr_revenue ‘(not shown)’, trial_saw_revenue = 0). Survey weight kept. | 2026-10-08 | |
banda_2022_filibuster |
Conjoint table (irw_conjoint_2). Factorial vignette (US, 2022): one Senate-candidate rally vignette per respondent, 2x3 of party x filibuster stance (control arm omits the stance sentence, ‘(not shown)’); rating = favorability 1-5 (higher = more favorable), rating_supports_elimination 1-5 (a manipulation check) and rating_ideology 1-7 (higher = more conservative). Level text and wording from the deposit codebook. | 2026-10-14 | |
banerjee_2025_arithmetic_market |
Study 3a wave 1 only, first attempts only. Each child solved one of two variants of each subtraction and division problem (randomised); the item code names the operands. The mystery-shopping and hypothetical-market questions are not included (each uses the child’s own goods and prices, so no two children answer the same item), nor are second attempts and post-hint attempts (asked only after a wrong answer or under a randomised hint). Study 3a wave 2 is a different instrument and is not here. cov_market is the authors’ corrected market id. | #819 | 2026-10-01 |
banerjee_2025_arithmetic_school |
Item codes are the survey’s. b1a-b4: ASER tool, adaptive (b1c, b2, b3, b4 are asked only on certain earlier outcomes, so they have few rows; b4 has 2). c = written abstract, d = oral abstract, e = oral anchored (word problem); the digit is the operation (1 add, 2 subtract, 3 multiply, 4 divide) and _v1-_v4 the survey version. Each child got one version, and versions use different numbers, so c1_v1 and c1_v2 are different items. f1 = oral anchored market problem; f2 = pretend market, scored as the authors do (a total within rounding of the exact one counts). Division answers within a rounding band count as correct, per the enumerators’ coding. The randomised-hint follow-up to f1 is not included. | #819 | 2026-10-01 |
bankert_2023_consumer_origin |
Outcome: “If you had to purchase one of the products above, which one would it be?” (Product A / Product B), forced choice, no opt-out; 5 unanswered tasks omitted. Displayed price rebuilt as base price x (1 + markup) from appendix Table A3; the follow-up purchase-likelihood rating is not in the deposit. Profile (screen) position is not recorded, so profile is the source row order; wave-1-only respondents and the Lucid replication (no product column) are excluded. | #2887 | 2026-10-07 |
bansak_2016_asylum |
Conjoint table (irw_conjoint). Fifteen country samples of one translated design in one table (cov_country), since the deposit holds only the English master labels. choice forced, no opt-out; rating 1-7 support for granting asylum. cov_survey_weight is the raw weight; the authors top-code it at 6 (missing for 147 respondents). 9 respondents with blank profiles dropped (18,021 vs ‘18,000’ in the paper). | 2026-10-07 | |
bansak_2019_satisficing_candidates_ssi |
Conjoint table (irw_conjoint). choice = ‘Which candidate would you prefer to vote for?’ forced choice of two profiles; rating = 1-7 per profile (1 definitely would NOT vote for, 7 definitely would vote for). Each respondent saw a randomly assigned number of filler attributes (trial_n_fillers), drawn at random from a pool; fillers not shown are ‘(not shown)’. Core attributes always shown; attribute order randomized but not recorded. Qualtrics IDs re-keyed; the authors’ recoded concordance variables dropped. | 2026-10-07 | |
bansak_2019_satisficing_hotels |
Conjoint table (irw_conjoint). choice = ‘Which hotel room do you prefer? (paraphrase)’ forced choice of two profiles; rating = 1-7 per profile (anchors not documented). Each respondent saw a randomly assigned number of filler attributes (trial_n_fillers), drawn at random from a pool; fillers not shown are ‘(not shown)’. Core attributes always shown; attribute order randomized but not recorded. Qualtrics IDs re-keyed; the authors’ recoded concordance variables dropped. | 2026-10-07 | |
bansak_2021_austerity_italy |
Conjoint table (irw_conjoint_2). choice = preferred package in the pair (forced, no opt-out; wording not quoted in the article); rating = ‘If you could vote on each of these options in a referendum, how likely is it that you would vote in favor or against each of the options?’ 1 = vote definitely against, 10 = vote definitely in favor (translated). Eight attributes are percentages 0-30 in steps of 2, stored as the source numbers (cuts negative); the order of the spending and tax blocks was randomized between respondents (trial_block_order); endorsing party shown only on pairs 6-10 for the endorsement arm. The article reports N = 3,950 for both countries; the file has 3,968. | 2026-10-09 | |
bansak_2021_austerity_spain |
Conjoint table (irw_conjoint_2). choice = preferred package in the pair (forced, no opt-out; wording not quoted in the article); rating = ‘If you could vote on each of these options in a referendum, how likely is it that you would vote in favor or against each of the options?’ 1 = vote definitely against, 10 = vote definitely in favor (translated). Eight attributes are percentages 0-30 in steps of 2, stored as the source numbers (cuts negative); the order of the spending and tax blocks was randomized between respondents (trial_block_order); endorsing party shown only on pairs 6-10 for the endorsement arm. The article reports N = 3,950 for both countries; the file has 3,968. | 2026-10-09 | |
bansak_2023_asylum_2022 |
choice = which applicant the respondent would prefer be allowed to stay (forced, no opt-out); rating 1 (absolutely send back) to 7 (definitely allow to stay). Attribute order randomized per respondent but not recorded; no restrictions documented. ratebin, cosmo_index, nat_index, Qualtrics IDs and 10 respondents with no attribute levels dropped (14,966 vs the paper’s 14,976). | #2887 | 2026-10-07 |
bansak_2024_algorithm_dm_crime |
Conjoint table (irw_conjoint_2). choice = which decision-maker in the pair the respondent would prefer making decisions (forced choice, no opt-out; paraphrase); rating = ‘On a scale from 1 to 7, where 1 indicates VERY BAD and 7 indicates VERY GOOD, how would you rate the performance of [each of these DMs]?’. Rates (10-50%) uniformly and independently randomized; DM types set by a between-respondent arm (trial_matchup), developer (algorithms only) and locale randomized per respondent. Three respondents with unsaved attribute values dropped; N = 4,784 matches the paper’s Table 1. | 2026-10-09 | |
bansak_2024_algorithm_dm_loans |
Conjoint table (irw_conjoint_2). choice = which decision-maker in the pair the respondent would prefer making decisions (forced choice, no opt-out; paraphrase); rating = ‘On a scale from 1 to 7, where 1 indicates VERY BAD and 7 indicates VERY GOOD, how would you rate the performance of [each of these DMs]?’. Rates (0-10% in steps of 0.5) uniformly and independently randomized; DM types set by a between-respondent arm (trial_matchup), developer (algorithms only) and locale randomized per respondent. Three respondents with unsaved attribute values dropped; N = 4,240 matches the paper’s Table 1. | 2026-10-09 | |
bansak_2025_odd_profiles_candidate |
Conjoint table (irw_conjoint). choice = preferred candidate of two (forced choice; wording not published, paraphrase). 60 tasks in four 15-task within-subject conditions (trial_condition): Normal excludes odd attribute combinations, the others allow incongruent and/or nonsensical ones, so estimators must account for the condition-specific restrictions; level shares are unequal. Lab sample (145 of 147 participants in this file); no respondent covariates deposited. | 2026-10-07 | |
bansak_2025_odd_profiles_immigrant |
Conjoint table (irw_conjoint). choice = preferred immigrant of two (forced choice; wording not published, paraphrase). 60 tasks in four 15-task within-subject conditions (trial_condition): Normal excludes odd attribute combinations, the others allow incongruent and/or nonsensical ones, so estimators must account for the condition-specific restrictions; level shares are unequal. Lab sample (142 of 147 participants in this file); no respondent covariates deposited. | 2026-10-07 | |
bard_2018_neuroenhancement |
Conjoint table (irw_conjoint). Factorial vignette experiment (contrastive vignettes), one employment and one education vignette per respondent (task 1/2 = context, display order randomized and not recorded). rating: ‘In (name)’s shoes, would you make the same choice?’, 11 points stored 0-10 (article: -5 to +5), higher = more willing; end labels not given. Four 2-level factors with the authors’ labels; protagonist gender was shown through a first name. 11 countries pooled (cov_country); no respondent ID in the deposit (row = respondent). | 2026-10-09 | |
bari_2025_deposit_inclusion |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical deposit products for slum dwellers (5 attributes) with a no-product option: choice = the option ranked first (0 on both profiles in 45 of 1,250 tasks where no product was ranked first, so opt_out = yes), choice_forced = the preferred of A and B, rating_rank = the profile’s rank among A, B and no product (1 = most preferred). The rank reading of the deposit’s Choice code is inferred from its agreement with the authors’ Internal/External choice columns; wording absent. Fixed blocked design (10 choice sets, 5 per respondent); A and B are not exchangeable. | 2026-10-12 | |
barnes_2024_tax_levers |
Conjoint table (irw_conjoint_2). choice = ‘If the government was only going to make one of these changes, which would you prefer?’ Option A / Option B; opt-out: ‘I think both of these changes are equally good or bad’ (2,911) and ‘Don’t know’ (1,709) are tasks with no chosen profile. One task per respondent; profile texts (headline, description, change, argument) assembled from the deposited prompt file; direction, revenue size (£10bn only for the five large taxes) and argument arm are task-level. N = 9,713, as in the article. | 2026-10-09 | |
barnett_2024_journal_choice |
Conjoint table (irw_conjoint). DCE: choice = “Select your preferred journal” between Journal A and Journal B, no opt-out. Task 1 is a dominant set, tasks 2-9 the respondent’s block of a fixed blocked efficient design, task 10 a retest of task 2 (or 3) with the same A/B order (inferred). Levels transcribed from the authors’ choice-set images because the design files are not deposited. Two designs (trial_design 1 = first sample of 96, 2 = second sample of 520) and two scenarios (trial_scenario). | 2026-10-09 | |
barnett_2025_politicians_morocco |
Conjoint table (irw_conjoint). Single-profile design: choice = would the respondent re-elect the politician in the profile (1 yes / 0 no; a no is the outside option, so opt-out yes); rating_competence = competence rating 1 (least) to 10 (most competent); wording not deposited (codebook descriptions). Attribute levels are the authors’ English labels; the survey’s display language and the selection of these 927 of 1,800 survey respondents are not documented in the deposit. Rows with neither outcome (7) omitted. Licence CC BY-NC 4.0. | 2026-10-08 | |
barnett_2025_quota_politicians |
Conjoint table (irw_conjoint). Factorial vignette (one per respondent): three 1-10 ratings of the described MP’s competence, whether a pawn of party elites, and cooperativeness (pre-registered wording; endpoint labels undocumented; higher = more of the named trait). The politician factor has 3 arms by design (man / woman district-elected, woman quota-elected; no quota man) with probabilities 25/25/50. Don’t know/refused are NA; 13 respondents with no rating dropped (914 of 927). cov_manip_correct is the post-vignette quota manipulation check (9 = don’t know/refused). | 2026-10-08 | |
barnett_2026_morocco_candidates |
Conjoint table (irw_conjoint). choice = does the respondent prefer this candidate over the paired one; choice_men / choice_women = would men / women in the respondent’s community prefer this candidate (codebook wording, paraphrased); each one per task, no documented opt-out, NA on both profiles when unanswered (302, 110, 92 of 3,492 tasks). rating = competence, 1 (least competent) to 10 (most competent). Question order randomized (trial_outcome_order_1/_2). CC BY-NC 4.0. | 2026-10-08 | |
barnfield_2023_policy_discounting |
Conjoint table (irw_conjoint_2). Paired forced choice between two life-saving policies (lives saved, policy area, cost, payoff delay); no opt-out. Respondents saw either a qualitative or a quantitative (years) delay; only the qualitative version is included because the years shown are not stored. 1,168 of 2,405 respondents; 38 respondents have 7 tasks and 1 has 4. | 2026-10-12 | |
barnum_2026_bystander_video |
Conjoint table (irw_conjoint_2). Factorial video vignette, one per respondent (US national online sample): offender race x victim race, 2x2, equal allocation. Seven 5-point ratings: likelihood of six bystander actions (1 = very likely, 5 = very unlikely, so lower = more likely) and perceived danger of physical intervention (5 = very unsafe); item wording paraphrased from truncated labels. No respondent id (id = row number); level text is the authors’ condition labels because the videos are not deposited. The paper’s analytic N is 1,901 (complete cases); 2,028 respondents are kept here. | 2026-10-14 | |
bayer_2026_migration_destinations |
Conjoint table (irw_conjoint). choice: “Generally speaking, if you had to choose, which one of these two options w…” (source label truncated; the intro asks which of a pair of countries the respondent would consider better to travel to and start a new life); forced, no opt-out. Attribute order randomized per respondent (attrpos_). 124 completes without conjoint answers omitted; priming-experiment arm in trial_priming; cov_gender NA includes 17 ‘Other / Do not want to answer’. | 2026-10-08 | |
beaucousin_2025_green_investments |
Conjoint table (irw_conjoint). choice: which of two proposed green investments the respondent preferred (paraphrase; wording not deposited), forced choice, no opt-out recorded. Dutch level text as shown; attribute order randomized per respondent (attrpos_). 33 unanswered tasks dropped; no covariates deposited. 1,541 respondents, as in the deposit description. | 2026-10-08 | |
bechard_2024_bond_referendums |
Conjoint table (irw_conjoint_2). Single-profile vote: each bond seen on a hypothetical Rhode Island ballot of 1, 3 or 5 bonds gets a yes/no vote (choice; a no is the opt-out). Attributes are bond type, price in millions of dollars and ballot size; ballot position is kept raw in trial_order_code because the source repeats positions. The authors’ sample filters are applied (731 of 851 respondents); the paper’s N was not checked. | 2026-10-14 | |
bechtel_2013_climate_agreements_de |
Conjoint table (irw_conjoint). choice = which of the two climate agreements the respondent prefers (forced choice, no opt-out); rating = “If you could vote on each of these agreements in a referendum, how likely is it that you would vote in favor or against each of the agreements?” from 1 (definitely against) to 10 (definitely in favor), stored unrescaled. One of four country tables (Germany); costs and sanctions are this country’s displayed amounts, other levels are the article’s English text; dimension order was randomized but its code is undocumented and not kept; profile position is inferred from row order. | 2026-10-07 | |
bechtel_2013_climate_agreements_fr |
Conjoint table (irw_conjoint). choice = which of the two climate agreements the respondent prefers (forced choice, no opt-out); rating = “If you could vote on each of these agreements in a referendum, how likely is it that you would vote in favor or against each of the agreements?” from 1 (definitely against) to 10 (definitely in favor), stored unrescaled. One of four country tables (France); costs and sanctions are this country’s displayed amounts, other levels are the article’s English text; dimension order was randomized but its code is undocumented and not kept; profile position is inferred from row order. | 2026-10-07 | |
bechtel_2013_climate_agreements_uk |
Conjoint table (irw_conjoint). choice = which of the two climate agreements the respondent prefers (forced choice, no opt-out); rating = “If you could vote on each of these agreements in a referendum, how likely is it that you would vote in favor or against each of the agreements?” from 1 (definitely against) to 10 (definitely in favor), stored unrescaled. One of four country tables (United Kingdom); costs and sanctions are this country’s displayed amounts, other levels are the article’s English text; dimension order was randomized but its code is undocumented and not kept; profile position is inferred from row order. | 2026-10-07 | |
bechtel_2013_climate_agreements_us |
Conjoint table (irw_conjoint). choice = which of the two climate agreements the respondent prefers (forced choice, no opt-out); rating = “If you could vote on each of these agreements in a referendum, how likely is it that you would vote in favor or against each of the agreements?” from 1 (definitely against) to 10 (definitely in favor), stored unrescaled. One of four country tables (United States); costs and sanctions are this country’s displayed amounts, other levels are the article’s English text; dimension order was randomized but its code is undocumented and not kept; profile position is inferred from row order. | 2026-10-07 | |
bechtel_2017_bailouts |
Conjoint table (irw_conjoint). rating: “If you could vote over each proposal in a direct-democratic vote, how likely would you vote against or in favour of each proposal?” stored 0 (vote definitely against) .. 6 (vote definitely in favour), as in the source (article: 1-7). choice: forced choice between the two proposals (Online Appendix), one chosen per comparison. Up to 4 comparisons per respondent; 6 independently randomized attributes, attribute order randomized per respondent (not recorded). Qualtrics IDs re-keyed; no weight. | 2026-10-08 | |
bechtel_2020_income_reforms |
Conjoint table (irw_conjoint). choice = “Which policy do you prefer?” (forced, no opt-out); rating = likelihood of voting for each proposal in a referendum, 1 = definitely against to 10 = definitely in favor. Fully randomized, attribute order randomized per respondent (not recorded). Respondents have 1-4 tasks (N matches the article). cov_party_id is missing for respondents who named no party; cov_survey_weight is the authors’ entropy-balancing weight. Derived dummies and groupings dropped. | 2026-10-08 | |
bechtel_2020_income_reforms_pilot |
Conjoint table (irw_conjoint). Pilot study with income changes in percent. choice = preferred policy of each pair (forced, no opt-out; the pilot’s wording is not deposited, the main study asked “Which policy do you prefer?”). Task and profile are inferred from row order (the file has neither). No rating or covariates in the deposit. | 2026-10-08 | |
bechtel_2020_income_reforms_validation |
Conjoint table (irw_conjoint). Validation study: choice = preferred proposal (or candidate) of each pair, forced, no opt-out; wording of this instrument not deposited. trial_version is the randomly assigned framing (generic policy, health care policy, trade policy, candidate platform), constant within respondent. Profile position is inferred from row order within each recorded task. 2,003 of the 2,300 fielded respondents are in the file (matches the article’s analysis N). | 2026-10-08 | |
bechtel_2022_climate_multilateralism |
Conjoint table (irw_conjoint_2). Paired climate-policy scenarios (8 tasks) with own-country and other-economies blocks of cost per month, cost path, adaptation and mitigation shares; choice = ‘Which of these scenarios do you prefer?’ (forced, no neither option); rating_effectiveness = expected effectiveness 1-10 (US respondents only, wording not in the deposit). Cost levels are shown in the respondent’s currency; attribute order and own/other block order randomized per respondent (attrpos_, trial_block_order). 8,075 respondents vs 8,080 in the paper. | 2026-10-11 | |
bechtel_2022_disaster_relief |
Conjoint table (irw_conjoint). rating is the share of a fixed $10 million relief budget (in $ million, 0-10) given to each of two counties; the two ratings of a task sum to 10 (no choice column, no opt-out). Presidential partisanship is randomized per task and the same for both counties; county attribute order randomized per respondent (attrpos_). cov_survey_weight is the authors’ entropy-balancing weight used in the main results. | 2026-10-09 | |
becker_2025_basic_income |
Conjoint table (irw_conjoint). Five paired UBI-proposal tasks with six German-text attributes; choice = preferred proposal (Vorschlag A/B; paraphrase, forced choice, no opt-out), rating = support for each proposal, 0 ‘Auf keinen Fall’ to 10 ‘Voll und ganz’ (paraphrase; integer from the answer label). Respondent IDs re-keyed; N = 3,076, no rows dropped. | 2026-10-09 | |
becker_2025_tax_packages |
Conjoint table (irw_conjoint). Two forced choices per pair of tax-reform packages (“To which of the reform packages do the following statements apply more?”): choice_inequality = “Contributes to the reduction of economic inequality.”, choice_growth = “Contributes to the growth of the economy.”; no opt-out and no overall preference question. Levels are the English text of the article’s Table 1 (respondents saw German); tax-row order was varied but is not recorded. The deposit holds 921 respondents, fewer than the roughly 1,214 the article implies. | 2026-10-07 | |
becker_2026_housing_text |
Conjoint table (irw_conjoint_2). One profile per respondent (a text profile); choice = 1 for Yes to ‘if it were put to a public vote, would you support it being built in your neighborhood?’ (yes/no; a single-profile vote, so No is the opt-out). Three attributes randomized independently and equally likely (12 combinations); the other presentation mode is a separate table. Authors’ exclusions (bot score, under 2 minutes, attention check) applied; N = 1,706 as in the article’s models; ZIP codes, panel ids and free text removed. | 2026-10-09 | |
begum_2026_hama |
Self-rated (survey) administration of the clinician-rated HAM-A, as deposited. | 2026-10-11 | |
beisermcgrath_2019_paris_cn |
Conjoint table (irw_conjoint). choice = “Which of the two policies do you prefer?” (forced choice between Policy A and B, no opt-out), 5 tasks of 6 attributes; respondents first saw one of six randomized information treatments (trial_info_treatment). Survey in Chinese; attribute text is an English reconstruction and the USA partner levels omit their emission-share figure. No restrictions documented; level probabilities not documented. | 2026-10-09 | |
beisermcgrath_2019_paris_us |
Conjoint table (irw_conjoint). choice = “Which of the two policies do you prefer?” (forced choice between Policy A and B, no opt-out), 5 tasks of 6 attributes; respondents first saw one of six randomized information treatments (trial_info_treatment). No restrictions documented; level shares observed unequal. | 2026-10-09 | |
beisermcgrath_2019_revenue_de |
Conjoint table (irw_conjoint_2). Forced choice between two carbon tax proposals (no opt-out) and a 1-7 support rating of each (7 = most supportive, inferred from the authors’ 5-7 = supportive coding; wording not in the deposit). The revenue-use attribute was shown to a random half of respondents, otherwise ‘(not shown)’ (trial_revenue_shown). Germany only; N = 3620 respondents, matching the paper. | 2026-10-11 | |
beisermcgrath_2019_revenue_us |
Conjoint table (irw_conjoint_2). Forced choice between two carbon tax proposals (no opt-out) and a 1-7 support rating of each (7 = most supportive, inferred from the authors’ 5-7 = supportive coding; wording not in the deposit). The revenue-use attribute was shown to a random half of respondents, otherwise ‘(not shown)’ (trial_revenue_shown). United States only; N = 3640 respondents, matching the paper. | 2026-10-11 | |
beisermcgrath_2022_vehicle_delhi |
Conjoint table (irw_conjoint). choice = which of two policy bundles the government should adopt and implement (forced, one per task); rating = 1-7 support for each bundle (1 strongly oppose, 7 strongly support). No opt-out; four attributes, randomized independently by the article, attribute order randomized once per respondent (not recorded). Profile position inferred from row order. Beijing sample of the same deposit not included (screen text unknown). N 1,501 (article: 750 per wave). | 2026-10-09 | |
beisermcgrath_2024_energy_policy |
Conjoint table (irw_conjoint). Outcomes: forced choice “Which policy do you prefer?” and “How much do you support or oppose |
2026-10-08 | |
belyakova_2026_bfi2s |
Four suspected duplicate submissions flagged by the depositors (same contact e-mail, identical answers to all 65 items) are dropped, so N = 143 rather than the 147 analysed. Reverse-keyed items are stored as answered. The same respondents are in belyakova_2026_pil and belyakova_2026_maas. | 2026-10-07 | |
belyakova_2026_maas |
Four suspected duplicate submissions flagged by the depositors are dropped (N = 143, not 147). 1 = almost always … 6 = almost never, so higher = more mindful. | 2026-10-07 | |
belyakova_2026_pil |
Four suspected duplicate submissions flagged by the depositors are dropped (N = 143, not 147). Items are on the 1-7 bipolar scale as recorded; the reversed items (2, 5, 6, 7, 10, 13, 14, 15, 18, 19, 20 per the codebook) are stored as answered. | 2026-10-07 | |
bennour_2022_location_decision |
Conjoint table (irw_conjoint). choice = ‘In which municipality would you prefer to live?’ (I pick municipality A/B), forced, no opt-out; rating = ‘How likely is it that you would choose to live in Municipality A [B]?’ 0-10, 10 = best. Eight binary attributes in a fixed grid. Fielded in English, German, French, Italian, Spanish and Portuguese; English level text stored, language in cov_language. Finished respondents only: 1,589 (codebook 1,596). | 2026-10-09 | |
berens_2024_crime_measures |
Conjoint table (irw_conjoint). Agreement 1-5 with ‘Para reducir y controlar el crimen |
2026-10-08 | |
berens_2024_demilitarisation |
Conjoint table (irw_conjoint). Agreement 1-5 with a proposal ‘De acuerdo con una propuesta reciente |
2026-10-08 | |
berens_2024_police_recruits |
Conjoint table (irw_conjoint). Two ratings of the same sentence ‘Suponga que el gobierno recluta a más policías municipales |
2026-10-08 | |
berens_2024_taxes_spending |
Conjoint table (irw_conjoint). Agreement 1-5 (Muy en desacuerdo … Muy de acuerdo) with ‘¿Qué tan de acuerdo estaría usted con pagar más impuestos |
2026-10-08 | |
bergquist_2020_gnd_energy |
Conjoint table (irw_conjoint). Experiment 2. choice: ‘Which policy package would you prefer?’ between Policy Package A and B, forced, 3 tasks. A category with no policy is ‘(not shown)’. Attribute order randomized per respondent (attrpos_). Respondents who saw an earlier attribute version (Sponsor instead of Transportation) dropped by the authors (2,427 of 2,476). Authors’ re-raked weight kept. | 2026-10-09 | |
bergquist_2020_gnd_social |
Conjoint table (irw_conjoint). Experiment 1. choice: ‘Which policy package would you prefer?’ between Policy Package A and B, forced, 3 tasks. A category with no policy is ‘(not shown)’. Attribute order randomized per respondent (attrpos_). Between-respondent benefits framing in trial_benefits. Authors’ raked weight kept. | 2026-10-09 | |
bergquist_2023_covid_climate_ca |
Conjoint table (irw_conjoint). choice: ‘Which policy package would you prefer?’ Policy A/B, no opt-out (a few skipped choices are NA); rating: support for each package 0 (no support) - 100 (full support). Attribute order randomized per respondent (attrpos_). French-language respondents’ level text is stored in English. ZIP codes and ResponseIds dropped. | 2026-10-09 | |
bergquist_2023_covid_climate_us |
Conjoint table (irw_conjoint). choice: ‘Which policy package would you prefer?’ Policy A/B, no opt-out (a few skipped choices are NA); rating: support for each package 0 (no support) - 100 (full support). Attribute order randomized per respondent (attrpos_). 1,695 reached the conjoint (article n = 1,695), 1,642 answered at least one task. ZIP codes and ResponseIds dropped. | 2026-10-09 | |
bergquist_2023_covid_climate_us2 |
Conjoint table (irw_conjoint). choice: ‘Which policy package would you prefer?’ Policy A/B, no opt-out (a few skipped choices are NA); rating: support for each package 0 (no support) - 100 (full support). Attribute order randomized per respondent (attrpos_). Follow-up survey; 1,040 reached the conjoint, 1,022 answered. ZIP codes and ResponseIds dropped. | 2026-10-09 | |
berinsky_2020_immigrant_admission |
choice = “Please choose one: Immigrant 1 / Immigrant 2” (forced, no opt-out); rating 1-7 per immigrant (1 = the US should absolutely not admit, 7 = definitely admit). Atheist immigrants always ‘Never’ attend services; attribute order randomized but not recorded. Self-monitoring index and unanswered tasks dropped; article N 1,571 vs 1,627 respondents answering in the deposit. | #2887 | 2026-10-07 |
bernauer_2022_ev_purchase_zurich |
Conjoint table (irw_conjoint). choice: “Wenn Sie sich für eines der beiden Autos entscheiden müssten, welches Auto würden Sie kaufen?” (forced, no opt-out); rating: “Würden Sie Auto A/B kaufen?” Ja = 1, Nein = 0. Levels independently randomized; German text as stored. trial_group records the source’s two survey versions (Conjoint Pos / Neg), whose difference is not documented. | 2026-10-08 | |
berntzen_2025_political_violence |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (US, Jan 2024): which party’s attendees beat which party’s disruptors at a political rally (no party labels = ‘(not shown)’; own party beats opposing party; opposing party beats own party) x severity of the disruption (shouting vs blocking the venue entrance). rating = how justified the beating is, 1 = completely unjustified, 7 = completely justified. Kept only the authors’ 1,521 attentive partisans (of 2,030); level text paraphrased from the codebook; the survey weight is kept although the authors analyse unweighted. | 2026-10-14 | |
berz_2022_candidate_selection |
Conjoint table (irw_conjoint). choice: “Welchen Kandidat bevorzugen Sie als Wahlkreiskandidaten für die Bundestagswahl?” (or “… für die Landtagswahl?”), forced choice between two candidates, 10 tasks (two blocks of 5 in random order; trial_parliament). 7 attributes in German; attribute row order randomized per respondent (attrpos_). Qualtrics IDs re-keyed, free text dropped; left-right end points exported as text mapped to 0/10. | 2026-10-08 | |
berz_2026_crisis_accountability |
Conjoint table (irw_conjoint). Three forced choices per pair of crisis news reports (options ‘The government in Crisis 1/2’, no opt-out): choice ‘Which government do you prefer to re-elect?’, choice_handled ‘Which government has handled their crisis better?’, choice_responsible ‘Which government is more responsible for the severity of their crisis?’. Attribute order fixed; levels as inserted into the report text. 12 partial respondents kept (the article’s N = 812 completes). Prolific IDs dropped. | 2026-10-09 | |
bhandari_2023_trade_partners |
Conjoint table (irw_conjoint). Each pair asked which deal the respondent would be more likely to choose and which would more likely end in breach, with Deal 1 / Deal 2 / Neither / Both allowed, so both are stored as per-profile 0/1 ratings (rating, rating_breach; higher rating_breach = more breach risk); refusals and don’t-knows are blank. Level text is the English codebook text; respondents’ display language is not documented. Attribute order varied by one of four undocumented schemes (trial_attribute_order). 7 respondents with no answers dropped; 8 implausible ages (above 100) blanked. | 2026-10-07 | |
bhandari_2026_clientelism_consumers |
Conjoint table (irw_conjoint). choice = “Which deal are you more likely to choose?” and choice_wrong = “Which deal is more likely to go wrong?”, each with Both and Neither options: those answers are 0 on both profiles, raw answer in trial_choice_response / trial_wrong_response (the authors count Both as chosen). rating = trust in the deal, 0-10 slider. Levels are the authors’ English labels; party is stored even for Businessman owners, display undocumented. Bank account rounds 1-2, TV rounds 3-4 (trial_product). 696 unanswered tasks and never-reached respondents dropped. | 2026-10-08 | |
bhatia_2025_afghan_leaders |
Conjoint table (irw_conjoint). choice: “Given a choice between these two profiles, which person would you prefer as a leader?” (forced A/B); rating: suitability of each profile, 1 = absolutely unsuitable to 5 = absolutely suitable (“an ideal leader for you”). Restrictions undocumented but some levels have unequal shares (Military, Kandahar, age 28). All four pre-conjoint prime arms kept (trial_prime); place-name free text dropped. N = 2,485 matches the codebook. | 2026-10-07 | |
binningsbo_2025_party_platforms |
Conjoint table (irw_conjoint). choice = which of two party platforms (Parti A/Parti B) the respondent chose (vote choice; exact wording not deposited), forced choice, 5 tasks, task 5 repeats task 1 with profiles swapped. No documented restrictions; all 360 combinations occur. Tasks left unanswered (9996) omitted; deposit N 3,317 (3,300 answering) vs article 2,962 native Norwegians. | 2026-10-09 | |
bischof_2026_simple_language_high |
Conjoint table (irw_conjoint_2). choice = which of the two MPs the respondent preferred (paraphrase), forced choice, no opt-out, up to 3 pairs. Task and profile are inferred from row blocks (exactly one chosen per rebuilt pair), so position effects are not identified. Attribute text is the authors’ English labels for a German survey; position (for/against/neutral) and sophistication (simple/sophisticated) describe a statement not included. The authors’ design object lists two prohibited party-position cells that nonetheless occur; N against the article not checked. | 2026-10-09 | |
bischof_2026_simple_language_low |
Conjoint table (irw_conjoint_2). choice = which of the two MPs the respondent preferred (paraphrase), forced choice, no opt-out, up to 3 pairs. Task and profile are inferred from row blocks (exactly one chosen per rebuilt pair), so position effects are not identified. Attribute text is the authors’ English labels for a German survey; position (for/against/neutral) and sophistication (simple/sophisticated) describe a statement not included. The authors’ design object lists two prohibited party-position cells that nonetheless occur; N against the article not checked. | 2026-10-09 | |
biswas_2024_digital_center_quality |
cov_age is an age-band code, not age in years: 1 = “?20”; 2 = “21~30”; 3 = “31~40”; 4 = “41~50”; 5 = “51~60”; 6 = “?61”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
blackman_2021_tunisia_household |
Conjoint table (irw_conjoint). choice = pick one of two candidates (wording not deposited), forced, 4 tasks; trial_election records the randomized municipal vs national-parliament prompt. Level text is the stored English (interviews in Tunisian Arabic). 2,571 unanswered tasks dropped; 709 of 1,193 respondents answered at least one (the paper’s main analysis uses those answering all 4). | 2026-10-09 | |
blackman_2021_tunisia_online |
Conjoint table (irw_conjoint). choice = pick Candidate A or B (wording not deposited), forced, 5 tasks, all answered. Level text as stored, without the authors’ renames (Jabha Shabiyya = Popular Front, ‘Studied outside of Tunis’); questionnaire in Arabic (540) or English (34), cov_language; YouGov weight kept. | 2026-10-09 | |
blackman_2025_aid_tunisia |
Conjoint table (irw_conjoint). Forced choice between two development projects (‘Which project do you prefer for your municipality?’), no opt-out, 4 tasks; 36 unanswered tasks omitted. Levels are the paper’s English Table 1 text mapped from the deposit’s short labels. Profile position inferred from the deposit’s stacked row order (which project was shown first is assumed). Attribute order randomized per respondent, not recorded. Municipality ID dropped. | 2026-10-08 | |
blais_2026_losers_consent |
Conjoint table (irw_conjoint_2). Five single-outcome vignettes per respondent; rating = how acceptable the respondent would personally find the election outcome, 0 (not acceptable at all) to 10 (completely acceptable); rating_protest = likelihood of joining a protest contesting it, 0-10, asked only when acceptance was below 5 (NA otherwise; the authors code those 0). Seat shares, coalition and Chancellor are generated under plausibility rules (restrictions); trial_flag_bug marks the 0.41% of outcomes the authors exclude as generation errors. Wording is paraphrased from the companion article; cabinet-seat shares not stored. 5,374 respondents (the paper’s 5,370 drops 4 with a missing fifth answer); IP addresses, GPS coordinates and panel IDs removed. | 2026-10-09 | |
bland_2024_gcos_autonomy |
Each of the 17 GCOS vignettes has three responses (autonomy, control, impersonal), split here into three tables by orientation as GCOS is scored; item GCk_A answers the same vignette as GCk_C and GCk_I. Item codes are the deposit’s. No codebook in the deposit. | 2026-10-09 | |
bland_2024_gcos_control |
Each of the 17 GCOS vignettes has three responses (autonomy, control, impersonal), split here into three tables by orientation as GCOS is scored; item GCk_C answers the same vignette as GCk_A and GCk_I. Item codes are the deposit’s. No codebook in the deposit. | 2026-10-09 | |
bland_2024_gcos_impersonal |
Each of the 17 GCOS vignettes has three responses (autonomy, control, impersonal), split here into three tables by orientation as GCOS is scored; item GCk_I answers the same vignette as GCk_A and GCk_C. Item codes are the deposit’s. No codebook in the deposit. | 2026-10-09 | |
bland_2024_mfq20 |
The deposit’s MFT1-22 follow the MFQ-20 order; the two catch items (MFT6 ‘good at math’, MFT17 ‘better to do good than bad’) are removed, identified by position and by their response profile (lowest and highest item means). Responses are stored 1-6, i.e. the MFQ’s 0-5 scales plus one. No codebook in the deposit. | 2026-10-09 | |
blaszczyk_2025_siechnice_development |
Conjoint table (irw_conjoint_2). Discrete choice experiment: respondents chose between two development visions for their town (5 attributes, Polish card text) or a fixed opt-out ‘none of the above, natural development’ (no attributes, so no profile row; opt_out = yes, 394 of 4,608 tasks chose it). Fixed blocked design from R idefix (24 cards, 5 blocks of 12 tasks; trial_block). Levels are the card sentences from the questionnaire PDF, mapped from the data file’s short keyword labels. Demographics are not in the deposit. | 2026-10-12 | |
blaszyk_2023_copsoq_js |
Stored as in the source: 1 = very satisfied … 4 = very dissatisfied, so low values are the agreeing/positive end. Missing code 9 dropped. | 2026-10-07 | |
blaszyk_2023_copsoq_se |
Stored as in the source: 1 = fits completely … 4 = does not fit at all, so low values are the agreeing/positive end. Missing code 9 dropped. | 2026-10-07 | |
blaszyk_2023_dtt |
266 cells that the source had filled with item means (fractional values shared by 29 respondents) are dropped. | 2026-10-07 | |
blaszyk_2023_gts |
Stored as in the source: 1 = strongly agree … 6 = strongly disagree, so low values are the agreeing/positive end. Missing code 9 dropped. | 2026-10-07 | |
blaszyk_2023_olbi |
Stored as in the source: 1 = agree … 4 = disagree, so low values are the agreeing/positive end. Missing code 9 dropped. | 2026-10-07 | |
blaszyk_2023_ptps |
242 cells that the source had filled with item means (fractional values shared by 19 respondents) are dropped, as are missing codes (9). | 2026-10-07 | |
blauberger_2025_sanctions_bg |
Conjoint table (irw_conjoint). choice: forced choice between two EU sanction proposals (no opt-out); rating: support for each proposal, 1 (no support) to 7 (high support). Task and profile inferred from row order (verified: exactly one chosen profile per pair). Attribute text is the deposit’s short English value labels; question wording paraphrased (article not seen). | 2026-10-09 | |
blauberger_2025_sanctions_hu |
Conjoint table (irw_conjoint). choice: forced choice between two EU sanction proposals (no opt-out); rating: support for each proposal, 1 (no support) to 7 (high support). Task and profile inferred from row order (verified: exactly one chosen profile per pair). Attribute text is the deposit’s short English value labels; question wording paraphrased (article not seen). | 2026-10-09 | |
blauberger_2025_sanctions_pl |
Conjoint table (irw_conjoint). choice: forced choice between two EU sanction proposals (no opt-out); rating: support for each proposal, 1 (no support) to 7 (high support). Task and profile inferred from row order (verified: exactly one chosen profile per pair). Attribute text is the deposit’s short English value labels; question wording paraphrased (article not seen). | 2026-10-09 | |
blumel_2018_mrs |
The deposit’s MRS1B-MRS11B (a 0/1 per symptom, summed in TRTotal) are not shipped: what was asked is not documented in the file or record. | 2026-10-09 | |
blumenau_2024_representation_de |
Conjoint table (irw_conjoint). Text conjoint: choice of which of two described politicians would better represent the respondent (‘Welcher dieser Politiker würde Sie ihrer Meinung nach besser in der Politik vertreten?’), forced, no opt-out. Sentence order randomized once per respondent (attrpos_); the policy issue is shared within a pair and only the stance varies; ethnicity and sexuality levels have unequal probabilities by design. Attention-check failures and speeders are kept, flagged in cov_attention_pass and cov_duration_sec (the article drops them). | 2026-10-09 | |
blumenau_2024_representation_uk |
Conjoint table (irw_conjoint). Text conjoint: choice of which of two described politicians would better represent the respondent (‘Which of these politicians do you think would better represent you in politics?’), forced, no opt-out. Sentence order randomized once per respondent (attrpos_); the policy issue is shared within a pair and only the stance varies; ethnicity and sexuality levels have unequal probabilities by design. Attention-check failures and speeders are kept, flagged in cov_attention_pass and cov_duration_sec (the article drops them). | 2026-10-09 | |
blumenau_2024_representation_us |
Conjoint table (irw_conjoint). Text conjoint: choice of which of two described politicians would better represent the respondent (‘Which of these politicians do you think would better represent you in politics?’), forced, no opt-out. Sentence order randomized once per respondent (attrpos_); the policy issue is shared within a pair and only the stance varies; ethnicity and sexuality levels have unequal probabilities by design. Attention-check failures and speeders are kept, flagged in cov_attention_pass and cov_duration_sec (the article drops them). | 2026-10-09 | |
boissonneault_2022_migration_experts |
Conjoint table (irw_conjoint). Factorial vignette survey of 138 European migration professionals (web, English, Nov 2021-Jan 2022), 4 vignettes each from a blocked full 2^7 design (trial_block). rating_family/work/refugees/return: ‘Based on [the situation described by the vignette], compared to 2019, the number of family [work/return] migrants [refugees] from the Middle East & North Africa to Europe will be in 2030…’ stored as multiplication factor 0.2-5 (1 = No change); rating_compact: more or less difficult to achieve safe, orderly and regular migration, -3 much more to 3 much less difficult. Factor order one of two fixed orders, recorded in attrpos_. | 2026-10-09 | |
bonnenfant_2026_cyprus_peace |
Conjoint table (irw_conjoint_2). Five forced-choice tasks of two hypothetical Cyprus peace-settlement packages (five attributes), Greek and Turkish Cypriot residents pooled (cov_community); choice = package chosen (question wording not in the deposit, no opt-out). Task and profile inferred from row blocks (one chosen in every pair, verified); the diaspora sample is not included. | 2026-10-11 | |
borbath_2025_new_party_brand |
Conjoint table (irw_conjoint_2). Three tasks of two hypothetical new party/movement profiles (name, organization, action form) in Austria, Italy and Hungary; rating of support (stored 1-11, article says 0-10) and forced choice of the one more likely to be voted for. Task and profile inferred from row order (one chosen per consecutive pair); English labels; the German sample is not included (tasks not recoverable). | 2026-10-11 | |
borbath_2025_party_rebranding |
Conjoint table (irw_conjoint_2). Three tasks of two counterfactual versions of the respondent’s own party (name change, organization, action form) in Austria, Italy and Hungary; rating of support (stored 1-11, article says 0-10) and forced choice of the version more likely to be voted for. Name-change levels depend on the party’s current name; task and profile inferred from row order; Germany not included; respondents without a party omitted. | 2026-10-11 | |
borbath_2025_volunteer_mobilization |
Conjoint table (irw_conjoint_2). Factorial vignette, two per respondent: mobilizing actor (friends, local or national aid organization, local administration) x recipient (older people / people in financial difficulties) x displayed emotion (anger, fear, sadness, guilt, hope, neutral). Ratings: likelihood of volunteering (0-10, in rating) and of donating (stored 1-11, 1 = very unlikely); no opt-out. Countries (DE, IT, PL) pooled with cov_country. The displayed emotion matches or differs from the respondent’s own emotion by random assignment (trial_emotion_matches_own), so the emotion mix is not uniform. Actor text is from the article; recipient and emotion levels are the deposit’s data labels. | 2026-10-14 | |
bordeleau_2024_democracy_security |
Conjoint table (irw_conjoint). choice: which of two societies ‘you consider to be the best one for you; that is, the society in which you would be the most content’ (forced, no opt-out); rating: each society 0 (very bad) to 10 (very good). Seven two-level attributes, fully randomized; attribute order said to be randomized but not recorded. Australia, Canada and UK samples pooled (cov_country); derived dummies and Qualtrics IDs dropped; N = 3,033 matches the supplement. | 2026-10-09 | |
borry_2018_rule_following |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent, 2x2: the organizational rule is written or unwritten and consistently or inconsistently enforced. Outcome rating_rulebending is 0 = rule following, 1 = rule bending (higher = more bending); no opt-out. Level text is the stored value labels because the vignette wording was not available. Study 2 only (N = 150). | 2026-10-14 | |
bostouwen_2017_selfmanagement |
Conjoint table (irw_conjoint). Factorial vignette survey, one vignette per task (12 per provider); rating = ‘How likely is it that you will support this patient in self-management?’ 1 very unlikely to 5 very likely, rating_success = expected success 1-5, rating_certainty = ‘I am confident about my answers’ 1-5. Fixed blocked design (96 vignettes from a Fedorov main-effects design after removing unrealistic combinations, 8 blocks); levels are the authors’ English value labels for the Dutch vignette sentences; N = 140 as in the paper. | 2026-10-09 | |
bowen_2023_nuclear_strikes |
Conjoint table (irw_conjoint). choice = which strike option (Option 1 / Option 2) the respondent chose; wording not deposited; forced choice, no opt-out; 1,439 unanswered pairs omitted. Restricted randomization in the ‘Realistic’ tasks (nuclear Option 1 vs conventional Option 2, fixed counts of advantages/disadvantages), full randomization in the ‘Conjoint’ tasks (trial_randomization). Level text is the authors’ short labels; survey weights in cov_survey_weight. | 2026-10-07 | |
bracic_2023_judge_fairness |
Conjoint table (irw_conjoint). choice = ‘Which of these judges do you trust more to fairly hear this case on |
2026-10-08 | |
brazzill_2020_social_investment |
Conjoint table (irw_conjoint). choice = which policy package was selected as preferable (wording not deposited); task 3 is omitted: in the deposit its selections duplicate task 2’s although the profiles differ, so its real answers are not available (the authors’ AMCEs include it). 686 unanswered tasks of tasks 1-2 omitted. Randomization rules not documented. Levels are the authors’ short English codes (sq = status quo), not the Japanese text shown. | 2026-10-07 | |
brehm_2025_job_training_germany |
Conjoint table (irw_conjoint). choice: which of two hypothetical people the respondent picked in the job-training conjoint (question wording not deposited; paraphrase), forced choice with exactly one chosen per task, no opt-out. Attribute levels are the authors’ English translations; the authors standardized Age to 25/35/46/62. Record is unequally distributed (None about 70%). 1,882 respondents, matching the README. Year of birth kept as recorded (3 respondents born after 2005); the authors’ main sample (Western, ethnic Germans) is a subset. | 2026-10-08 | |
brehm_2025_job_training_korea |
Conjoint table (irw_conjoint). choice: which of two hypothetical people the respondent picked in the job-training conjoint (question wording not deposited; paraphrase), forced choice with exactly one chosen per task, no opt-out. Attribute levels are the authors’ English translations; the authors standardized Age to 25/35/46/62. Record is unequally distributed (None about 70%). 1,768 respondents, matching the README. Year of birth kept as recorded (12 respondents born after 2005). | 2026-10-08 | |
brehm_2025_politicians_germany |
Conjoint table (irw_conjoint). Forced choice between two hypothetical politicians (no opt-out); question wording not deposited. Attribute levels are issue statements in the survey language. Sample as prepared by the authors (quota sample with attention check plus Eastern German oversample); one of three samples from the same study, each its own table. | 2026-10-08 | |
brehm_2025_politicians_korea |
Conjoint table (irw_conjoint). Forced choice between two hypothetical politicians (no opt-out); question wording not deposited. Attribute levels are issue statements in the survey language. Sample as prepared by the authors; one of three samples from the same study, each its own table. | 2026-10-08 | |
brehm_2025_politicians_nk_migrants |
Conjoint table (irw_conjoint). Forced choice between two hypothetical politicians (no opt-out); question wording not deposited. Attribute levels are issue statements in the survey language. Sample as prepared by the authors; one of three samples from the same study, each its own table. | 2026-10-08 | |
breitenstein_2019_corrupt_mayors |
Conjoint table (irw_conjoint). rating: probability of voting for each candidate, 0 (would never vote for) to 10 (would definitely vote for), stored 0-10 (source 1-11); choice: forced choice ‘Imagina que hay elecciones en tu municipio. ¿Qué alcalde o alcaldesa preferirías…’ (García/Martínez), no opt-out. Fully randomized, no restrictions; attribute order randomized per respondent but not recorded. The authors drop fast tasks (page time < 15/10/9 s for tasks 1/2/3, in trial_page_seconds); all tasks kept here. | 2026-10-07 | |
bremer_2022_social_spending |
Conjoint table (irw_conjoint_2). Paired conjoint, 5 tasks x 2 profiles: six social-spending attributes (training for the unemployed, childcare, education, unemployment benefits, child benefits, pensions), each decrease, no change or increase. Packages are zero-sum: the number of increases equals the number of decreases (141 of 729 combinations occur), so AMCEs must respect the restriction. choice is the forced pick between the two packages (no opt-out); rating is 0-10 support for each package. Attribute level text is taken from the article (Table 2); the deposit stores tags. Germany, Italy and UK pooled via cov_country; cov_survey_weight is NA for 116 respondents. | 2026-10-14 | |
breyer_2025_social_status |
Conjoint table (irw_conjoint). choice: which of two persons stands higher in Swiss society today (forced, no opt-out); rating: each person’s position on the 1 (unten) to 10 (oben) status ladder; wording German. Choice and rating files have different respondent numbering and were joined by each respondent’s unique set of 10 profiles (2,319 matched; 63 respondents have choice only, 117 rating only, these without covariates). Attribute text is the authors’ English labels; occupation and education combine only in fixed ways (observed restriction). | 2026-10-08 | |
brooks_2026_mining_argentina |
Conjoint table (irw_conjoint). choice: “¿Qué proyecto es mejor?” (Which project is better?), Proyecto A / Proyecto B, no opt-out; rating: “Califica cada proyecto de 1 estrella (malo) a 5 estrellas (bueno)” (1 star bad .. 5 stars good; 77 ratings of 0 kept as recorded; stars skipped on 1,604 profile rows). Respondents faster than 200 seconds (157), whom the authors drop, are kept with cov_duration_sec; duplicate panel IDs reduced to the first record; break-off tasks omitted. | 2026-10-08 | |
bruckmann_2020_ev_policy |
Conjoint table (irw_conjoint). Forced choice between two policy proposals (paraphrase: which proposal is preferred; no opt-out) and a 7-point rating of each proposal (anchors not documented; higher = more support by the data). Funding attribute shown to a random half only (‘(not shown)’ elsewhere), with ‘No additional funding’ shown iff no chargers and no subsidy. Task inferred from row order; 5 tasks with invalid choices set to NA; 352 respondents with no answers dropped (4,973 of 5,325). Non-EV n = 3,768 and no-funding non-EV n = 1,897 reproduce the article. | 2026-10-08 | |
brueckmann_2023_charging_stations |
Conjoint table (irw_conjoint_2). Forced choice between two charging stations (“Which charging station would you use?”, German; no opt-out), 5 tasks. Waiting time is a set of 2 or 3 possible values (shown as text, e.g. ‘0 or 5 or 10 min’); amenities text is taken from the article. Profile 1/2 is inferred from the file’s _c2/_c3 suffix. N = 2,831 respondents, as in the paper. | 2026-10-11 | |
brugge_2024_circular_products |
Conjoint table (irw_conjoint). Outcome: choice of which of two products to buy (wording not deposited; paraphrase), exactly one per round, no opt-out. Each respondent saw one randomly assigned product type (trial_product_type), and price and functional-duration levels are specific to that product. Level text is the authors’ English recode of German labels. The authors’ survey weight and attribute-order variables are not deposited; respondents with any missing conjoint answer were dropped before deposit. | 2026-10-08 | |
bueno_2017_credit_claiming |
Conjoint table (irw_conjoint). choice: ‘Qual prefeito você prefere?’ (forced, no opt-out); rating: performance of each mayor and rating_federal: performance of the federal government (asked once per task, repeated on both profiles), both 7-point Péssimo (1) to Ótimo (7). Attribute order randomized per respondent (attrpos_); 105 respondents have 3 tasks and 3 have 2; survey weight kept; Qualtrics ids re-keyed. | 2026-10-09 | |
bulman_2025_bailout_en |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_bailout_es |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_bailout_indonesia |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_bailout_malaysia |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_fdi_argentina |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_brazil |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_colombia |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_indonesia |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_kenya |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_malaysia |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language (Malay, English or Chinese by respondent; trial_language); attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_nigeria |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_peru |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_philippines |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_fdi_southafrica |
Conjoint table (irw_conjoint). choice: which of two investment projects the respondent preferred (article: respondents ‘indicated which project they preferred’; verbatim wording not deposited; forced, no opt-out). Levels as displayed in the survey language; attribute order randomized per respondent (attrpos_). Migrant workers always come from the firm’s country. Only respondents passing two attention checks were deposited; free text and the separate investment vignette dropped. | 2026-10-08 | |
bulman_2025_project_en |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_project_es |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_project_indonesia |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
bulman_2025_project_malaysia |
Conjoint table (irw_conjoint). choice = “Which foreign loan would you prefer for |
2026-10-09 | |
butler_2019_elite_trust_game |
Conjoint table (irw_conjoint_2). Four rounds per respondent of a game with a hypothetical partner described by ethnicity (Serb/Croat/Bosniak) and gender, both randomized; rating is the 0-10 game decision (the authors’ trust-game outcome; wording and unit not in the deposit, raw). Partner levels are numeric codes mapped from the authors’ own recode code; trial_eu_benefits_asked marks the respondent-level EU-benefits prime. One round dropped for a missing partner gender. | 2026-10-11 | |
butt_2022_actual_usage |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
butt_2022_cognitive_absorption |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
butt_2022_institutional_factors |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
butt_2022_performance_impact |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
butt_2022_task_tech_fit |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
butt_2022_user_satisfaction |
cov_age is an age-band code, not age in years: 1 = “20 or less”; 2 = “21-30”; 3 = “31-40”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
buttler_2026_hiring_vignettes |
Conjoint table (irw_conjoint). Factorial vignette survey, one vignette per task. rating_interview and rating_employ: ‘How likely is that you will invite this person for the interview / that this person would be employed …’, 0 very unlikely to 10 very likely (codebook’s 1-10 is wrong). rating_final_comparison keeps raw d2 codes: 0 not in the final pairwise comparison of the two highest-d11 candidates, 1 best (preferred), 2 second best. Fixed blocked design (144 vignettes in 24 decks); natives are always host-educated and C2-proficient by design. Level text is the English codebook labels; ‘other ethnic group’ resolved to the country’s nationality per the codebook. | 2026-10-09 | |
buzzi_2025_cyrm12 |
The CYRM-12 statements were answered retrospectively: ‘Ripensando ai periodi di maggiori restrizioni in pandemia’ (thinking back to the periods of strictest pandemic restrictions). Item codes are the LimeSurvey codes (G05Q36[SQ012] -> G05Q36_SQ012) in the questionnaire’s order. | 2026-10-09 | |
cai_2026_contract_repair_s1 |
Conjoint table (irw_conjoint_2). Between-subject 2x2x2 vignette (pre-violation trust in the manager, repair tactic, repair speed), one vignette per respondent (Study 1 of a three-study deposit; Study 3 is a longitudinal field study and is not built). Eight 1-5 ratings of psychological contract violation after the repair, four toward the manager and four toward the company, stored raw (anchors not in the deposit); no opt-out. Level text is the deposit’s English labels expanded with translated check wording, since the Chinese vignette is not deposited. | 2026-10-14 | |
cai_2026_contract_repair_s2 |
Conjoint table (irw_conjoint_2). Between-subject 2x2x2 vignette (pre-violation trust in the manager, repair tactic, repair speed), one vignette per respondent (Study 2 of a three-study deposit; Study 3 is a longitudinal field study and is not built). Eight 1-5 ratings of psychological contract violation after the repair, four toward the manager and four toward the company, stored raw (anchors not in the deposit); no opt-out. Level text is the deposit’s English labels expanded with translated check wording, since the Chinese vignette is not deposited. | 2026-10-14 | |
calvo_2026_milei_framing |
Conjoint table (irw_conjoint_2). Factorial vignette, one Facebook-style news post per respondent (Argentina, 20 Nov 2023): candidate (Milei or Massa) x claim (can or cannot control inflation) x label (It is TRUE or FALSE that). All 12 outcomes are 0/1 rating_* indicators (would like, share, comment or ignore the post; the emotion it evoked: one of Ekman’s categories or indifferent); no choice among profiles. Level text is the English translation of SIF Table 2 (the Spanish post is not deposited); 2,044 of 2,075 respondents were assigned a post and the authors’ regressions use 1,901 (complete cases on runoff vote). | 2026-10-14 | |
cammett_2024_clientelism_lebanon |
Conjoint table (irw_conjoint). Four paired candidate tasks with nine Arabic-text attributes; choice = vote for candidate A or B (paraphrase; forced, no opt-out). trial_prime records a priming experiment fielded before the conjoint. Not kept: an undocumented 1-7 per-candidate ‘rally’ item. IP, GPS and enumerator ids dropped; N = 2,455, no rows dropped. | 2026-10-09 | |
campbell_2019_dissent_frequency |
Conjoint table (irw_conjoint). choice = “Based on this information, which ONE of these two MPs would you prefer to have as your MP in the House of Commons?” (MP 1 / MP 2; a response was required, no opt-out). Attributes completely independently randomized with equal probabilities (article). Level text is the bold part of each bullet (e.g. dissent ‘never’..’often’ within ‘ |
2026-10-08 | |
campbell_2019_dissent_type |
Conjoint table (irw_conjoint). choice = “Based on this information, which ONE of these two MPs would you prefer to have as your MP?” (the Study 2 question without ‘in the House of Commons’; no opt-out). Motivation (‘own personal views’ / “constituents’ views”) and dissent type (three levels from ‘nevertheless tends not to speak out’ to ‘tends to speak out at internal party meetings and also publicly’) are the article’s quoted wording mapped from the authors’ labels; profile position is inferred from row order. N = 1,919 as in the article. | 2026-10-08 | |
campbell_2019_local_roots |
Conjoint table (irw_conjoint). choice: “Based on this information, which ONE of these two MPs would you prefer to have as your MP?” (MP 1 / MP 2, response required, no opt-out). Attributes assigned randomly and independently (full factorial of 768 profiles). Constituency, browser/device and derived recodes dropped; N = 1,719 matches the article and local-roots AMCEs (10-13 points vs living elsewhere) reproduce. | 2026-10-07 | |
caprara_2020_mnesres |
The paper analyses a 15-item MNESRES (three items per emotion); the deposit holds 40 administered items and its standardised ZAE columns identify the 15 as AE3, 5, 7, 10, 13, 15, 19, 21, 24, 26, 29, 31, 32, 35, 39. The emotion each belongs to is not recorded. Six deposit rows aged 19, which the authors left without an age group, are dropped to match the paper N of 1,695. Values of 99 and stray out-of-scale codes are set missing. | #2997 | 2026-10-14 |
caprara_2020_panas |
All 20 PANAS items ship; the paper scores only the ten negative-affect items and the deposit does not say which columns they are. Six deposit rows aged 19 are dropped to match the paper N of 1,695; 99 codes are set missing. The deposit also carries LS1-LS12 on 1-5, which cannot be the 5-item 1-7 SWLS the paper describes, so they are held. | #2997 | 2026-10-14 |
carey_2020_democracy_donors |
Conjoint table (irw_conjoint). choice: “Based on the information above, which of the following two candidates would you be more likely to support?” Candidate 1 / Candidate 2, no opt-out. Candidate race and gender are shown only through names (gender crosswalk from the authors’ candidate_names.xlsx). Platform-row order randomized per respondent (attrpos_). 3 respondents with unsaved levels dropped; the authors’ filters (US resident, not insincere) leave 573 of 622. | 2026-10-09 | |
carey_2020_democracy_public |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent supports (YouGov ‘Conjoint screen |
2026-10-09 | |
carey_2020_democracy_public_oct |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent supports (YouGov ‘Conjoint screen |
2026-10-09 | |
carey_2020_hiring_unm_faculty |
Conjoint table (irw_conjoint). choice: “If you had to choose between them, which of these two candidates should be given priority to be hired as a new faculty member at the University of New Mexico?”, forced choice between Candidate 1 and Candidate 2, no opt-out. Gender drawn 0.45/0.45/0.10 (Woman/Man/Non-binary), race White and Hispanic 0.35 each, and Non U.S. Citizen never paired with American Indian or Alaska Native; attribute order randomized per respondent (attrpos_). Respondents who did not finish the survey are not included, as in the paper; 870 respondents, one more than the authors’ estimates (869), whose reader drops the first row of the export. | 2026-10-08 | |
carey_2020_hiring_unm_students |
Conjoint table (irw_conjoint). choice: “If you had to choose between them, which of these two candidates should be given priority to be hired as a new faculty member at the University of New Mexico?”, forced choice between Candidate 1 and Candidate 2, no opt-out. Gender drawn 0.45/0.45/0.10 (Woman/Man/Non-binary), race White and Hispanic 0.35 each, and Non U.S. Citizen never paired with American Indian or Alaska Native; attribute order randomized per respondent (attrpos_). Respondents who did not finish the survey are not included, as in the paper; 1387 respondents, one more than the authors’ estimates (1386), whose reader drops the first row of the export. | 2026-10-08 | |
carey_2020_hiring_unr_faculty |
Conjoint table (irw_conjoint). choice: “If you had to choose between them, which of these two candidates should be given priority to be hired as a new faculty member at the University of Nevada, Reno?”, forced choice between Candidate 1 and Candidate 2, no opt-out. Gender drawn 0.45/0.45/0.10 (Woman/Man/Non-binary); attribute order randomized per respondent (attrpos_). Respondents who did not finish the survey are not included, as in the paper; 203 respondents, one more than the authors’ estimates (202), whose reader drops the first row of the export. | 2026-10-08 | |
carey_2020_hiring_unr_students |
Conjoint table (irw_conjoint). choice: “If you had to choose between them, which of these two candidates should be given priority to be hired as a new faculty member at the University of Nevada, Reno?”, forced choice between Candidate 1 and Candidate 2, no opt-out. Gender drawn 0.45/0.45/0.10 (Woman/Man/Non-binary); attribute order randomized per respondent (attrpos_). Respondents who did not finish the survey are not included, as in the paper; 622 respondents, one more than the authors’ estimates (621), whose reader drops the first row of the export. | 2026-10-08 | |
carlson_2023_free_expression |
Conjoint table (irw_conjoint). One scenario per task: rating is “How likely would you be to express your true opinions in that conversation?” on 5 points (1 = very unlikely, 5 = very likely; the source stores it rescaled 0-1). No choice question. Randomization restricted: the partisan news outlet (Fox News or MSNBC) follows the partner’s party and is not recorded for Independent partners, and the partisan news level appears in about two thirds of tasks. Unanswered tasks (207) omitted; the authors’ derived dummies and an unlabelled education code dropped. | 2026-10-07 | |
carnegie_2022_trade_partners |
Conjoint table (irw_conjoint). choice = ‘Which country should America increase trade with?’ (Country A/B, forced, no opt-out); rating = ‘How much would you support or oppose increasing trade with Country A [B]’ 1 = oppose .. 10 = support. Attribute row order randomized per respondent (attrpos_). Unequal level shares (Adversary 75%) and an observed restriction: conflict-reduction levels only for adversaries. 1 respondent without conjoint answers dropped; occupation free text and authors’ scales dropped. | 2026-10-09 | |
carreras_2024_corrupt_bureaucrats |
Conjoint table (irw_conjoint). Three forced choices per pair (choice = the authors’ main outcome outcome_q3; choice_q1, choice_q2 = alternative outcomes); the question wording is not in the deposit, no opt-out. Task and profile are inferred from row order (verified one chosen per pair). Attribute levels are the authors’ English translations; the authors’ matched subsample is not included. | 2026-10-08 | |
castanhosilva_2023_populism_us |
Conjoint table (irw_conjoint). choice: which of the two candidates the respondent would rather vote for (article paraphrase), forced, no opt-out; 132 unanswered tasks omitted (20 respondents drop out; 1,485 of the article’s 1,505). Priorities: exactly two of 12 per candidate, stored as one column per priority with “(not shown)” when absent; first/second order not in the deposit. Other restrictions and attribute order undocumented. | 2026-10-07 | |
caton_2025_sanction_vignettes |
Conjoint table (irw_conjoint_2). One news-story vignette per respondent (country pair x dispute type x costs of sanctions mentioned or (not shown); 2 x 2 x 2); choice = ‘Do you think the decision to impose sanctions was the right one?’ (single-profile yes/no, opt-out yes) and choice_lift_sanctions (should the President lift them if costly), plus 0-10 ratings of handling, approval, competence, concern about costs to the US and to oneself, and importance to national interest. Levels are short descriptions of the story elements; IP address and coordinates in the deposit were dropped; N = 858. | 2026-10-11 | |
cebul_2021_assurances |
Conjoint table (irw_conjoint_2). Single-vignette factorial survey (US-China crisis): three 0-4/0-3 ratings of the same vignette (credibility of China’s threat, higher = more credible; credibility of China’s assurance, coded so higher = more credible; resolve, 0 accept China’s demand to 3 reject at any cost). The authors’ inattentive respondents are already excluded (n = 1,028); id is a new integer. | 2026-10-11 | |
celikbas_2022_bpi |
Both the pain group and the control group answered the BPI; cov_group (1 kontrol, 2 agri) separates them. | 2026-10-11 | |
celikbas_2022_pcs |
The deposit carries 14 columns under the Pain Catastrophizing Scale prefix; items 1-13 reproduce the deposited rumination/magnification/helplessness sums exactly and column 14 does not belong to any of them, so it is not shipped. | 2026-10-11 | |
charnysh_2026_refugee_assistance |
Conjoint table (irw_conjoint). Forced choice of the preferred refugee profile (don’t know/refusal NA, 5,411 of 15,000 tasks), willingness to host 1-4 (higher = more willing) each task, donate food/clothes and money 1-5 (higher = agree) and in-need/crime/burden 0/1 in the first task of each block; wording from variable labels (paraphrase). Task/profile inferred from row order; Syrian block (tasks 1-3) then Ukrainian block (tasks 4-6). Derived and standardized variables dropped; 24 respondents with no answers omitted. | 2026-10-09 | |
chatzoudes_2021_emotional_exhaustion |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chatzoudes_2021_ethical_leadership |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chatzoudes_2021_job_satisfaction |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chatzoudes_2021_service_delivery |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chatzoudes_2021_trust |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chatzoudes_2021_turnover_intention |
cov_age is an age-band code, not age in years: 1 = “18-25”; 2 = “25-35”; 3 = “35-45”; 4 = “45-55”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
chauchard_2019_wealth_accumulation |
Conjoint table (irw_conjoint). Single-profile vignettes (3 per respondent). choice: would consider voting for the politician (yes/no; paraphrase), accept/reject so opt-out; rating_repr / rating_personal / rating_corrupt / rating_violence: how good a job addressing constituents’ problems, how helpful personally, how likely corrupt, how likely violent, 1-5 (paraphrases; anchors not deposited). Level text is the authors’ value labels rather than displayed wording. Restrictions: legality only with a wealth increase; one vignette per respondent shows the respondent’s own subcaste. Attribute order randomized and recorded (attrpos_). Article N = 1,020 vs 1,021 in deposit. | 2026-10-09 | |
chen_2024_cesd |
The paper scores the CES-D 0-3; the deposit stores 1-4 (its own total equals the item sum minus 20 on every row). Shipped as stored. Positive items (YY4, 8, 12, 16) are not reversed. All 623 deposited rows are kept although the paper analyses 476. | 2026-10-07 | |
chen_2024_learning_effectiveness |
No codebook: block membership was recovered by regressing each of the deposit’s 15 construct means on the items (exact equal-weight fits). | 2026-10-09 | |
chen_2024_learning_engagement |
No codebook: block membership was recovered by regressing each of the deposit’s 15 construct means on the items (exact equal-weight fits). | 2026-10-09 | |
chen_2024_learning_motivation |
No codebook: block membership was recovered by regressing each of the deposit’s 15 construct means on the items (exact equal-weight fits). | 2026-10-09 | |
chen_2024_personality |
No codebook: trait membership was recovered from the deposit’s own trait means. Those means put C14 in both Openness and Agreeableness and C13 in no trait; both are shipped. | 2026-10-09 | |
chen_2025_bere_barley_scot |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 512 cards; both profiles 0). One respondent answered only 2 cards (kept, as in the deposit). Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_hanging_tomato_es |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 330 cards; both profiles 0). Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_ideal_tomato_bg |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 193 cards; both profiles 0). Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_legume_farmers_it |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 68 cards; both profiles 0). choice_forced = follow-up pick between the two plans after Neither. Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_parmigiano_it |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 202 cards; both profiles 0). Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_pasta_pea_einkorn_it |
Conjoint table (irw_conjoint). Discrete choice experiment, 6 cards of Option 1 / Option 2 / Neither from a fixed blocked design (36 cards in 6 sets); choice = selected option (wording not deposited, paraphrase), opt-out yes (Neither chosen on 345 cards; both profiles 0). Level text from the codebook’s English labels; postcodes dropped. | 2026-10-09 | |
chen_2025_phone_addiction |
Two cells of -2 (q7_6, q7_12) are dropped as entry errors. | 2026-10-09 | |
chen_2025_social_anxiety |
Only four of the deposit’s seven social-anxiety items ship: the three reverse-worded items (8.3, 8.5, 8.7) are stored on 1-3 only (with two -2s), a lossy recode rather than the response. The club-participation block is not shipped: its 0 (non-member) coding disagrees with the file’s own membership column in 39 rows. | 2026-10-09 | |
chen_2026_clubs_participation |
Same depositor as the published chen_2026_* tables (DVN/QS5D8C) but a different study and sample; the clubs_ infix keeps the names apart. Item codes q |
2026-10-09 | |
chen_2026_clubs_phone_addiction |
Same depositor as the published chen_2026_* tables (DVN/QS5D8C) but a different study and sample; the clubs_ infix keeps the names apart. Item codes q |
2026-10-09 | |
chen_2026_clubs_social_anxiety |
Same depositor as the published chen_2026_* tables (DVN/QS5D8C) but a different study and sample; the clubs_ infix keeps the names apart. Item codes q |
2026-10-09 | |
chen_2026_clubs_ucla_loneliness |
Same depositor as the published chen_2026_* tables (DVN/QS5D8C) but a different study and sample; the clubs_ infix keeps the names apart. Item codes q |
2026-10-09 | |
chewning_2024_campaign_volunteers |
Conjoint table (irw_conjoint). choice = “Which of the following two individuals would you prefer to have work on your campaign as a volunteer?”, forced choice between two volunteer profiles, 3 tasks; the strong-ideology level depends on the campaign’s party (Strong Conservative for Republican, Strong Progressive for Democratic campaigns). 124 unanswered tasks dropped; 194 of 233 responses answered at least one task; ids are survey responses (3 respondents answered twice, linked by cov_person). | 2026-10-09 | |
chile_2023_children-adolescents-survey_cp_c |
The time-use items cp7_1 to cp7_4 and cp9_1 to cp9_4 (hh:mm durations, in hours) are not included. They are measured quantities, not item responses. | #1700 | 2026-10-01 |
chile_2023_social-welfare-survey_u |
The CASEN time-use follow-ups (u1_a to u11_a, u15_a to u17_a: minutes per day spent on each activity, 0 to 1440) are not included. They are measured quantities, not item responses. The yes/no and rating items they follow (u1 to u22) are all here. | #1700 | 2026-10-01 |
chilton_2020_fdi_reciprocity_cn |
Conjoint table (irw_conjoint). One vignette per task; choice = 1 if the respondent answered Yes to “should the Chinese government prevent the proposed transaction?” (English version of the Chinese instrument; single-profile yes/no, so opt_out = yes). Four attributes; level text is the authors’ short English labels, not the Chinese display text. No restrictions documented. Task order inferred from row order; three ids with 10 rows dropped; 1,613 respondents vs 1,659 recruited. | 2026-10-08 | |
chilton_2020_fdi_reciprocity_us |
Conjoint table (irw_conjoint). One vignette per task; choice = 1 if the respondent answered Yes to “should the United States government prevent the proposed transaction?” (0 = No; single-profile yes/no, so opt_out = yes). Six attributes filled into a text template; no restrictions or probabilities documented (authors analyse it as uniform). Task order inferred from row order; one id with 10 rows dropped; 2,003 respondents vs 2,010 recruited. | 2026-10-08 | |
cho_2026_asian_coethnic |
Conjoint table (irw_conjoint). Paired candidates, 10 tasks + task 11 repeating task 1 (same positions; trial_repeat_of = 1); choice = ‘Which candidate are you most likely to vote for?’ (no opt-out); rating = ‘Looking at just Candidate k, how likely are you to vote for this candidate for Congress?’ 1 very unlikely - 5 very likely (authors’ coding). Level probabilities unequal by design. All 1403 respondents with saved profiles and any answer are kept; the authors’ filter (complete, reCAPTCHA >= 0.5, not a pure independent) leaves 1176; respondents with no saved attribute levels dropped. | 2026-10-09 | |
cho_2026_asian_panethnic |
Conjoint table (irw_conjoint). Paired candidates, 10 tasks + task 11 repeating task 1 (same positions; trial_repeat_of = 1); choice = ‘Which candidate are you most likely to vote for?’ (no opt-out); rating = ‘Looking at just Candidate k, how likely are you to vote for this candidate for Congress?’ 1 very unlikely - 5 very likely (authors’ coding). Level probabilities unequal by design. All 1403 respondents with saved profiles and any answer are kept; the authors’ filter (complete, reCAPTCHA >= 0.5, not a pure independent) leaves 1179; respondents with no saved attribute levels dropped. | 2026-10-09 | |
chou_2021_party_candidates |
Conjoint table (irw_conjoint). Four candidates per screen, one per party (Die Linke, SPD, CDU, AfD; attr_party): choice of one candidate (no opt-out) and a 1-7 favourability rating of each (7 = most favorable). Level sets differ by party (the AfD candidate never has several terms of experience or the two liberal refugee policies) and each screen has one or two likely winners, so effects should be estimated within party. Attribute text is the codebook’s English translation. | 2026-10-09 | |
christensen_2020_ep_candidates |
Conjoint table (irw_conjoint). choice = which of two candidates the respondent would be most likely to vote for in the EP elections (paraphrase), forced, no opt-out. Every attribute had a randomized ‘not shown’ level (stored as (not shown)); no restrictions. Levels are the article’s English text; covariates limited to EU interest and the authors’ 3-category ideology. | 2026-10-09 | |
christensen_2020_participatory |
Conjoint table (irw_conjoint). Five pairs of participatory processes; choice = the alternative the respondent prefers (forced choice, no opt-out). No randomization restrictions (article). Levels stored as the article’s English Table 1 wording; task/profile inferred from row order (one chosen per pair verified). N = 1,050 as in the article; the article’s AMCEs (advisory +9.3, moderated discussion +11.5, regional reform +14 points) reproduce. | 2026-10-09 | |
christensen_2021_endorsements |
Conjoint table (irw_conjoint). choice = which of two hypothetical candidates the respondent chose (wording not available; forced choice, exactly one per task). Seven attributes, each with a ‘NO INFO’ level stored as (not shown) (display undocumented). Levels from value labels (English). N 1,021 matches the abstract. | 2026-10-09 | |
christensen_2021_online_platforms |
Conjoint table (irw_conjoint). Six pairs of online participatory platforms; choice = the platform the respondent would like to see introduced (forced choice, no opt-out); rating_participate = follow-up, asked only for the chosen platform, whether the respondent would participate on it (1 yes / 0 no; NA on the unchosen profile). Levels are the preprint’s English Table 1 wording (Finnish display text not deposited); restrictions not documented. N = 1,048 as in the article; the article’s AMCEs (strong verification +0.16, anonymity +0.093) reproduce. | 2026-10-09 | |
christensen_2023_backsliding_issues |
Conjoint table (irw_conjoint). choice: which of two leader profiles the respondent chose (wording not deposited; forced, one per task); choice_vote: follow-up ‘would vote for selected profile’ (No/Yes) coded 1 on the chosen profile if Yes, so opt-out (9,276 of 14,436 tasks have none). Attribute text is the authors’ English labels of Finnish displays; profiles from 180 fixed design versions; typed age kept only when a whole number; municipality dropped. | 2026-10-09 | |
christian_2025_imagination |
Three three-to-four-item blocks shipped as one table; the deposit has no description, legend or covariates. | 2026-10-11 | |
chriswell_2023_organizing_formal |
Conjoint table (irw_conjoint_2). rating = ‘How difficult do you think it would be for the neighbors to organize among themselves, without the government, in the first/second colonia?’; rating_government = ‘… to organize to ask the government for help …’ (translated), both 1 = Extremely difficult … 5 = Extremely easy (coded from the answer text). One task; response and result texts depend on the problem; the monitoring response appears in two spellings (stored as displayed). No choice outcome. GPS coordinates and colonia names dropped. | 2026-10-09 | |
chriswell_2023_organizing_pairs |
Conjoint table (irw_conjoint_2). rating = ‘How difficult do you think it would be for the neighbors to organize in the first/second colonia?’ (translated), 1 = Extremely difficult … 5 = Extremely easy (coded from the answer text); rating_organize = 1 if the colonia was named in ‘Which colonia(s) do you think the neighbors will organize?’ (both/one/neither), else 0. Two vignette sets per respondent; the private-response and result texts depend on the problem (water vs shootings); trial_problem_asked = the randomized problem the questions refer to. No choice outcome. One respondent with unsaved attribute values dropped; GPS coordinates and colonia names dropped. | 2026-10-09 | |
chu_2020_intent_consequences_college |
Conjoint table (irw_conjoint_2). Factorial vignette about wartime conduct, one per respondent, 2x2: positive or negative consequence of the action, civilian or UNESCO heritage-site target. Ratings: intentionality (1-6) and immorality (1-6); no opt-out. Level text is the stored/do-file labels only (the appendix wording is not in the deposit); question wording, anchors and direction are not documented. | 2026-10-14 | |
chu_2020_intent_consequences_mturk |
Conjoint table (irw_conjoint_2). Factorial vignette about wartime conduct, one per respondent, 2x2x2: positive or negative consequence of the action, civilian or UNESCO heritage-site target, low or high severity. Ratings: intentionality (1-6), immorality (1-6) and responsibility (1-5); no opt-out. Level text is the stored/do-file labels only (the appendix wording is not in the deposit); question wording, anchors and direction are not documented. | 2026-10-14 | |
chu_2024_democracy_eg |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (ليست ديمقراطية على الإطلاق) to 10 (ديمقراطية تمامًا); wording in the survey language of Egypt. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,008; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2024_democracy_in |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (बिल्कुल लोकतांत्रिक नहीं) to 10 (पूर्णतः लोकतांत्रिक); wording in the survey language of India. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,022; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2024_democracy_it |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (Per niente democratico) to 10 (Del tutto democratico); wording in the survey language of Italy. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,047; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2024_democracy_jp |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (まったく民主的でない) to 10 (完全に民主的); wording in the survey language of Japan. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,012; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2024_democracy_th |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (ไม่เป็นประชาธิปไตยเลย) to 10 (เป็นประชาธิปไตยโดยสมบูรณ์); wording in the survey language of Thailand. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,036; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2024_democracy_us |
Conjoint table (irw_conjoint). choice = which of the two countries is more democratic (Country A / B, forced); rating = how democratic each country is, 1 (Not at all Democratic) to 10 (Completely Democratic); wording in the survey language of United States. Attribute order randomized per respondent (attrpos_*); 3 tasks. N = 1,024; the six samples total 6,149 here vs 6,150 in the article (one Thai respondent without saved attributes dropped). | 2026-10-09 | |
chu_2025_democracy_tradeoff_eg |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,008 as in the paper; attribute text as displayed in the survey language. | 2026-10-08 | |
chu_2025_democracy_tradeoff_in |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,022 as in the paper; attribute text as displayed in the survey language. Respondents chose Hindi or English (cov_survey_language) but the export stores the attribute text in Hindi only; one respondent-wealth level has two Unicode spellings (as stored; the authors merge them). | 2026-10-08 | |
chu_2025_democracy_tradeoff_it |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,047 as in the paper; attribute text as displayed in the survey language. | 2026-10-08 | |
chu_2025_democracy_tradeoff_jp |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,012 as in the paper; attribute text as displayed in the survey language. In Japan an administrative error showed respondents drawn for ‘somewhat dangerous’ or ‘very dangerous’ both descriptors in one level (stored as displayed; public safety has 3 levels). | 2026-10-08 | |
chu_2025_democracy_tradeoff_th |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,037 as in the paper; attribute text as displayed in the survey language. | 2026-10-08 | |
chu_2025_democracy_tradeoff_us |
Conjoint table (irw_conjoint). choice = which of two hypothetical countries (A/B) the respondent would prefer to be born and grow up in (forced, no opt-out); rating = how happy the respondent would be to have been born and grown up in each country, 1 = extremely unhappy to 7 = extremely happy. Levels independently randomized with equal probability; attribute order randomized per respondent (only the position of the democracy block is recorded). N = 1,024 as in the paper; attribute text as displayed in the survey language. | 2026-10-08 | |
cifar10h |
resp scores the annotator’s choice against the CIFAR-10 ground-truth label (itemcov_true_category), which is itself a label assigned when the dataset was built and is wrong for some images; Peterson et al. (2019) use the human label distribution, not the key, as the target. The chosen class is in resp_raw (core) and text (cifar10h_nom). | #2655 | 2026-10-01 |
claassen_2026_democratic_backsliding |
Conjoint table (irw_conjoint_2). Twelve forced-choice pairs of hypothetical candidates (seven attributes) in the US, South Korea and Poland (cov_country); choice = ‘Which candidate do you prefer?’, no opt-out. The political-system attribute is constrained (one anti-democratic and one pro-democratic position per pair), gender is 80% male and Polish parties are drawn with unequal probabilities. All respondents kept; the authors’ speed exclusion (109) can be applied with cov_duration_sec. | 2026-10-11 | |
clarke_2021_birth_weight |
Conjoint table (irw_conjoint). Forced choice ‘Which of these two birth scenarios would you choose?’ between two hospital births of a first child (sex, out-of-pocket expenses $250-$10,000, birth weight 5 lb 8 oz-8 lb 13 oz, season), 7 tasks, no opt-out; orthogonal balanced main-effects design, attribute order randomized once per respondent. All 2,005 responses kept; the article’s analysis sample of 1,894 is cov_main_sample = 1 (the rest were outside the US by IP, failed a repeated question or took under 2 minutes). | 2026-10-08 | |
clayton_2021_france_immigrants |
choice = “S’il vous fallait choisir entre les deux, lequel ou laquelle de ces deux immigrant(e)s devrait être admis(e)…” (forced, no opt-out, 10 tasks). Origin weighted toward Western Europe, with education/profession and origin/reason/visit restrictions; the education restrictions failed in the 28 Dec soft launch (cov_fielding = 1), whose tasks are kept here (authors drop 332). Postal code, département, IP, location and free text dropped; N = 1,500 matches the article. | #2887 | 2026-10-07 |
clayton_2025_gender_quotas |
Conjoint table (irw_conjoint_2). Six ratings of the same vignette on 1-4 scales taken from the authors’ recode scripts (agree: 1 strongly disagree to 4 strongly agree; fair: 1 very unfair to 4 very fair): right decision for citizens, right decision for women/animals, fairness to women/animals, fairness of the process, overturn (stored raw, higher = less legitimacy) and trust; no choice or opt-out. The two attributes are council composition (3) and policy issue (2); Australia, France and Norway saw harassment only. Authors’ speeder and manipulation-check exclusions are not applied (cov_manipulation_pass, cov_duration_sec); The paper reports over 17,000 surveyed and an analytic sample of 13,274 after exclusions; the file has 17,037 raw rows and the table 16,427 with a vignette and an answer. | 2026-10-11 | |
clemente_2024_attitude_father |
The deposit’s variable label for FATHER9 (‘Does not see child’) does not match item 9 of the published questionnaire (‘I believe that he/she does not love his/her children’); the other nine labels match by position. | 2026-10-04 | |
clemente_2024_attitude_mother |
The deposit’s variable label for MOTHER9 (‘Does not see child’) does not match item 9 of the published questionnaire (‘I believe that he/she does not love his/her children’); the other nine labels match by position. | 2026-10-04 | |
clemente_2024_pmd |
The PMD is in the deposit and the SI questionnaire but is not described in the paper; the deposit’s own Moral_Disengagement composite averages only the first seven items, leaving out Attribution_of_blame. | 2026-10-04 | |
coffe_2024_quota_fr |
Conjoint table (irw_conjoint_2). French sample: one pair of hypothetical parliamentary candidates per respondent (7 attributes: gender, marital status, policy priority, experience, age, career, skill) after a randomized framing of positive-action measures (Control/Parity/Quota, trial_framing). choice = which candidate, no opt-out; rating_qualified = 1-11 qualified-to-be-MP rating. Wording not in the deposit; N=1038. | 2026-10-11 | |
coffe_2024_quota_uk |
Conjoint table (irw_conjoint_2). UK sample: one pair of hypothetical parliamentary candidates per respondent (7 attributes: gender, marital status, policy priority, experience, age, career, skill) after a randomized framing of positive-action measures (Control/Parity/Quota, trial_framing). choice = which candidate, no opt-out; rating_qualified = 1-11 qualified-to-be-MP rating. Wording not in the deposit; N=1734. | 2026-10-11 | |
cohen_2024_poll_watchers_s1 |
Conjoint table (irw_conjoint). Pairs of hypothetical polling locations; choice = which location would conduct the election more fairly (paraphrase of the pre-registration; forced choice, no opt-out, unanswered tasks NA or omitted); rating = five-point intensity of that belief for each location, 1-5 raw, higher = fairer inferred, anchors unknown. Attributes randomized independently per the pre-registration. Respondents with empty attribute fields dropped (as the authors do); exact question wording not in the deposit. | 2026-10-09 | |
cohen_2024_poll_watchers_s2 |
Conjoint table (irw_conjoint). Pairs of hypothetical polling locations; choice = which location would conduct the election more fairly (paraphrase of the pre-registration; forced choice, no opt-out, unanswered tasks NA or omitted); rating = five-point intensity of that belief for each location, 1-5 raw, higher = fairer inferred, anchors unknown. Attributes randomized independently per the pre-registration. Respondents with empty attribute fields dropped (as the authors do); exact question wording not in the deposit. | 2026-10-09 | |
cohen_2024_poll_watchers_s3 |
Conjoint table (irw_conjoint). Pairs of hypothetical polling locations; choice = which location would conduct the election more fairly (paraphrase of the pre-registration; forced choice, no opt-out, unanswered tasks NA or omitted). Attributes randomized independently per the pre-registration. Task and profile inferred from the stacked row order (three blocks of pairs). Respondents with empty attribute fields dropped (as the authors do); exact question wording not in the deposit. | 2026-10-09 | |
conroykrutz_2025_media_restrictions |
Conjoint table (irw_conjoint). Single-scenario factorial: rating = what should be done to the station, 1 Nothing, 2 written warning, 3 fine, 4 temporary shutdown, 5 permanent shutdown (higher = harsher; authors use 0-4); no opt-out; randomization restrictions not documented. Refused/don’t know (504 rows) dropped; 4,975 respondents fielded, 4,920 with an answer; vignette wording and interview language not in the deposit. | 2026-10-07 | |
contreras_valdez_2022_edeq |
Items edeq_i13 to edeq_i18 are the EDE-Q’s behavioural frequency questions (number of episodes or days over the past 28 days: overeating, loss of control, binge days, self-induced vomiting, laxative use, compulsive exercise), so resp is a self-reported count (here 0 to 50), not a 0-6 rating. The EDE-Q’s subscale and global scores use only the 22 rating items. | #1700 | 2026-10-01 |
costa_2023_research_ethics |
Conjoint table (irw_conjoint). choice: ‘If you had to choose, which of the two studies would you say is more ethical?’ 1/2, forced (no opt-out). Attribute order randomized once per respondent (attrpos_ columns; attributes move in fixed blocks); unanswered tasks dropped, leaving 247 of 362 consenting respondents; ranking and harm batteries not kept. | 2026-10-09 | |
cozza_2025_party_leaders |
Conjoint table (irw_conjoint_2). Five forced choices per pair (‘Which of these leaders earned their position?’, ‘… would be more effective in passing legislation?’, ‘… will work harder on behalf of their party?’, ‘… would be more effective in unifying the party?’, ‘… would help their party win more seats in the next election?’), plus choice_vote (vote; wording not deposited) and rating_lr (0-10 left-right placement of each party, wording not deposited); answers not forced, so some NA. Restriction: 23 years as MP never with age 38 or 46. Task 5 repeats task 1 with profiles swapped (trial_repeat_of). Per-profile trust item dropped (text only). 1,195 respondents with any answer of 1,198 in the article. | 2026-10-09 | |
crabtree_2026_civilian_violence |
Conjoint table (irw_conjoint). choice_anger = which description made the respondent angrier, choice_punish = which perpetrator should be punished more (both forced, no opt-out); rating_punish = support for life in prison for each description, 1-5 raw, higher = more punitive (wording paraphrased, anchors unknown). Attribute levels are the author’s stored labels; no restrictions documented. Black and White respondents only; the author’s entropy-balancing weights (white subsample) are not kept. | 2026-10-09 | |
crawfurd_2025_aid_projects |
Conjoint table (irw_conjoint_2). choice: ‘(k/6) Now I am going to ask you about some hypothetical aid projects. I am going to describe two aid projects and ask you to choose which one you would prefer.’ Project 1 / Project 2 / None of these options; OPT-OUT ‘None’ = choice 0 on both (252 of 4,470 tasks; no in-person respondent chose it). 6 rounds, 3 attributes (project, technical assistance, budget), English text (some respondents answered in French). Tasks with unsaved attribute values (175 respondents) dropped; country and job title dropped for confidentiality. The article reports 931 officials in all. | 2026-10-09 | |
crawfurd_2025_education_outcomes |
Conjoint table (irw_conjoint_2). choice: ‘There are multiple possible goals of an education system … Imagine you had to choose between the outcomes of two education systems, which do you prefer?’ System 1 / System 2 / None of these options; OPT-OUT ‘None’ = choice 0 on both (163 of 3,483 tasks). 4 rounds; attributes Gain foundational literacy (%) 60/80/100, Complete secondary school (%) 30/50/70, Dutiful citizens (%) 70/80/90/100, stored as numbers. Tasks with unsaved attribute values (45 respondents) dropped; country and job title dropped for confidentiality. | 2026-10-09 | |
crowdermeyer_2023_rcv_candidates |
Conjoint table (irw_conjoint). Mock local elections with 4 named candidates per office (party shown in the partisan arm, else (not shown)); trial_rcv: plurality or ranked-choice ballot. choice = top choice in mayor and district races (plurality vote or first rank), NA for the two-winner at-large race; rating_vote = plurality mark (1 = voted for), rating_rank = RCV rank (1 = first preference, lower = more preferred). No abstention. Wave-2 recontacts (trial_wave = 2) have new, unlinkable ids. Wording not deposited. | 2026-10-09 | |
cuong_2022_rice_labels |
Conjoint table (irw_conjoint). choice: “I prefer:” Alternative A, Alternative B or the status quo, one of three; the status quo rice (no certification, VND 20,000/kg) is profile 3 in every task. Fixed blocked orthogonal design (4 blocks of 4 cards, trial_block). Level text from the article’s example card in English; certified options also carried a label logo. Task order within block inferred. 410 respondents, as in the article. | 2026-10-08 | |
cuttler_2017_cupit |
Each CUPIT item keeps its own published response scale (ranges from 0-3 to 1-9), so ranges differ by design. CUPIT7 runs 1 = ‘6 months or longer’ .. 9 = ‘no days at all’ and CUPIT9 0 = never .. 4 = always able to stop; both as stored. | 2026-10-09 | |
cwalina_2023_affect |
Affect ratings after the framing manipulation (treat). | 2026-10-10 | |
cwalina_2023_covid_attitudes |
Semantic-differential items rated after a gain/loss framing manipulation (treat); 8 cases the authors excluded from analysis are kept and flagged by cov_excluded_by_authors. | 2026-10-10 | |
dafoe_2021_coercion_provocation |
Conjoint table (irw_conjoint_2). Between-subject 2x5 factorial scenario (US military capability inferior/superior x air incident: attack, collision, non-collision, accident, or none shown), one per respondent. Four ‘how far should the US go to assert its claims’ ratings, higher = more resolve: military force (0-3, in rating), economic cost (0-9), military deaths (0-7), risk of war (0-6); no opt-out. The control arm shows no incident (‘(not shown)’). N matches the paper (1,761). | 2026-10-14 | |
dahlweg_2026_citizen_collaboration |
Conjoint table (irw_conjoint). rating = willingness to collaborate with the candidate, 1 (very unlikely) to 7 (very likely), single-profile, no choice; wording paraphrased from the codebook. Levels are the codebook’s English labels, not the displayed German/Chinese text. 4 profiles per respondent, order not recorded (task = row order). N = 744 (379 DE + 365 TW), as the codebook states. | 2026-10-09 | |
dai_2024_populist_rhetoric |
Conjoint table (irw_conjoint). choice = ‘If you had to choose between these two candidates in the upcoming primary, who would you vote for? If neither of the two candidates appeals to you, please still indicate who you would rather vote for.’ (forced choice, no opt-out). Candidates shown as prose; attributes are job, office, polls and four campaign statements drawn from lists of 12, 12, 8 and 8 (online appendix Table B1). Randomization restrictions undocumented (shares look uniform). The authors’ derived populism codings and populist-attitude flags are dropped; cov_attention_pass flags the respondents who passed the attention check. | 2026-10-07 | |
dai_2025_fear_of_domination |
Conjoint table (irw_conjoint_2). Paired forced choice between two international organizations (4 attributes); choice = the chosen IO, no opt-out. Question wording, codebook and article are not in the deposit, and attribute levels are the authors’ short labels, not the full text shown; US involvement x decision rule combinations are unbalanced. Position is recorded (IO A/B, C/D, …). Derived dummies, dates and attitude items dropped; 1,633 respondents. | 2026-10-12 | |
darr_2026_engagement_news |
Conjoint table (irw_conjoint). Forced choice between two news stories per task (the authors’ chosen), no opt-out; question wording not in the deposit. Headline is the authors’ short label, not the displayed wording; photo levels describe photographs. Qualtrics Conjoint with 260 design versions (trial_design_version); restrictions and level probabilities undocumented. Respondents can recur across the four darr_2026 tables but IDs are not linkable. |
2026-10-08 | |
darr_2026_good_conflict_news |
Conjoint table (irw_conjoint). Forced choice between two news stories per task (the authors’ chosen), no opt-out; question wording not in the deposit. Headline and note are the authors’ value labels (headline stored as codes GC1/GC2/GC bad 1/GC bad 2), not the displayed wording; photo levels describe photographs. Qualtrics Conjoint with 260 design versions (trial_design_version); restrictions and level probabilities undocumented. Respondents can recur across the four darr_2026 tables but IDs are not linkable. |
2026-10-08 | |
darr_2026_solutions_news |
Conjoint table (irw_conjoint). Forced choice between two news stories per task (the authors’ chosen), no opt-out; question wording not in the deposit. The respondent’s state piped into the headline is stored as [your state]; photo levels describe photographs. Qualtrics Conjoint with 260 design versions (trial_design_version); restrictions and level probabilities undocumented. Respondents can recur across the four darr_2026 tables but IDs are not linkable. |
2026-10-08 | |
darr_2026_transparency_news |
Conjoint table (irw_conjoint). Forced choice between two news stories per task (the authors’ chosen), no opt-out; question wording not in the deposit. Headline, note and pledge are the authors’ short labels, not the displayed wording; 39 tasks of 7 respondents who left early are dropped; photo levels describe photographs. Qualtrics Conjoint with 260 design versions (trial_design_version); restrictions and level probabilities undocumented. Respondents can recur across the four darr_2026 tables but IDs are not linkable. |
2026-10-08 | |
dasandi_2022_climate_frames |
Conjoint table (irw_conjoint). Forced choice between two climate statements: “Indicate which of the two statements would make you more likely to support policies to tackle climate change.” (exactly one chosen per task, no opt-out); four frames randomized without constraints (96 statements). Attribute levels are the authors’ frame names, not the displayed sentences; the willingness-to-pay rating is dropped as undocumented. 7,512 respondents = the article. Deltapoll IDs re-keyed; survey weight kept. | 2026-10-08 | |
dasandi_2022_climate_frames_pilot |
Conjoint table (irw_conjoint). Forced choice between two climate statements: “Indicate which of the two statements would make you more likely to support policies to tackle climate change.” (exactly one chosen per task, no opt-out); four frames randomized without constraints (96 statements). Attribute levels are the authors’ frame names, not the displayed sentences; the willingness-to-pay rating is dropped as undocumented. Pilot: the article mentions earlier pilots in the five countries (Supplementary Figs. S24-S25) but gives no n or date; 4,935 respondents in the file. Deltapoll IDs re-keyed; survey weight kept. | 2026-10-08 | |
dassonneville_2021_mp_vote |
Conjoint table (irw_conjoint_2). One vignette per respondent; rating = how the MP should vote on reducing the number of immigrants by 10%, stored raw (1 = For reducing, 2 = Against reducing). Full 2 x 2 x 2 factorial with independent bullets; bullet order randomized per respondent (trial_bullet_order, letters not mapped to bullets); 0-10 follow-up with truncated wording dropped. 2,001 respondents, as in the article. | 2026-10-09 | |
davenport_2026_black_classification |
Conjoint table (irw_conjoint). choice = which person the respondent selected as more Black (Person A/B, forced; wording not deposited); rating = how Black each person is, 0 = Not at all to 10 = Very. Four forms randomized between respondents: parents’ race shown or not (attr_parents = ‘(not shown)’ when not shown; trial_ancestry_shown) and skin tone near the top or at the bottom (trial_skin_tone_row). Authors’ filters applied (2,397 respondents; README’s 2,385 excludes 12 outside the three race groups); 2 tasks without saved levels dropped. | 2026-10-07 | |
davis_2026_ethnic_federalism |
Conjoint table (irw_conjoint). Forced choice between two federal arrangements (question wording not deposited); no opt-out. Task inferred from row order (A,B,A,B per respondent; every pair has one chosen profile). 102 of 2,108 pairs dropped for missing attributes or outcome; Endline and Expansion samples pooled as in the authors’ code (cov_sample). | 2026-10-09 | |
debele_2026_opendata_portal |
Conjoint table (irw_conjoint). Discrete choice experiment, fixed design (same 5 sets of 3 portals for everyone) plus a status quo alternative without attributes. rating: rank of the alternative among the four, 1 = most preferred to 4 = least preferred (record description), stored as is; the status quo’s rank is the one missing from a task. No choice column. | 2026-10-09 | |
debenedictiskessner_2022_opioid_identity |
Conjoint table (irw_conjoint). Factorial vignette: one news story per respondent with five randomized attributes; rated federal spending on treatment and on law enforcement for those addicted to opioids (1 Increase a lot .. 5 Decrease a lot) and agreement that addicts are to blame for their addiction (1 Strongly agree .. 5 Strongly disagree), stored in source code order; order of the two spending items randomized (trial_question_order). Race and gender were conveyed by a name, a photo of a hand and pronouns: the attr_ columns hold the authors’ condition labels, the name itself is not deposited. 5 respondents who skipped all three outcomes dropped (3,107 of 3,112). | 2026-10-08 | |
debruin_2026_crisis_accountability |
Conjoint table (irw_conjoint). Single text vignette per task (5 per respondent): approval of the president’s action, 1 Strongly disapprove to 6 Strongly approve (higher = more approval; wording paraphrased, no questionnaire deposited). Attributes are the displayed text pieces (situation, party, pronoun, age, electoral margin, action); all level pairs occur. Task-5-only follow-up questions and manipulation checks dropped; one duplicate panelist removed (4,191 respondents). | 2026-10-08 | |
deception_professors |
The reason_less_rigorous_* (9) and reason_no_deception_* (8) items are not answered by respondents: they are themes the study’s authors coded from two open-ended answers (“Do you think that studies that use deception are inherently less rigorous…? Please elaborate” and “Why don’t you use deception in your experimental designs?”), as deposited. 1 = the answer was coded with that theme, 0 = it was coded with others; one answer can carry several themes. Respondents who left the question blank have no rows. The deposit does not document the coding procedure. *_considerdecept (1-5) and you_limit_decept (1-3) are unordered answer choices, not rating scales. | #2846 | 2026-10-05 |
deckman_2022_genz_candidates |
Conjoint table (irw_conjoint_2). One candidate per respondent (age 22 or 68 x woman or man x White or Black, picture-based; 2 x 2 x 2); eight ratings of how likely the respondent is to take part in political activities, stored only as the authors’ 0-1 normalisations of 5-point items (direction inferred as higher = more likely). No opt-out; levels are the authors’ condition codes, no stimulus or item wording in the deposit; 2,205 respondents. | 2026-10-11 | |
degirolamo_2022_brs |
Three of the Brief Resilience Scale’s six items (BRS 2, 3 and 4), asked on a 7-point agreement scale rather than the BRS’s 5-point one. pag15_1_1 and pag15_1_3 are negatively worded; responses are unreversed. | #2602 | 2026-09-30 |
degirolamo_2022_brs |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2602 | 2026-09-30 |
degirolamo_2022_cmq |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_covid_perception |
Each item is a pair of opposite phrases with its own poles (for example 1 = close to me, 7 = far from me), and the direction is not aligned across items: pag8_1_2 runs slowly to fast spreading, while the others run from the more threatening pole to the less. | #2605 | 2026-09-30 |
degirolamo_2022_covid_perception |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_health_literacy |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_info_use |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_k6 |
Stored 1-5 as deposited, with 1 = “always” and 5 = “never”. Standard K6 scoring is 0-4 with 4 = all of the time, so here a higher resp means lower distress; the standard item score is 5 - resp. | #2605 | 2026-09-30 |
degirolamo_2022_protective_behaviour |
For pag7_2_9 (stayed home from work or school), “not applicable” (code 8) answers are dropped, so that item has fewer respondents than the others. | #2605 | 2026-09-30 |
degirolamo_2022_protective_behaviour |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_risk_perception |
pag5_2_1 (vulnerability) runs from 1 = completely vulnerable to 7 = completely invulnerable, the opposite direction from pag5_1_1 (likelihood) and pag5_3_1 (severity), where 7 is the higher risk. | #2605 | 2026-09-30 |
degirolamo_2022_risk_perception |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_trust_institutions |
For pag12_2_8 (employer) and pag12_2_9 (parish), “not applicable” (code 8) answers are dropped, so those items have fewer respondents than the others. | #2605 | 2026-09-30 |
degirolamo_2022_trust_institutions |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_trust_sources |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2605 | 2026-09-30 |
degirolamo_2022_who5 |
Stored 1-6 as deposited, with 1 = “all of the time” and 6 = “at no time”. Standard WHO-5 scoring is 0-5 with 5 = all of the time, so here a higher resp means lower well-being; the standard item score is 6 - resp. | #2602 | 2026-09-30 |
degirolamo_2022_who5 |
The four survey waves are separate samples of different people, not a panel. responseid restarts in each wave, so id is prefixed with the wave (w1_ to w4_) and the wave is recorded as cov_survey_wave. | #2602 | 2026-09-30 |
deilkas_2019_collaboration |
Responses of 0 ‘Not applicable’ were dropped; the hospital, out-of-hours and prioritization-office ratings have the fewest responses because many staff do not collaborate with them. | 2026-10-10 | |
deilkas_2019_saq |
Responses of 0 ‘Not applicable’ were dropped; SAQ59 (adequate and timely training) has the most. The deposit’s gender and age variables carry each other’s labels; the script follows the variable names and value sets. | 2026-10-10 | |
dekleer_2025_protest_support |
Conjoint table (irw_conjoint_2). Two text vignettes per respondent, one profile each. rating = ‘Steunt u de protestacties van de [protestgroep] of keurt u deze af?’ 1 = Keur ze sterk af … 5 = Steun ze in sterke mate; rating_prosecute = ‘Vindt u dat de [protestgroep] vervolgd moeten worden?’ 1 = Ja, 0 = Nee (4 whitespace answers as missing). Levels are the Dutch vignette fragments; the full template is not deposited. 2 tasks with neither answer omitted. N = 2,528, as in the README. | 2026-10-09 | |
delponte_2020_debt_dilemmas |
Conjoint table (irw_conjoint_2). One vignette per respondent (4 x 2 between-subject: jobs lost by repaying, full vs partial default). choice = whatDo, a single-profile decision about the government (opt-out = default); its coding 1 = repay is inferred from the authors’ Study 2 comment and the data pattern, not stated for Study 1. Ratings: moral wrongness of stopping repayment / of making cuts (0/1 and 0-10) and punishment strength (0-10). Levels are the authors’ condition labels; Study 2 (single factor) not built; id = row number; free text and MTurk ids dropped. | 2026-10-11 | |
denny_2022_police_cooperation |
Conjoint table (irw_conjoint). Two paired scenario tasks; choice = the scenario in which the respondent would cooperate with the police (paraphrase; wording not deposited), forced choice, no opt-out. Levels are the authors’ English labels (displayed wording not deposited); profile order within a task is the deposit’s row order, screen position unknown. N = 1,841 (not checked against the closed article). | 2026-10-09 | |
depression_anxiety_stress |
This table bundles three instruments from the same online survey: DASS-42 (items 1 to 42), TIPI (43 to 52, scored 1 to 7) and a vocabulary check list (53 to 68, scored 0/1). Model them separately. | #2513 | 2026-09-28 |
deryahanoglu_2024_exercise_motivation |
Instrument inferred from the deposit’s subscale variables (intrinsic, introjected, external, amotivation); the record has no description or codebook. | 2026-10-10 | |
deschepper_2026_firefly_lights |
Conjoint table (irw_conjoint_2). Paired choice between two firefly-light products described by five attributes (Dutch text), with a ‘neither’ option (opt-out; both profiles 0 in 475 of 1,840 tasks); the question wording is not in the deposit. Fixed design: all 184 respondents saw the same 10 pairs in the same order. The installation level is missing from most chosen-answer texts in the source and was filled from the fixed design, so that respondents saw it is inferred. | 2026-10-12 | |
development_delay_anunciacao_asqse |
Item codes are q |
#2834 | 2026-10-05 |
development_delay_anunciacao_concerns |
Item codes are q |
#2834 | 2026-10-05 |
devine_2024_climate_policy |
Conjoint table (irw_conjoint). choice: “Which proposal would you prefer to be implemented?” (Proposal A / Proposal B, no don’t-know option). Attribute order randomized per respondent and fixed across tasks (attrpos_; policy always first); restrictions undocumented. Level text is English (instrument labels; pricing, GDP and recommended-by abbreviated as in the article’s Table 1) although respondents saw German. N = 1,558 matches the article. | 2026-10-07 | |
devine_2025_class_political_success |
Conjoint table (irw_conjoint_2). Two people rated per task, no choice and no opt-out; rating = chance of being a successful MP candidate, 0-10 (group 2 only; ‘not sure’ = NA, 2,828 of 12,500 profile ratings), rating_class = rated on a scale from working class (1) to upper class (5), both groups. Group 1 saw a ninth attribute, political office; for group 2 it is ‘(not shown)’. Attribute order randomized per respondent (string kept as trial_attr_order). Survey weight kept. N = 4,157 respondents (paper’s analysis sample = group 2, 2,084). | 2026-10-11 | |
dey_2017_mistreatment_observed |
DVN/GLXDPR is an identical copy of this deposit and is not separately processed. cluster_id is the delivery facility. | 2026-10-09 | |
dey_2017_mistreatment_selfreport |
DVN/GLXDPR is an identical copy of this deposit and is not separately processed. cluster_id is the delivery facility. | 2026-10-09 | |
diaz_2025_mayor_credit |
Conjoint table (irw_conjoint_2). One factorial vignette per respondent (Argentina, Netquest panel, 2021): mayor gender x information on programme distribution (strictly need-based / need-based in theory but contacts favoured in practice / omitted) x mayor party (Peronist / PRO / omitted) x mayor’s name on the food box (Pedro or Alicia / omitted); omitted levels are ‘(not shown)’. Seven ratings 1-4 (1 = not at all likely, 4 = very likely). Level text paraphrased from the article (Spanish shown); 21 all-missing respondents dropped. | 2026-10-14 | |
diehl_2025_campus_speakers |
Conjoint table (irw_conjoint). Factorial vignette (text), 5 topic vignettes per respondent, each randomizing speaker (professor/journalist), statement (conservative/progressive x with/without political demand) and a critical reaction (‘(not shown)’ when absent); four yes/no judgements stored as ratings (Ja = 1, Nein = 0): cancel the talk, revoke his teaching assignment, remove his book from the library, allow disruptive protests. Task numbers are topics, not display order (first four topics in random unrecorded order); 19 soft-launch respondents’ item order corrected as in the authors’ code. Studies 2-4 of the deposit are not built. | 2026-10-09 | |
diessner_2026_welfare_deservingness |
Conjoint table (irw_conjoint). choice = which of two workers the respondent would prefer to receive greater government income assistance (forced choice, no opt-out); rating = 0-10 support for each worker receiving income assistance (0 strongly oppose, 10 strongly support). Attribute order randomized per respondent (attrpos_). Task 6 repeats task 1 with workers swapped, rebuilt from task 1 (trial_repeat_of = 1). trial_ict_info marks the between-subjects ICT information arm. All respondents with an answered task kept (4,013); the authors’ 3,916 = cov_finished 1 and cov_duration_sec > 180. | 2026-10-07 | |
dietrich_2023_welfare_priority |
Conjoint table (irw_conjoint). choice = “Wessen Antrag lassen Sie zuerst bewilligen?” (whose application do you approve first; Person A or B, forced, no opt-out). German level text; attrpos_ give the per-respondent attribute row order. Four samples pooled (cov_sample; cov_sector for the general-population sample); preview, incomplete and non-consenting responses excluded as by the authors; one task with an invalid answer dropped. N = 3,882, not checked against the article. | 2026-10-09 | |
digiuseppe_2018_trade_agreements |
Conjoint table (irw_conjoint). choice is a forced choice of the preferred trade agreement (wording paraphrased in the article; instrument appendix not available), no opt-out. Three randomized arms (trial_arm): the global-influence and relationship attributes appear only in their own arm and are (not shown) elsewhere. One respondent with no regime values dropped; Qualtrics IDs, timings and the authors’ derived dummies dropped. | 2026-10-07 | |
digiuseppe_2022_tshirt_usa |
Conjoint table (irw_conjoint). choice: one of three T-shirts per task (question wording not deposited; exactly one chosen in every task, no opt-out recorded). Attribute levels are the .dta value labels; price shown as 5/10/15/20 (display format unknown). Restrictions and attribute order undocumented. MTurk IDs, ZIP codes and derived dummies dropped; 44 respondents lack wave-1 covariates. | 2026-10-08 | |
digiuseppe_2022_tv_india |
Conjoint table (irw_conjoint). choice: one of three televisions per task (question wording not deposited; exactly one chosen in every task, no opt-out recorded). Attribute levels as stored text; price as value labels such as 1,00,000. Restrictions, attribute order and survey language undocumented. Worker-ID column, free-text city and occupation, and derived dummies dropped; birth year kept as recorded, with implausible values. | 2026-10-08 | |
ding_2020_covid_knowledge |
Each tick-all-that-apply option of the knowledge questions is its own item, and resp is scored against the questionnaire’s printed key (1 = the respondent’s tick matches the key: a keyed option ticked, or a distractor left unticked). resp_raw holds the tick itself (1/0) and, for Q11, the option chosen (1-6; option 5 is keyed). The key is the authors’ as of February 2020; Q10D (fecal-oral transmission) is keyed incorrect and 68% ticked it. Most items are near ceiling. | #2940 | 2026-10-10 |
ding_2020_covid_risk_perception |
Four author-written items on 1-5 scales whose anchors differ by item (likelihood, frequency of reminders, worry); see the item text table. | #2940 | 2026-10-10 |
diredja_2025_utaut2 |
The record title says 31 items; the file has 29. | 2026-10-10 | |
dobson_2026_legislative_professionalism |
Conjoint table (irw_conjoint). Three forced choices per pair of legislatures: choice = which is best positioned to benefit society, choice_people_like_you = to benefit people like you (wordings from the authors’ figure titles), choice_q3 = the paper’s Figure 1 outcome, wording not deposited. Task 2 is dropped because one attribute of its profile A is missing for all respondents in the deposit. Salary $0k and session Unlimited appear less often than other levels (non-uniform weights). Attribute names are assigned from the level text; attribute order randomized per respondent (attrpos_*). | 2026-10-07 | |
doherty_2019_party_chairs |
Conjoint table (irw_conjoint). choice: ‘Which candidate do you think would fare best in a [Democratic/Republican] primary in the area for which you are responsible as a party chairperson?’, forced choice, no opt-out. Each profile showed a name plus 4 of 10 attribute rows drawn by a fixed scheme with non-uniform level weights (restriction); undrawn rows are ‘(not shown)’. Occupation, experience, military and descriptor levels are the authors’ classes, not the displayed wording. The 0-100 ‘how much better’ follow-up and noised county measures are not included. | 2026-10-07 | |
doherty_2020_blame_inaction |
Conjoint table (irw_conjoint_2). Factorial newsletter vignette, one per respondent: policy area, author’s chamber and party, bill status (passed / unlikely to pass) and an optional blame sentence (opposing chamber or opposing party), the last only when the bill is unlikely to pass. Six outcomes: 0-100 feelings ratings of the author, Congress, Senate, House, Democrats and Republicans, and a 1-4 policy-support rating (1 = most supportive, as stored); no opt-out. The party named in the party-blame arm is not stored and appears as a placeholder. Dollar-allocation outcomes dropped. | 2026-10-14 | |
doidge_2019_vet_antimicrobials |
Conjoint table (irw_conjoint). Factorial vignette survey, one vignette per task. rating: how likely the vet in the scenario is to prescribe without visiting the farm, circled -5 (definitely would not) to +5 (definitely would), stored by the authors as 1-11; rating_others: proportion of vets who would prescribe, grouped to 10% bands and coded 1-10. Vignettes drawn from 384 combinations by a resolution-V D-efficient design; factor order fixed. Level text is the clause printed in the example survey. 10 vignettes with neither outcome omitted. | 2026-10-08 | |
donnaloja_2022_citizenship |
“please choose to whom you want to grant citizenship. You may choose ONE, BOTH or NEITHER in each pair” is coded per profile: rating 1 = “Should be granted British citizenship”, 0 = not; no choice column. Restrictions: refugee status shown only for Pakistan/Nigeria/Syria/Somalia (otherwise attr_refugee = “(not shown)”), Ireland and Australia always excellent English, Poland never Muslim; attr_ancestry “(not shown)” (no ancestry clause) is the paper’s “Neither” level. All 1,648 respondents kept (the paper analyses the 1,597 British citizens); the paper’s Table 1 appears to swap the Germany and Poland counts. | #2887 | 2026-10-07 |
donnaloja_2023_birthright_citizenship |
“Per favore, indica a chi daresti la cittadinanza italiana alla nascita” (Profilo A / Profilo B / Entrambi / Nessuno), coded per profile: rating 1 = would grant citizenship at birth, 0 = not. Romania never appears with “Senza permesso di soggiorno”; attribute order randomized but not recorded. 58 non-citizen respondents and the YouGov panel ID dropped; N = 1,463 matches the paper. | #2887 | 2026-10-07 |
donnan_2023_cannabis_flower |
Conjoint table (irw_conjoint). Discrete choice experiment, Sawtooth CBC, 300 D-efficient versions of 8 tasks with two unlabelled options (Option A/B); choice = the option chosen, forced, no opt-out. Level text from the deposit Read Me (English; survey also fielded in French). Only complete respondents kept, as in the papers (891). Licence carried as CC BY-NC 4.0 (Read Me forbids commercial use). | 2026-10-09 | |
donnan_2023_cannabis_vape |
Conjoint table (irw_conjoint). Discrete choice experiment, Sawtooth CBC, 300 D-efficient versions of 8 tasks with two unlabelled options (Option A/B); choice = the option chosen, forced, no opt-out. Level text from the deposit Read Me (English; survey also fielded in French). Only complete respondents kept, as in the papers (385). Licence carried as CC BY-NC 4.0 (Read Me forbids commercial use). | 2026-10-09 | |
donnan_2024_cannabis_edible |
Conjoint table (irw_conjoint). Discrete choice experiment, Sawtooth CBC, 300 D-efficient versions of 8 tasks with two unlabelled options (Option A/B); choice = the option chosen, forced, no opt-out. Level text from the deposit Read Me (English; survey also fielded in French). Only complete respondents kept, as in the papers (684). Licence carried as CC BY-NC 4.0 (Read Me forbids commercial use). | 2026-10-09 | |
donnan_2024_cannabis_retailer |
Conjoint table (irw_conjoint). Discrete choice experiment, Sawtooth CBC, 300 D-efficient versions of 6 tasks with two unlabelled options (Option A/B); choice = the option chosen, forced, no opt-out. Level text from the deposit Read Me (English; survey also fielded in French). Only complete respondents kept, as in the papers (1,626). Licence carried as CC BY-NC 4.0 (Read Me forbids commercial use). | 2026-10-09 | |
dopmeijer_2022_loneliness |
resp runs in the opposite direction to the source’s SPSS file and paper: here 1 = ‘No! Totally disagree!’ through 5 = ‘Yes! Totally agree!’, whereas the deposit codes 1 = ‘Yes! Totally agree!’ through 5 = ‘No! Totally disagree!’ on all eleven items. IRW numbers the response labels from least to most agreement, as for this study’s other scales. To join resp back to the deposit or read it with the deposit’s codebook, use 6 - resp. | #2120 | 2026-10-01 |
dow_2023_vigilantism |
Conjoint table (irw_conjoint). choice = “Of the two scenarios, in which case do you think it would be better for someone in the community rather than the police to punish the perpetrator directly?” and choice_punish = “which case do you think should be punished more severely?”, both forced choice between Case 1 and Case 2, no opt-out. Attribute text is the authors’ English factor labels (survey in Spanish); perpetrator labels differ slightly from the codebook’s. Restrictions and attribute order not documented; N = 9,365 matches the abstract’s “over 9000 households”. | 2026-10-09 | |
drent_2026_teacher_expectations |
Conjoint table (irw_conjoint). Single-vignette ratings: rating_recommendation = transfer recommendation (0 lowest track, 1 highest track); rating_expectation = ‘How likely do you think it is that this student will graduate from the highest attainable track?’ 1 Unlikely - 5 Highly likely (English translation). Full 2^5 factorial, 6 random vignettes per teacher; levels are codebook category labels, not the Dutch display text; task order not recorded. Free-text explanations dropped; 722 rows vs codebook 721. | 2026-10-09 | |
drewniak_2022_ecls_start |
Conjoint table (irw_conjoint). Factorial vignette survey, ECLS initiation scenario, 5 vignettes per respondent. rating_self ‘From my point of view, ECLS/ECMO should be used in this patient’, rating_clinic ‘In my clinic, ECLS/ECLS would be used in this patient’, 1 not correct at all to 6 fully correct (translated). Fixed blocked design; 80- and 90-year-old patients never bridged to transplant. Task is not display order (not recorded). Deposit has 510 respondents vs 420 analysed in the article (exclusions not applied). Level text is the article’s English vignette wording. | 2026-10-09 | |
drewniak_2022_ecls_stop |
Conjoint table (irw_conjoint). Factorial vignette survey, ECLS withdrawal scenario, 5 vignettes per respondent. rating_self ‘From my point of view, the ECLS/ECLS treatment should be discontinued’, rating_clinic ‘In my clinic, the ECLS/ECLS treatment would be discontinued’, 1 not correct at all to 6 fully correct (translated); higher = more readiness to withdraw. Fixed blocked design; 80- and 90-year-old patients never bridged to transplant. Task is not display order (not recorded). Deposit has 467 respondents vs 420 analysed in the article. Level text is the article’s English vignette wording. | 2026-10-09 | |
driscoll_2023_court_curbing |
Conjoint table (irw_conjoint). choice is which of two candidates the respondent would be more likely to support (no opt-out described; 2,430 tasks unanswered, choice blank); rating and rating_mobilize are 5-point candidate-rating and mobilization items stored in the authors’ recode (higher = more favourable; wording and anchors not deposited). Within every pair the candidates take opposite positions on all issues and on court curbing, and opposite parties in the 10 general-election trials (trial_election); the court-curbing proposal is fixed by trial (trial_court_proposal). Level text is the authors’ labels; the candidate-name attribute (unlabelled codes) and the legitimacy factor score are dropped. 962 respondents vs about 1,000 in the article. | 2026-10-07 | |
duch_2021_vaccine_priority |
Conjoint table (irw_conjoint). Outcome: “Which of the Persons do you think should get the vaccine immediately? Select one of them.” (UK questionnaire), Person A / Person B, forced choice, no opt-out; the 1-7 priority ratings were asked but are not deposited. 13 countries pooled (cov_country); levels stored in English, other display languages undocumented. cov_survey_weight is 1 throughout Canada, Spain, Uganda and India. | 2026-10-07 | |
duch_2022_destination_country |
Conjoint table (irw_conjoint). choice = ‘which of the employment destinations do you prefer?’ (forced choice, no opt-out); rating = approval of each destination, 1 strongly disapprove to 7 strongly approve. Levels randomly assigned; attribute order randomized but not stored. Four student pools pooled with cov_country; stored levels are the authors’ English text (Chile and China saw Spanish/Chinese). Conjoint 3 of 3 (country-label version). | 2026-10-07 | |
duch_2022_destination_deportation |
Conjoint table (irw_conjoint). choice = ‘which of the employment destinations do you prefer?’ (forced choice, no opt-out); rating = approval of each destination, 1 strongly disapprove to 7 strongly approve. Levels randomly assigned; attribute order randomized but not stored. Four student pools pooled with cov_country; stored levels are the authors’ English text (Chile and China saw Spanish/Chinese). Conjoint 2 of 3 (deportation version). | 2026-10-07 | |
duch_2022_destination_muslim_ban |
Conjoint table (irw_conjoint). choice = ‘which of the employment destinations do you prefer?’ (forced choice, no opt-out); rating = approval of each destination, 1 strongly disapprove to 7 strongly approve. Levels randomly assigned; attribute order randomized but not stored. Four student pools pooled with cov_country; stored levels are the authors’ English text (Chile and China saw Spanish/Chinese). Conjoint 1 of 3 (Muslim-ban version). | 2026-10-07 | |
duch_2025_covid_incumbents |
Conjoint table (irw_conjoint). Single profile per round. choice: ‘Please indicate whether you think YES this Government Leader should be re-elected or NO…’ (1 = YES; NO is the opt-out); rating: 1 (Very Strongly Oppose) to 7 (Very Strongly Support) re-election. Pooled 16 countries with cov_country; attribute text is the English display text (translations shown outside English-speaking countries); 221 Brazilian rounds with a mistranslated jobs level dropped; respondents 22,180 vs 22,147 in the article. | 2026-10-09 | |
duell_2023_rebel_mps |
Conjoint table (irw_conjoint). Five MP vignettes per respondent; rating_represented ‘How well do you feel represented by this MP?’ (1 not at all - 7 very well), rating_like_mp ‘Every [nationality] MP should be like this MP’ (1 strongly disagree - 7 strongly agree), rating_trait_* = traits chosen (up to 3) after the last vignette, on task 5 only. Bill direction and public opinion appear only in some information arms (trial_info_version) and are otherwise ‘(not shown)’; bill direction depends on the respondent’s preference; free text and unlabelled items dropped; N = 14,000 as in the article. | 2026-10-09 | |
duell_2025_party_responses |
Conjoint table (irw_conjoint). Factorial vignette, single profile: rating = how much the respondent likes the described response of their first-ranked party, 1 (do not like at all) - 7 (like a lot) (paraphrase). Levels are author labels (attack, response) and article paraphrases (policy demand), not the displayed text. 39 vignettes without rating or attack type omitted; 66 respondents have 3 vignettes. | 2026-10-09 | |
dufresne_2026_immigrant_drivers |
Conjoint table (irw_conjoint_2). Single 2x2 vignette per respondent (speaks the language x driving training from the receiving country) about an immigrant driver; choice = 1 if the respondent prefers the immigrant driver, 0 if the automated driving system (a fixed alternative with no attributes, so opt_out = yes). Level text is the authors’ condition labels, not the questionnaire wording; countries pooled (CA, JP, US); 4 rows without an outcome omitted; N against the paper not checked. | 2026-10-11 | |
dukovska_2022_p2p_electricity |
Conjoint table (irw_conjoint_2). Choice of one of three electricity bundles (4 attributes) in 10 tasks; forced choice, no none option. Attribute levels are the authors’ English translation of the Dutch survey, with code order taken from the Qualtrics level list; the design file and responses were joined on the Qualtrics version number. 329 respondents (12 did not finish the later background questions but answered all tasks); gender, birth year, education (Dutch), province kept. | 2026-10-12 | |
dur_2023_trade_agreement_groups |
Conjoint table (irw_conjoint). Paired choice between two trade agreements (Agreement A, Agreement B or Neither; Neither = opt-out, 411 tasks) and a rating of each agreement 1 (not favourably at all) to 10 (very favourably); wording paraphrased from the article. Respondents are interest groups, not individuals. 5 tasks with no answer omitted. | 2026-10-08 | |
dur_2025_trade_poland |
Conjoint table (irw_conjoint). Outcomes: forced choice between two trade agreements and a 1-7 rating of each (1 ‘muy en contra’ .. 7 ‘muy a favor’); no opt-out. Payoff levels are an English rendering of the article’s example (zloty, 5 x the euro value); randomization excluded agreements with losses on all three payoffs and pairs identical on the payoffs. 2,999 respondents (article: 3,000; one respondent has no conjoint rows). The authors pooled this sample with the Spanish one (dur_2025_trade_spain). | 2026-10-08 | |
dur_2025_trade_spain |
Conjoint table (irw_conjoint). Outcomes: forced choice between two trade agreements and a 1-7 rating of each (1 ‘muy en contra’ .. 7 ‘muy a favor’); no opt-out. Payoff levels are an English rendering of the article’s example (euro); randomization excluded agreements with losses on all three payoffs and pairs identical on the payoffs. 2,999 respondents (article: 3,000; one respondent has no conjoint rows). The authors pooled this sample with the Polish one (dur_2025_trade_poland). | 2026-10-08 | |
dur_2025_trade_uk |
Conjoint table (irw_conjoint). Outcomes: forced choice between two trade agreements and a 1-7 rating of each (1 ‘strongly oppose’ .. 7 ‘strongly favour’); no opt-out. Pairs identical on the two payoff attributes were redrawn; attribute order randomized (payoffs kept together) but not recorded. N = 2,128 matches the article, and its Table 4 reproduces exactly. | 2026-10-08 | |
duraj_2026_pe |
Wave-1 item 2 is the deposit’s misnamed column PE2_M2 and item 10 its ‘@_M1’; the authors’ reverse-scored copies are not shipped. See data/duraj_2026_math_anxiety_expectancies.py. | 2026-10-11 | |
duraj_2026_pee |
Wave-1 item 2 is the deposit’s misnamed column PEE2_M2; the authors’ reverse-scored copies are not shipped. | 2026-10-11 | |
duraj_2026_te |
One value of 5 on the 1-4 item TE1 at wave 2 dropped as a data-entry error. | 2026-10-11 | |
dusoruth_2020_food_waste |
Conjoint table (irw_conjoint_2). Single-profile design: each respondent rated 8 product scenarios (ground beef or spinach; appearance photo, date label, days to expiration, package size, price); rating = food waste tendency 0-100 (higher = more discarding), stored raw; no choice. Task is inferred from row order; appearance levels are described from the article, not shown text; the 12 profiles per product are an orthogonal fractional design in which price is nested in package size. 333 respondents (matches the article); 4 have 7 rows. | 2026-10-12 | |
eckerehrhardt_2025_energy_governance |
Conjoint table (irw_conjoint). choice = ‘Which proposal do you find most appropriate?’, forced, no opt-out; rating = ‘How much confidence would you have in this new organization if created as proposed?’, 1 = none at all to 7 = a lot of confidence. Four Yes/No attributes (which groups get a lot of say), randomly drawn from the 16 combinations; between-subject vignette factors (government participation, scope, costs and benefits) in trial_ fields. Four countries pooled with the authors’ weight; covariates in the deposit cannot be linked to respondents. | 2026-10-07 | |
eckerehrhardt_2025_internet_governance |
Conjoint table (irw_conjoint). choice = ‘Which proposal do you find most appropriate?’, forced, no opt-out; rating = ‘How much confidence would you have in this new organization if created as proposed?’, 1 = none at all to 7 = a lot of confidence. Four Yes/No attributes (which groups get a lot of say), randomly drawn from the 16 combinations; between-subject vignette factors (government participation, scope, costs and benefits) in trial_ fields. Four countries pooled with the authors’ weight; covariates in the deposit cannot be linked to respondents. | 2026-10-07 | |
eckhard_2022_administrative_language |
Conjoint table (irw_conjoint). One audio vignette per respondent: administrative language (Control / Highly informational / Highly relational / Highly informational and relational, each a bundle of linguistic features) x service decision (negative/positive). Ratings 1-10, higher = more: rating = “All in all, how satisfied are you with the bureaucratic encounter?”, plus comprehensibility, reification, emotionality, complaisance. Level text is the codebook’s condition names (the German recordings are not deposited). trial_listened_complete = 0 for 124 who did not listen to the end; missing codes 0/-66/-77/-99 -> NA; free text dropped. | 2026-10-08 | |
edelson_2025_political_ads |
Conjoint table (irw_conjoint). choice = which of two ads is more political (‘Which of these two ads is more political?’, pilot wording), forced choice, no opt-out. Ads were images; attribute values are the authors’ labels (sponsor, message short label, image), not the ad text. 96 tasks without an answer omitted; task order as display order not documented. | 2026-10-08 | |
eggers_2018_mp_misconduct |
Conjoint table (irw_conjoint). choice: “If you were living in this constituency at the next general election, which party would you vote for?” (current MP = profile 1, challenger = profile 2); opt-outs (third-party candidate, another party, would not vote) leave choice = 0 on both profiles in 4,305 of 9,810 tasks. The MP is always Labour or Conservative and the challenger a different party; the conduct line describes the MP only. Respondent party-identification flags dropped as derived; N = 1,962 matches the article and Table 1 col. 1 reproduces exactly. | 2026-10-07 | |
eggers_2025_council_office |
Conjoint table (irw_conjoint). rating = interest in running for the described city-council seat, 1 ‘No interest whatsoever’ to 10 ‘Highest possible interest’ (question wording not deposited); single-profile text vignette, 8 per respondent. Sentence order randomized once per respondent (attrpos_); level probabilities not stated. N = 3,000 as in the paper’s Table 1; YouGov weight kept. | 2026-10-08 | |
ehret_2022_ai_governance |
Conjoint table (irw_conjoint_2). Which of two AI algorithms should be prohibited (choice = 1 for the one chosen for prohibition, forced, no opt-out); five samples pooled (cov_country). Level text is the appendix Table 1 wording recoded from short codes; profile order inferred from row order; the 1-7 approval rating is not in the deposit. N = 932 respondents, as in the paper’s Table 7. | 2026-10-11 | |
elia_2024_scandal_women_candidates |
Conjoint table (irw_conjoint_2). rating_vote = likelihood of voting for the candidate, 1 very unlikely to 5 very likely; rating_tell = a second 5-point likelihood item (same scale; its wording is not in the deposit or the paper’s open text); rating_trust = trust in the candidate, 1 none to 4 a lot (‘No se’ is missing); question wording is not in the deposit. Attribute levels are the authors’ condition labels (Man/Woman, No scandal/Corruption scandal); the Spanish vignette text sits in the paper’s SI. The 3,215 respondents are the authors’ analysis sample. | 2026-10-11 | |
elkjaer_2025_inheritance_tax |
Conjoint table (irw_conjoint_2). Pairs of tax plans (A/B), 5 tasks; each plan sets four income-tax and four inheritance-tax rates; choice = preferred plan on income tax, choice_inheritance = preferred plan on inheritance tax (paraphrases; both forced, no opt-out). The deposit has no questionnaire, so the joint display of the two rate sets is inferred from the row structure. Survey weight (w8) and the information-treatment arm are kept. | 2026-10-14 | |
elmendorf_2024_housing_platforms |
Conjoint table (irw_conjoint). choice: “Given this choice, which set of positions would you prefer?” between Set A and Set B, forced choice, no opt-out. Each set lists three policy positions; the two sets take opposite positions on the same three issues, so attributes are per issue with ‘(not shown)’ when absent. Task 6 repeats task 1 with the sets swapped (and an issue-name column shown). Authors’ filters applied (finished, screening passed, speeders dropped): 5,127 respondents. | 2026-10-08 | |
elshafie_2020_edsc |
Items above a child’s age band are recorded 0 (No); the deposit does not distinguish ‘not yet achieved’ from ‘not asked’ (the form is Yes/No only). Q1-Q3 are passed by every child. | 2026-10-05 | |
encceja_2018_ef_ci |
Item 90388 (biserial below -0.15) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (90388); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_ef_ci_nom |
Item 90388 (biserial below -0.15) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (90388); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_ef_hg |
Items 56503 (BILOG run failed), 92152 (BILOG run failed), 94391 (BILOG run failed) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_ef_hg_nom |
Items 56503 (BILOG run failed), 92152 (BILOG run failed), 94391 (BILOG run failed) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_ef_lp |
Items 56947 (biserial below -0.15), 76391 (BILOG run failed) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (56947); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_ef_lp_nom |
Items 56947 (biserial below -0.15), 76391 (BILOG run failed) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (56947); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_ef_ma |
Items 59204 (BILOG run failed), 76773 (biserial below -0.15), 77815 (BILOG run failed), 94604 (biserial below -0.15) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (76773, 94604); consider leaving them out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_ef_ma_nom |
Items 59204 (BILOG run failed), 76773 (biserial below -0.15), 77815 (BILOG run failed), 94604 (biserial below -0.15) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (76773, 94604); consider leaving them out when fitting a model. | #2644 | 2026-10-05 |
encceja_2018_em_cn |
Item 91064 (excluded because it prevented convergence) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_em_cn_nom |
Item 91064 (excluded because it prevented convergence) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_em_lc |
Item 91564 (BILOG run failed) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_em_lc_nom |
Item 91564 (BILOG run failed) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_em_mt |
A very hard form: 17 of the 30 items have proportion correct below chance (.25). The key is INEP’s: it matches the per-candidate key string in the microdata, the summed score correlates .75 with INEP’s own score NU_NOTA_MT, and INEP’s 3PL difficulties for these items run from b = 1.1 to 6.5. Expect weak item-total correlations, not a keying error. | #2644 | 2026-10-05 |
encceja_2018_em_mt |
Item 57841 (BILOG run failed) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2018_em_mt_nom |
A very hard form: 17 of the 30 items have proportion correct below chance (.25). The key is INEP’s: it matches the per-candidate key string in the microdata, the summed score correlates .75 with INEP’s own score NU_NOTA_MT, and INEP’s 3PL difficulties for these items run from b = 1.1 to 6.5. Expect weak item-total correlations, not a keying error. | #2644 | 2026-10-05 |
encceja_2018_em_mt_nom |
Item 57841 (BILOG run failed) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_ci |
Item 93848 (biserial below -0.15) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (93848); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2019_ef_ci |
Item 56777 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 1.25, b = 3.05), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A stays at about .2 from the bottom 5% to the top 5%, while the share choosing C rises from .23 to .58. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2019_ef_ci_nom |
Item 93848 (biserial below -0.15) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below -0.15 means the item runs against ability (93848); consider leaving it out when fitting a model. | #2644 | 2026-10-05 |
encceja_2019_ef_ci_nom |
Item 56777 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 1.25, b = 3.05), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A stays at about .2 from the bottom 5% to the top 5%, while the share choosing C rises from .23 to .58. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2019_ef_hg |
Item 57191 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_hg_nom |
Item 57191 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_lp |
Item 80884 (abandoned in an earlier edition) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_lp_nom |
Item 80884 (abandoned in an earlier edition) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_ma |
Items 76092 (convergence problems), 94314 (convergence problems), 94651 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_ma |
Item 59479 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on B, and INEP kept the item in its calibration (a = 3.00, b = 2.66), so the table scores B. But ranking candidates by their NU_NOTA in the other areas, the share choosing B stays at about .2 from the bottom 5% to the top 5%, while the share choosing C rises from .29 to .45. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2019_ef_ma_nom |
Items 76092 (convergence problems), 94314 (convergence problems), 94651 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_ef_ma_nom |
Item 59479 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on B, and INEP kept the item in its calibration (a = 3.00, b = 2.66), so the table scores B. But ranking candidates by their NU_NOTA in the other areas, the share choosing B stays at about .2 from the bottom 5% to the top 5%, while the share choosing C rises from .29 to .45. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2019_em_cn |
Items 58710 (convergence problems), 91878 (convergence problems), 92179 (convergence problems), 94194 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_em_cn_nom |
Items 58710 (convergence problems), 91878 (convergence problems), 92179 (convergence problems), 94194 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_em_lc |
INEP’s two keys disagree on item 58882 (position 19): the item file says C, the key string INEP ships per candidate says B. The table scores B, the letter whose choice correlates with INEP’s own score NU_NOTA_LC (r = .18, against -.10 for C). | #2644 | 2026-10-05 |
encceja_2019_em_lc |
Items 79691 (convergence problems), 93468 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_em_lc_nom |
INEP’s two keys disagree on item 58882 (position 19): the item file says C, the key string INEP ships per candidate says B. The table scores B, the letter whose choice correlates with INEP’s own score NU_NOTA_LC (r = .18, against -.10 for C). | #2644 | 2026-10-05 |
encceja_2019_em_lc_nom |
Items 79691 (convergence problems), 93468 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_em_mt |
Items 87599 (convergence problems), 93013 (convergence problems), 94554 (convergence problems), 117356 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2019_em_mt_nom |
Items 87599 (convergence problems), 93013 (convergence problems), 94554 (convergence problems), 117356 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_ef_ci |
Items 91305 (convergence problems), 125054 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_ef_ci_nom |
Items 91305 (convergence problems), 125054 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_ef_hg |
Item 56638 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 0.27, b = 0.44), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A peaks at .70 mid-distribution and falls to .56 in the top 5%, while the share choosing D rises from .07 to .18. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2020_ef_hg_nom |
Item 56638 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 0.27, b = 0.44), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A peaks at .70 mid-distribution and falls to .56 in the top 5%, while the share choosing D rises from .07 to .18. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2020_ef_ma |
Items 81263 (convergence problems), 93906 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_ef_ma_nom |
Items 81263 (convergence problems), 93906 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_em_ch |
Items 93057 (convergence problems), 124698 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_em_ch |
Item 57810 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 1.38, b = 2.85), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A falls from .42 to .33 before recovering to .40 in the top 5%, while the share choosing D rises from .12 to .38. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2020_em_ch_nom |
Items 93057 (convergence problems), 124698 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_em_ch_nom |
Item 57810 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on A, and INEP kept the item in its calibration (a = 1.38, b = 2.85), so the table scores A. But ranking candidates by their NU_NOTA in the other areas, the share choosing A falls from .42 to .33 before recovering to .40 in the top 5%, while the share choosing D rises from .12 to .38. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2020_em_lc |
Items 93992 (convergence problems), 126136 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2020_em_lc_nom |
Items 93992 (convergence problems), 126136 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2022_ef_ci |
Item 137253 (biserial below 0.01) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below 0.01 means the item barely discriminates (137253). | #2644 | 2026-10-05 |
encceja_2022_ef_ci_nom |
Item 137253 (biserial below 0.01) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. A biserial below 0.01 means the item barely discriminates (137253). | #2644 | 2026-10-05 |
encceja_2022_ef_ma |
Items 76191 (convergence problems), 94028 (convergence problems), 117449 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2022_ef_ma_nom |
Items 76191 (convergence problems), 94028 (convergence problems), 117449 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2022_em_ch |
Item 124903 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2022_em_ch_nom |
Item 124903 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2022_em_cn |
Item 80762 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on C, and INEP kept the item in its calibration (a = 1.62, b = 2.71), so the table scores C. But ranking candidates by their NU_NOTA in the other areas, the share choosing C stays near .3 throughout, while the share choosing B rises from .14 to .45. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2022_em_cn_nom |
Item 80762 may be miskeyed or flawed. INEP’s item file and per-candidate key string agree on C, and INEP kept the item in its calibration (a = 1.62, b = 2.71), so the table scores C. But ranking candidates by their NU_NOTA in the other areas, the share choosing C stays near .3 throughout, while the share choosing B rises from .14 to .45. Consider leaving it out when fitting a model; the _nom table has the letters for rescoring. | #2644 | 2026-10-06 |
encceja_2022_em_mt |
Items 117247 (biserial below 0.01), 117412 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below 0.01 means the item barely discriminates (117247). | #2644 | 2026-10-05 |
encceja_2022_em_mt_nom |
Items 117247 (biserial below 0.01), 117412 (convergence problems) are ones INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). They still have valid keys, so they stay in the table, as ENEM’s abandoned items do. A biserial below 0.01 means the item barely discriminates (117247). | #2644 | 2026-10-05 |
encceja_2023_em_ch |
Item 136996 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2023_em_ch_nom |
Item 136996 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2023_em_cn |
Item 58369 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2023_em_cn_nom |
Item 58369 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2023_em_mt |
Item 142265 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2023_em_mt_nom |
Item 142265 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2024_ef_lp |
Item 125160 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2024_ef_lp_nom |
Item 125160 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2024_ef_ma |
Item 77200 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
encceja_2024_ef_ma_nom |
Item 77200 (convergence problems) is one INEP dropped from its own IRT calibration and scoring (IN_ITEM_ABAN = 1 in the item file). It still has a valid key, so it stays in the table, as ENEM’s abandoned items do. | #2644 | 2026-10-05 |
endo_2025_candidate_ambition |
Conjoint table (irw_conjoint). 2 x 2 factorial vignette (candidate gender x ambition-revealing vs ambition-hiding motivation), one vignette per respondent, shown in Japanese; attribute text is the authors’ English translation. Ratings: favorability 1-4 (4 very favorable), expected popularity 1-5 (5 very likely), six perceived traits 5-point with undocumented anchors; wording paraphrased. 3,151 respondents in the deposit; the 2023 working paper reports 3,168 and 3,867. | 2026-10-08 | |
enem_2024_1mil_ch |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_ch_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_cn |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_cn_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_lc |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_lc_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_mt |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2024_1mil_mt_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_ch |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_ch_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_cn |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_cn_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_lc |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_lc_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_mt |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
enem_2025_1mil_mt_nom |
id is INEP’s NU_SEQUENCIAL from the RESULTADOS file, a surrogate that is unique within a year but not across years: INEP appears to reuse the same shuffle, so the same number is a different person in 2024 and 2025. Do not use id to follow anyone between years, or to link to INEP’s PARTICIPANTES file, which INEP ships with no key to RESULTADOS. | #2812 | 2026-10-05 |
erlich_2023_ukraine_candidates |
Conjoint table (irw_conjoint). rating = ‘Would you vote for these candidates?’ and rating_win = ‘Do you think these candidates could win local election?’, each asked of both candidates (1 = yes, 0 = no; one, both or neither), translated from the Ukrainian/Russian survey. Attribute order below the name randomized per respondent (attrpos_). A 25-year-old never has 15 years of experience (the authors’ design file also excludes age 35 with 15 years, but the data contain it). 46 speeders dropped as in the paper; N = 2,155 matches. | 2026-10-07 | |
erlich_2025_bribery_clinic |
Conjoint table (irw_conjoint). choice = in which of two side-by-side scenarios (A/B) Petro would be more likely to pay the doctor a bribe (forced, no opt-out); rating = how likely Petro would be to pay a bribe to get treatment more quickly, rating_self = how likely the respondent would be in Petro’s position, both 1 = definitely no to 7 = definitely yes (mapped from Russian answer text; higher = more likely to bribe). Attribute levels are stored as displayed in Ukrainian or Russian, so each level has two spellings and cross-language level pairs never co-occur. Tasks without a forced-choice answer dropped; 3,037 of the 3,060 analysed respondents have this scenario, matching the authors’ long file. | 2026-10-08 | |
erlich_2025_bribery_license |
Conjoint table (irw_conjoint). choice = in which of two side-by-side scenarios (A/B) Nina would be more likely to pay the instructor a bribe (forced, no opt-out); rating = how likely Nina would be to pay a bribe to get the license more quickly, rating_self = how likely the respondent would be in Nina’s position, both 1 = definitely no to 7 = definitely yes (mapped from Russian answer text; higher = more likely to bribe). Attribute levels are stored as displayed in Ukrainian or Russian, so each level has two spellings and cross-language level pairs never co-occur. Tasks without a forced-choice answer dropped; 3,031 of the 3,060 analysed respondents have this scenario, matching the authors’ long file. | 2026-10-08 | |
erlich_2025_corruption_harms |
Conjoint table (irw_conjoint). Five forced choices per pair of corruption scenarios (Scenario A/B): which harms you and your family most, harms Armenia’s economy more, reduces confidence in the political system more, is more morally wrong, happens more often (wording from the English instrument); 98/99 answers set to NA; no opt-out. Restricted randomization: favors to family and use of public resources for own firm occur only with personal use and keeping all funds; political use never with keeping all funds. Tasks with a mismatched screen number dropped. cov_survey_weight = authors’ age x gender post-stratification weight. N 1,497 (authors 1,495). | 2026-10-09 | |
escribafolch_2021_democracy_promotion |
Conjoint table (irw_conjoint). For each of two autocratic countries per task respondents said whether they would support US military intervention, economic sanctions and democracy aid (rating_military_intervention, rating_economic_sanctions, rating_democracy_aid; 1 = yes, 0 = no; any number of tools or none), so there is no choice between profiles. Attribute row order was randomized per task and is recorded in attrpos_; task 5 of the 101 wave-2 records is dropped because its levels were not saved. The deposit’s 1,591 records (2017) exceed the article’s 1,464 (dated 2016). | 2026-10-07 | |
eshima_2022_age_candidates |
Conjoint table (irw_conjoint). choice is a forced choice between two hypothetical upper-house candidates; question wording not deposited. Attribute levels are the authors’ English labels for Japanese text; task and profile are inferred from row order of the authors’ processed file. trial_treat_group holds four randomized arms whose content is undocumented. Qualtrics response IDs dropped; no weights. | 2026-10-07 | |
esteves_2026_dddt |
Only the deposit’s first sample (Sujeito 1-300 of 600) answered the Dirty Dozen. | 2026-10-11 | |
everitt_2021_candidate_orientation |
Conjoint table (irw_conjoint_2). Within-subject two-screen vignette: respondents rate the likelihood of voting for a candidate (1-10, in rating) first when only gender is shown (orientation and marital status ‘(not shown)’), then after sexual orientation and marital status are added; no opt-out. Missing ratings drop that task (888 of 935 respondents have both, matching the paper’s analysed N). Question wording and anchors are not in the deposit; marital-status text is paraphrased from the article. | 2026-10-14 | |
evers_2019_audience_costs |
Conjoint table (irw_conjoint_2). Single-vignette factorial survey (president x action, 3x3): rating = presidential approval, 0 = strongly disapprove to 4 = strongly approve (value labels; question wording not in the deposit). Levels are the authors’ short treatment labels, not the scenario text. All 1,991 respondents kept (cov_attention_pass = the authors’ manipulation-check flag; the authors analyse 1,699); id is the row number. | 2026-10-11 | |
fabbe_2025_refugee_sites_citizens |
Conjoint table (irw_conjoint). Three paired proposals for an asylum-seeker host site; choice = which mayoral candidate (backing Proposal A or B) the respondent would vote for (paraphrase; forced, no opt-out), rating = each proposal on a 0-7 scale (anchors not deposited; higher = more support, agrees with the choice in ~95% of pairs). Levels are English as exported; 232 unanswered tasks omitted; N = 5,916. | 2026-10-09 | |
fabbe_2025_refugee_sites_councilors |
Conjoint table (irw_conjoint). Three paired proposals for an asylum-seeker host site; choice = the preferred proposal (A or B) (paraphrase; forced, no opt-out), rating = each proposal on a 0-7 scale (anchors not deposited; higher = more support, agrees with the choice in ~95% of pairs). Levels Greek as exported; 25 non-responders omitted; N = 626 vs 586 reported; answered task-3 choice kept (authors impute it from ratings). | 2026-10-09 | |
fang_2021_sn |
The paper’s questionnaire has six social-norm items but the data file has five, and the data’s SN1-SN5 do not correspond to the paper’s SN1-SN5 (by means, data SN1/2/4/5 are the paper’s SN3/4/5/6 and data SN3 matches no printed item). Treat the item identities as uncertain. | 2026-10-04 | |
fang_2026_anime_museum |
Item codes from the processed sheet; stems from the raw workbook by the deposit’s item number. | 2026-10-10 | |
faure_2025_climate_migrants_nl |
Conjoint table (irw_conjoint). Forced choice between two migrant profiles (which to admit; wording not deposited, paraphrase), no opt-out, 6 tasks. Levels stored in Dutch as displayed. Some country x reason pairs never occur (Maldives with earthquake or drought; Pakistan and Vietnam with drought), undocumented. Qualtrics response IDs dropped; education and party preference kept as unlabelled codes; survey weight kept. | 2026-10-08 | |
favero_2020_covid_attitudes |
One respondent’s second submission (ResponseId R_1FJTuRyeH7msWDG) is dropped, as in the authors’ do-file. Five arms, so the arm is cov_arm rather than treat. | 2026-10-09 | |
favero_2020_prosocial_motivation |
One respondent’s second submission (ResponseId R_1FJTuRyeH7msWDG) is dropped, as in the authors’ do-file. Five arms, so the arm is cov_arm rather than treat. | 2026-10-09 | |
favero_2020_social_distancing |
One respondent’s second submission (ResponseId R_1FJTuRyeH7msWDG) is dropped, as in the authors’ do-file. Five arms, so the arm is cov_arm rather than treat. | 2026-10-09 | |
feng_2024_org_support |
cov_age is an age-band code, not age in years: 1 = “25岁及以下”; 2 = “26-35岁”; 3 = “36-45岁”; 4 = “46-55岁”; 5 = “56-65岁”; 6 = “66岁以上”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
feng_2024_professional_development |
cov_age is an age-band code, not age in years: 1 = “25岁及以下”; 2 = “26-35岁”; 3 = “36-45岁”; 4 = “46-55岁”; 5 = “56-65岁”; 6 = “66岁以上”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
feng_2024_work_engagement |
cov_age is an age-band code, not age in years: 1 = “25岁及以下”; 2 = “26-35岁”; 3 = “36-45岁”; 4 = “46-55岁”; 5 = “56-65岁”; 6 = “66岁以上”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
feng_2024_work_values |
cov_age is an age-band code, not age in years: 1 = “25岁及以下”; 2 = “26-35岁”; 3 = “36-45岁”; 4 = “46-55岁”; 5 = “56-65岁”; 6 = “66岁以上”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
ferrari_2025_status_threat |
Conjoint table (irw_conjoint). choice: “which one would you choose?” (forced, no opt-out); rating: 1 = definitely would NOT vote to 7 = definitely would vote. About a quarter of respondents saw ‘Independent’ on every profile (no-party arm, trial_status_threat holds the vignette arm). Issue order randomized but not recorded. 2,042 in the deposit vs 2,080 collected per the paper. | 2026-10-07 | |
ferree_2023_covid_neighbor_mw |
Conjoint table (irw_conjoint). Single text vignette per respondent (phone survey) with five randomized fill-ins; three 0/1 outcomes (would help the neighbour to hospital; should be allowed to move freely vs stay at home; thinks the neighbour has Covid-19), the authors’ recodes with don’t-know/refusal (and ‘not sure what Covid is’) as NA. No opt-out. Respondents with no outcome omitted. Malawi rounds 1 and 2 pooled (trial_round), ids not linked across rounds. | 2026-10-09 | |
ferree_2023_covid_neighbor_zm |
Conjoint table (irw_conjoint). Single text vignette per respondent (phone survey) with five randomized fill-ins; three 0/1 outcomes (would help the neighbour to hospital; should be allowed to move freely vs stay at home; thinks the neighbour has Covid-19), the authors’ recodes with don’t-know/refusal (and ‘not sure what Covid is’) as NA. No opt-out. Respondents with no outcome omitted. | 2026-10-09 | |
ferrer_2025_liberal_democracy |
Conjoint table (irw_conjoint_2). Five tasks of two hypothetical countries (6 binary institutional attributes plus relative citizen income) in seven countries; choice of one country and a 0-10 rating of each (rating higher = liked more, 567 NA); no opt-out. Levels are the UK English text for the translations respondents saw in their own language (authors’ recode reproduced); income stored as relative level with the displayed amount in trial_income_shown; no covariates or weights in the deposit. | 2026-10-11 | |
ferwerda_2023_guideline_violators |
Conjoint table (irw_conjoint). choice = the individual judged more likely to violate public-health guidelines (“quale dei due individui sarebbe più probabile di violare le linee guida di sanità pubblica?”; forced, no opt-out; being chosen is unfavourable, kept as asked). Italian level text as displayed; attribute order randomized per respondent (attrpos_). Questions Q15.3 and Q15.5 displayed randomizer sets F-3 and F-2; task = question order with the profiles actually shown (trial_profile_set), whereas the authors’ code pairs them the other way round. A rare birthplace level ‘Africa’ (83 profiles) is kept (the authors drop it). All 2,734 respondents with an answer kept; cov_in_authors_sample marks the authors’ two-wave sample (1,237). | 2026-10-07 | |
ferwerda_2023_vaccine_priority |
Conjoint table (irw_conjoint). choice = the individual who should get vaccine priority (“quale dei seguenti cittadini italiani dovrebbe avere la priorità secondo lei?”; forced, no opt-out). Italian level text as displayed (accents stripped in the export restored for ‘Più di 80’ and ‘Povertà’); attribute order randomized per respondent (attrpos_). All 1,268 respondents kept; cov_in_authors_sample marks the authors’ two-wave sample (1,236). | 2026-10-07 | |
fesenfeld_2020_food_policy_cn |
Conjoint table (irw_conjoint). choice: “Which policy package do you prefer?” (forced: “please choose the one you oppose less”; no opt-out); rating: “How much do you personally support policy package A/B?” 1 Strongly oppose .. 7 Strongly support. Revenue use is ‘No tax revenues’ exactly when there is no new tax; attribute order random per respondent (tax before revenue use), not recorded. trial_info_treatment holds the respondent’s earlier information/framing arm. Level text is the English instrument; respondents saw Chinese. | 2026-10-08 | |
fesenfeld_2020_food_policy_de |
Conjoint table (irw_conjoint). choice: “Which policy package do you prefer?” (forced: “please choose the one you oppose less”; no opt-out); rating: “How much do you personally support policy package A/B?” 1 Strongly oppose .. 7 Strongly support. Revenue use is ‘No tax revenues’ exactly when there is no new tax; attribute order random per respondent (tax before revenue use), not recorded. trial_info_treatment holds the respondent’s earlier information/framing arm. Level text is the English instrument; respondents saw German. | 2026-10-08 | |
fesenfeld_2020_food_policy_us |
Conjoint table (irw_conjoint). choice: “Which policy package do you prefer?” (forced: “please choose the one you oppose less”; no opt-out); rating: “How much do you personally support policy package A/B?” 1 Strongly oppose .. 7 Strongly support. Revenue use is ‘No tax revenues’ exactly when there is no new tax; attribute order random per respondent (tax before revenue use), not recorded. trial_info_treatment holds the respondent’s earlier information/framing arm. | 2026-10-08 | |
fesenfeld_2023_climate_legit_de |
Conjoint table (irw_conjoint). choice: forced choice between two climate-policy packages (German wording not deposited; the US instrument asks ‘Which policy package do you prefer?’), no opt-out; rating: 1-7 per package, wording and anchors not deposited, stored raw. 4 rounds x 2 packages, 6 attributes; ‘No restrictions’ appears twice as often as each specific restriction. No weight. | 2026-10-08 | |
fesenfeld_2023_climate_legit_us |
Conjoint table (irw_conjoint). choice: forced choice between two climate-policy packages (‘Which policy package do you prefer?’, choose the one you oppose less if you support neither), no opt-out; rating: 1-7 per package, wording and anchors not deposited, stored raw. 4 rounds x 2 packages, 6 attributes; ‘No restrictions’ appears twice as often as each specific restriction. 93 incomplete single-package tasks dropped (1,350 of 1,360 respondents kept). No weight. | 2026-10-08 | |
fhi360_2020_mnp_india |
Conjoint table (irw_conjoint). choice: “Which of these methods do you prefer?” (English master questionnaire; asked in Hindi), forced choice between two patch options, no None option. Level text is the Hindi text shown (each level also had an illustration), decoded from the Sawtooth design and study files. The fixed holdout task and unanswered tasks are dropped; 31 respondents without a design version dropped. 496 respondents. | 2026-10-08 | |
fhi360_2020_mnp_nigeria_s1 |
Conjoint table (irw_conjoint). choice: “Which of these methods do you prefer?” (English master questionnaire; asked in Yoruba), forced choice between two patch options, no None option. Level text is the Yoruba text shown (each level also had an illustration), decoded from the Sawtooth design and study files. The fixed holdout task and unanswered tasks are dropped; 13 respondents without a design version dropped. 530 respondents. | 2026-10-08 | |
fhi360_2020_mnp_nigeria_s2 |
Conjoint table (irw_conjoint). choice: “Which of these methods do you prefer?” (English master questionnaire; asked in Yoruba), forced choice between two patch options, no None option. Level text is the Yoruba text shown (each level also had an illustration), decoded from the Sawtooth design and study files. The fixed holdout task and unanswered tasks are dropped; no menstruation attribute in this design. 416 respondents. | 2026-10-08 | |
findrik_2026_bioplastic_packaging |
Conjoint table (irw_conjoint). Discrete choice experiment: ‘Welches dieser Produkte würden Sie kaufen?’ (which of these products would you buy) among 3 cherry-tomato packages; ‘Keines dieser 3 Produkte’ (none) is an opt-out, coded as choice = 0 on all three cards (646 of 10,428 sets). Fixed design of 12 cards in 12 sets shown to everyone in the same order; respondents randomly got an information text about bioplastics or not (trial_info). | 2026-10-09 | |
finseraas_2026_group_appeals_britain |
Outcomes: forced vote choice between two candidates; rating_represent, 1-7 extent the candidate represents “someone like you” (higher = more); rating_leftright, 1-7 placement (higher = further right, not favourability). The deposit collapses the six displayed job titles to working vs middle class, so attr_occupation holds the class with its possible titles. The authors’ dummy and interaction variables are dropped. | #2887 | 2026-10-07 |
finseraas_2026_group_appeals_norway |
Same outcomes as the British table (forced vote choice; 1-7 representation; 1-7 left-right placement, higher = further right); attribute text is the authors’ English translation of the Norwegian. Occupation is collapsed to class in the deposit. 228 tasks have ratings but no choice (choice empty) and 19 respondents with no answers are dropped; the paper reports 2,931 analysed versus 3,171 here (unreconciled). | #2887 | 2026-10-07 |
finseraas_2026_muslim_candidates |
Conjoint table (irw_conjoint). Outcome: forced choice between two candidates (question wording not deposited; the authors describe willingness to vote for them), no opt-out. Pools the first-stage baseline sample (trial_stage 1) with the second-stage sample randomized to five arms that add party, general-policy, minority-rights or personal attributes (trial_arm; attributes not shown are (not shown)). Task 6 repeats task 1 with profiles swapped. Levels are the deposit’s English labels (display presumably Norwegian). | 2026-10-07 | |
fischer_2018_promis29_physfunc |
The deposit’s value labels list the options in questionnaire order (1 = without any difficulty), but the stored codes are PROMIS item scores: 5 = without any difficulty (77-83% of this general-population sample; the domain correlates -0.58 with pain interference). | 2026-10-09 | |
fischer_2018_promis29_sleep |
Stored codes are PROMIS item scores: P5 (sleep quality) and P6_1 (sleep was refreshing) are stored reversed, so higher means more sleep disturbance on all four items, although the deposit’s value labels list the options in questionnaire order. | 2026-10-09 | |
fischer_2024_error_handling |
Conjoint table (irw_conjoint). Paired vignette options (A/B) after a fixed scenario; choice = option chosen, forced, no opt-out (question wording not deposited). 4 of 18 fixed choice sets per respondent (random subset, display order not recorded: task = ascending set number). Protagonist Frau/Herr Müller randomized between subjects (attr_protagonist). Level text German original. The deposit’s two code files disagree on which error-type code is skill- vs rule-based (table follows the .do). N 276 / 1,104 obs matches the article; 3 respondents with only 3 sets dropped, as by the authors. | 2026-10-09 | |
flores_2025_gasoline_brands |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each driver rated 12 fixed gas-station visits (brand, waiting time 5 or 20 minutes, prices of 87- and 91-octane gasoline) on a 1-10 scale, stored raw (the deposit says only ‘rate given to a hypothetical visit’; direction undocumented, though ratings fall with waiting time). Three fixed questionnaire versions (trial_version); that both prices were displayed is inferred from the deposit description; 49 respondents rated only some cards and unrated cards are omitted. | 2026-10-12 | |
floresmacias_2022_militarization |
Conjoint table (irw_conjoint). One pair of photographs per respondent (task = 1); attributes are shown in the photo and stored as the authors’ labels. choice is the forced choice of who is more effective (no opt-out); rating_effective, rating_civil_liberties, rating_corrupt and rating_neighborhood are on the authors’ 0-1 rescaling of a 10-point item, wording and anchors undocumented, and the direction of rating_corrupt is not documented. trial_prompt is the security framing shown before the photos. Municipality identifiers and the authors’ dummies are dropped. | 2026-10-07 | |
folke_2022_harassment_job_choice |
Conjoint table (irw_conjoint). Three choices between Job A and Job B; choice = ‘Based on the information in the table, which of the two jobs would you prefer?’ (authors’ translation from Swedish; no opt-out). Work environment levels non-uniform (each harassment vignette 11%, other values 22%); the vignette set (victim sex) follows the respondent’s workplace sex composition; row order randomized but not saved. 11,722 answered tasks from 3,925 respondents (paper N = 3,987); dominant-job pairs, which the paper drops, are kept. | 2026-10-09 | |
fong_2023_hads |
Responses are stored 1-4 as administered, not on the scale’s usual 0-3 scoring. | 2026-10-04 | |
fong_2023_sf12 |
Items keep their native formats: item 1 and item 8 are 1-5, items 2-3 are 1-3, items 4-7 are yes/no coded 0/1, and items 9-12 are 1-6. Response direction is not harmonised across items. | 2026-10-04 | |
fong_2023_ssqol12 |
Items are scored 1 = agree completely to 5 = disagree completely, so higher values mean less interference from stroke (better quality of life). The follow-up (wave 2) has 148 of the 184 baseline respondents. | 2026-10-04 | |
forster_2024_hiring_vignettes |
Conjoint table (irw_conjoint). Factorial survey of application mock-ups, one applicant per task: rating = likelihood of inviting the candidate to the next step, 0-100 percent in steps of 10; choice = final yes/no invitation marked for each of the eight applicants on a summary page (several or none could be invited, opt-out yes). Fixed blocked D-efficient design; field of study shown only for the some-college level (‘(not shown)’ otherwise). Level text is the authors’ English category labels, not the German materials. 480 recruiters x 8 = 3,840 vignettes, as in the article. | 2026-10-09 | |
fox_2023_bct |
Study 2 reworded BCT_14_Climate (‘financial interests’) and BCT_3_Sim; both studies share item codes and are pooled under cov_study. | wording | 2026-10-10 |
fox_2023_pss10 |
Study 2’s cov_age includes a few implausible values (minimum 4) kept as deposited. | 2026-10-10 | |
franchino_2015_candidate_valence |
Conjoint table (irw_conjoint). choice = ‘PER CHI VOTEREBBE?’ (who would you vote for), Candidato A or B, forced, no opt-out. Fixed set of 27 pairs (54 profiles) shown in random order, not independently randomized, so attribute levels are correlated across the design; task is the display position and trial_contest the pair number. Italian displayed text, opinion quotes without quotation marks; 23 unanswered pairs omitted. | 2026-10-07 | |
franchino_2017_economic_policy |
Conjoint table (irw_conjoint_2). choice = ‘Between these two programs, which one do you prefer?’ (forced, no opt-out, 2 pairs); rating = for/against each program, 1 molto contrario (strongly against) to 4 molto favorevole (strongly for). Task and profile are inferred from row pairs; level text completed from the appendix; sample weight kept; N against the article not checked. | 2026-10-11 | |
franchino_2017_fiscal_union |
Conjoint table (irw_conjoint_2). choice = ‘Between these two schemes, which one do you prefer?’ (forced, no opt-out, 2 pairs); rating = for/against each scheme, 1 molto contrario (strongly against) to 4 molto favorevole (strongly for). Task and profile are inferred from row pairs; sample weight kept; N against the article not checked. | 2026-10-11 | |
frankowski_2026_housing_invest_cz |
Conjoint table (irw_conjoint). Discrete choice experiment, 5 tasks of OPTION A vs OPTION B (‘Please check the option that suits you best’), forced choice, no opt-out; rating_confidence = ‘To what extent are you confident with your choice?’ 0-100 (task level). Czech labels; cost and value attributes stored as the percentage behind the amount shown. Only the Czech online sample (N = 1,531) is built; the Polish online and both face-to-face samples of the deposit are held for data defects. | 2026-10-09 | |
frederiksen_2025_undemocratic_dominance |
Conjoint table (irw_conjoint). Candidate ratings, no choice: each respondent was randomly assigned one outcome - rating_dominance (‘aggressive, threatening, and intimidating’), rating_warmth (‘warm, reliable, and friendly’), rating_competence (‘competent, knowledgeable, and skillful’), each 7-point Strongly disagree to Strongly agree, or rating_vote (how likely to vote for the candidate) - stored rescaled 0-1 as deposited; Don’t know omitted. Second Study 1 experiment only; non-uniform weights for age bands, gender and background by design. | 2026-10-09 | |
freihardt_2024_flood_de |
Conjoint table (irw_conjoint_2). rating_allocation = millions of a $10 million grant given to this community (EUR in the German sample; 0-10, the two communities sum to 10, higher = more funding); rating_support = 1-5 support for funding it (5 = strongly support); no choice, no opt-out. Task 6 repeats task 1 with A and B swapped (trial_repeat_of = 1); attribute order randomized per respondent (attrpos_). IP and lat/long dropped; 5 respondents with US wording dropped, 13-17 tasks with missing answers dropped. | 2026-10-11 | |
freihardt_2024_flood_us |
Conjoint table (irw_conjoint_2). rating_allocation = millions of a $10 million grant given to this community (0-10, the two communities sum to 10, higher = more funding); rating_support = 1-5 support for funding it (5 = strongly support); no choice, no opt-out. Task 6 repeats task 1 with A and B swapped (trial_repeat_of = 1); attribute order randomized per respondent (attrpos_). IP and lat/long dropped; 13-17 tasks with missing answers dropped. | 2026-10-11 | |
freire_2021_climate_treaty |
Conjoint table (irw_conjoint). Forced choice of the treaty “that has the characteristics that you believe are best for your country” (no opt-out), up to 7 pairs per respondent. Ten countries pooled with cov_country and cov_elite_group (blank for 778 rows as in the source). Level text is the authors’ English labels of a Spanish/Portuguese instrument. Randomization uniform with one restriction: Punishment = None never appears with repeated violations = Less penalty. Duplicate task records for 3 respondents collapsed; Qualtrics GPS coordinates and the location-derived country dropped. | 2026-10-07 | |
freire_2023_lynching |
Conjoint table (irw_conjoint_2). choice = ‘Em qual dos dois casos você acha que o linchamento é mais justificado?’ (In which of the two cases do you think the lynching is more justified?), forced choice between Caso 1 and Caso 2, no opt-out, 5 pairs. Restrictions: female perpetrators never paired with rape; car theft never with a child or teenage victim. 2,084 of the 2,460 respondents in the export answered the conjoint (the appendix reports 2,406 consenting respondents). Latitude/longitude and free-text answers dropped. | 2026-10-09 | |
frost_2023_cbq |
Six cells coded 0 on the 1-3 scale were dropped as off-scale. | 2026-10-10 | |
fu_2024_war_casualties |
Conjoint table (irw_conjoint_2). One hypothetical war scenario per respondent (Taiwan, online): whose casualties (US, Taiwan, PLA, civilians) x casualty size (1,000 / 10,000 / 100,000 deaths); rating = war-support score 1-5 stored raw (higher = more support, inferred from the authors’ recode). Attribute level text is the short labels from the authors’ Stata code, not the displayed vignette text (the deposit holds no questionnaire); the question wording and anchors are unknown. | 2026-10-14 | |
fu_2026_victory_abstract |
Conjoint table (irw_conjoint). choice = the scenario chosen (perceived chance of Taiwan’s victory; exact wording and display language not in the deposit), forced, exactly one per task. Levels are the authors’ short English labels. Randomization rules and attribute order not documented. 1,444 respondents as in the authors’ run.log. | 2026-10-07 | |
fu_2026_victory_concrete |
Conjoint table (irw_conjoint). choice = the scenario chosen (perceived chance of Taiwan’s victory; exact wording and display language not in the deposit), forced, exactly one per task. Levels are the authors’ short English labels. Randomization rules and attribute order not documented. 1,011 respondents as in the authors’ run.log. | 2026-10-07 | |
fullscaleiq_memory_options |
Each question showed 8 options, 3 to 5 of them correct, and asked respondents to tick only those they were sure of. Each row is one option: resp = 1 if ticked, 0 if not; a question with nothing ticked is all zeros. Option numbering: k is the icon’s position in that question’s option list in demo/subtest-ISM.js (seq_key); the source records icon ids. Key (correct options; the rest are incorrect): MQ1 opt1-opt4; MQ2 opt1-opt3; MQ3 opt1-opt4; MQ4 opt1-opt5; MQ5 opt1-opt5; MQ6 opt1-opt5. The key was derived from the stored scores and reproduces all of them. The order of clicks is not kept. The hits-minus-false-alarms score is in fullscaleiq_memory, with the same ids. | #2660 | 2026-10-01 |
fullscaleiq_mentalrotation_options |
Each question showed 8 options, 3 to 5 of them correct, and asked respondents to tick only those they were sure of. Each row is one option: resp = 1 if ticked, 0 if not; a question with nothing ticked is all zeros. Option numbering: k is the source’s answer code minus 10 x the question number (code 23 = RQ2_opt3), the shape image key in demo/shapes.json. Key (correct options; the rest are incorrect): RQ1 opt1-opt3; RQ2 opt1-opt5; RQ3 opt1-opt4; RQ4 opt1-opt5; RQ5 opt1-opt4; RQ6 opt1-opt5. The key was derived from the stored scores and reproduces all of them. The order of clicks is not kept. The hits-minus-false-alarms score is in fullscaleiq_mentalrotation, with the same ids. | #2660 | 2026-10-01 |
fullscaleiq_vocab_options |
Each question showed 8 options, 3 to 5 of them correct, and asked respondents to tick only those they were sure of. Each row is one option: resp = 1 if ticked, 0 if not; a question with nothing ticked is all zeros. Option numbering: k is the source’s answer code, which indexes the word list for that question in the test’s demo/subtest-V.js (e.g. VQ1 ‘big’: opt1 large … opt8 crumbled). Key (correct options; the rest are incorrect): VQ1 opt1-opt4; VQ2 opt1-opt3; VQ3 opt1-opt4; VQ4 opt1-opt4; VQ5 opt1-opt4; VQ6 opt1-opt5; VQ7 opt1-opt5. The key was derived from the stored scores and reproduces all of them. The order of clicks is not kept. The hits-minus-false-alarms score is in fullscaleiq_vocab, with the same ids. | #2660 | 2026-10-01 |
furnas_2023_staff_access |
Conjoint table (irw_conjoint). Single-profile factorial vignette (requester identity x action x evidence, 4x2x4, slot text as stored); rating = take the meeting, rating_use_info, rating_recommend, rating_representative, each 1 Very unlikely .. 5 Very likely coded from answer text; question wording paraphrased (not deposited). Survey weight kept; 437 staffers (article: 436). | 2026-10-09 | |
gagnon_2025_local_candidates |
Conjoint table (irw_conjoint). choice = Q26 etc. vote (forced choice, Candidate A/B); choice_benefit_me / _prefecture / _japan = which candidate would bring more benefits to you personally / your prefecture / Japan (forced). Level text is Japanese as stored; piped fields not saved in the export are kept as placeholders ([respondent_prefecture], [outside_prefecture], [university_local], [university_outside], [todai_kyodai]), so text reads as design categories, not the names respondents saw. Restrictions undocumented; shares near uniform. 44 unfinished respondents kept (cov_finished = 0). | 2026-10-07 | |
gagnon_2025_local_status |
Conjoint table (irw_conjoint). choice = vote (forced, Candidate A/B); choice_ |
2026-10-07 | |
gaikwad_2025_climate_country_in |
Conjoint table (irw_conjoint_2). Paired conjoint, 4 tasks (numbered 5-8) x 2 profiles, 7 donor-country attributes; choice is the preferred country (no opt-out), rating a 1-10 rating of each country. Question wording and rating anchors are not in the deposit. | 2026-10-11 | |
gaikwad_2025_climate_country_us |
Conjoint table (irw_conjoint_2). Paired conjoint, 4 tasks (numbered 5-8) x 2 profiles, 7 country attributes incl. regime, climate risk, income level and race (white / not white); choice is the preferred country (no opt-out), rating a 1-10 rating of each country (19 missing). Question wording and rating anchors are not in the deposit. | 2026-10-11 | |
gaikwad_2025_climate_policy_in |
Conjoint table (irw_conjoint_2). Paired conjoint, 4 tasks x 2 profiles, 7 attributes (duration, goal, compensation, partners, reciprocity, monitoring, conditionalities); choice is the preferred policy (no opt-out), rating a 1-10 rating of each policy. Question wording and rating anchors are not in the deposit. | 2026-10-11 | |
gaikwad_2025_climate_policy_us |
Conjoint table (irw_conjoint_2). Paired conjoint, 4 tasks x 2 profiles, 7 attributes (target, cost, duration, goal, compensation, partners, reciprocity); choice is the preferred policy (no opt-out), rating a 1-10 rating of each policy. Question wording and rating anchors are not in the deposit. Article reports 2,006 US respondents; the file holds 2,019. | 2026-10-11 | |
galan_cuevas_2021_kidscreen52 |
The negatively worded KIDSCREEN items (EMO1-7, SEL3-5, BUL1-3) are stored reverse-scored by the source, so higher = better quality of life on every item. IRW keeps the source’s coding. | 2026-10-07 | |
gampfer_2014_climate_funding_de |
Conjoint table (irw_conjoint_2). Forced choice between two climate-funding proposals (Proposal 1/2; no opt-out), 6 tasks, 7 attributes with levels from the Stata value labels (German); question wording is not in the deposit. One funding-use level is cut at 80 characters in the source and completed by inference. YouGov weight kept as cov_survey_weight. N = 1,500 respondents, as in the codebook; the German sample is a separate table from the US one. | 2026-10-11 | |
gampfer_2014_climate_funding_us |
Conjoint table (irw_conjoint_2). Forced choice between two climate-funding proposals (Proposal 1/2; no opt-out), 6 tasks, 7 attributes with levels from the Stata value labels (English); question wording is not in the deposit. One funding-use level is cut at 80 characters in the source and completed by inference. YouGov weight kept as cov_survey_weight. N = 1,500 respondents, as in the codebook; the US sample is a separate table from the German one. | 2026-10-11 | |
gan_2026_alternative_meat |
Conjoint table (irw_conjoint). Discrete choice experiment: choice = which burger (Option A or B), forced, no opt-out; question wording not deposited. Fixed Bayesian-efficient design of 20 sets (plant-based and cell-cultured are one Type attribute); supermarket or restaurant framing randomized between respondents (trial_setting), with restaurant prices £10 higher. Level text taken from the choice-card images. | 2026-10-09 | |
gao_2024_gad7 |
Item codes are the questionnaire’s running numbers (Q61-Q67; PHQ-9 is Q68-Q76, APSQ Q51-Q60, ABQ Q77-Q91), located by the deposit’s own scale scores. GAD-7 and PHQ-9 are stored 1-4 (the file’s score columns subtract 1 to give the usual 0-3); shipped as stored. | 2026-10-07 | |
garbe_2025_digital_id |
Conjoint table (irw_conjoint). choice: ‘Which of the following two policy proposals for a new Digital ID would you prefer?’ (forced; a refusal was re-asked as which they dislike least; 17 tasks still refused, choice NA). rating_support and rating_register: 1-4 (Not at all - Very supportive / likely), per policy; five round-1-only agreement ratings 1-4 (Strongly disagree - Strongly agree; higher = more worried for the three worry items). DK/refusals NA. 3 rounds x 2 policies, 5 binary attributes as showcard text; identical pairs excluded; attribute order randomized per respondent (attrpos_). Kept only interviews the survey firm approved (as the authors). | 2026-10-09 | |
gatto_2025_candidate_selection |
Conjoint table (irw_conjoint). choice = “Based on the information provided, please select the aspirant that you would prefer to place on a prominent placement in your party’s candidate list” (article’s English rendering; Kandidat A/B, forced, no opt-out). Fully randomized profiles; the share of women already on the list (10-50%) is randomized per task and stored as trial_list_share. Ideology levels are relative (‘Etwas linker/rechter/moderater als Sie’) with wording that depends on the respondent’s left/centre/right group. 1,856 respondents incl. non-selectors; cov_in_authors_sample = the article’s 1,389 (1,343 with at least one choice). | 2026-10-07 | |
gazmararian_2024_net_metering |
Conjoint table (irw_conjoint). Four paired solar-donation-program tasks with six attributes; choice = ‘Which program do you prefer?’ (Program A/B, forced, no opt-out), rating = ‘What percent (from 0 to 100) of your yearly solar profits would you choose to donate under each program?’ (free-typed 0-100, stored as typed). Qualtrics IDs re-keyed; zip/county/panel ids dropped; N = 2,006, no rows dropped; a stray fifth round present for 21 respondents is ignored. | 2026-10-09 | |
gazmararian_2024_transition_fair |
Conjoint table (irw_conjoint). choice: ‘If you had to choose, which proposal would you prefer the government to pursue?’ Proposal A/B, forced (no opt-out). Randomized introduction (coal-decline information or not) in trial_intro; cov_survey_weight is the trimmed raking weight that reproduces SI Table 8.1 (the deposited code uses the untrimmed weights2, kept as cov_weight_untrimmed); one unanswered task dropped; 248 respondents as in the SI. | 2026-10-09 | |
gazmararian_2024_transition_national |
Conjoint table (irw_conjoint). choice: ‘If you had to choose, which proposal would you prefer the government to pursue?’ Proposal A/B, forced (no opt-out). Randomized introduction (coal-decline information or not) in trial_intro; authors’ raking weight in cov_survey_weight (trimmed version in cov_weight_trimmed); ZIP codes and panel ids dropped; 1,001 respondents as in SI Table 4.1. | 2026-10-09 | |
gengler_2021_survey_participation |
Conjoint table (irw_conjoint). Factorial vignette, three per respondent, one profile each: ‘Imagine that the survey was being conducted by {SPONSOR} and would take place {MODE}. It would last about {TIMING} and cover important {TOPIC} issues facing the country. How likely would you be to participate in such a survey?’ rating 1 = Very likely, 2 = Somewhat likely, 3 = Somewhat unlikely, 4 = Very unlikely (stored as answered; lower = more willing). Timing levels depend on mode (restriction). Don’t know/refused omitted; survey weight kept. | 2026-10-09 | |
gest_2025_climate_migration |
Conjoint table (irw_conjoint). Forced choice of the scenario in which the respondent would be more likely to migrate to the United States (no opt-out), then a 1-7 rating of each scenario (1 highly unlikely .. 7 highly likely, raw); scenarios were read aloud by interviewers. Attribute row order shuffled per task (attrpos_). trial_prime marks an experimental arm whose content is not documented in the deposit. One preview record and one respondent without choices dropped. | 2026-10-08 | |
getmansky_2024_refugees |
Conjoint table (irw_conjoint). Three outcome blocks over the same pairs, in randomized order (trial_outcome_order): rating 1 = definitely don’t want to 7 = definitely want the refugee as neighbor / given a work permit / granted citizenship, then a forced choice. About 2.6% of choices contradict the ratings and are kept as answered. Level text from the authors’ pre-analysis plan (respondents saw Turkish); education, employment and income covariates dropped because their codes are undocumented. | 2026-10-07 | |
ghasemy_2021_poly_conceptual_skills |
Item numbering gaps are the deposit’s (items not retained); the file is the authors’ outlier-removed sample. The poly_ infix separates these from ghasemy_2021_* tables built from another Ghasemy deposit. | 2026-10-09 | |
ghasemy_2021_poly_creating_value |
Item numbering gaps are the deposit’s (items not retained); the file is the authors’ outlier-removed sample. The poly_ infix separates these from ghasemy_2021_* tables built from another Ghasemy deposit. | 2026-10-09 | |
ghasemy_2021_poly_job_satisfaction |
Item numbering gaps are the deposit’s (items not retained); the file is the authors’ outlier-removed sample. The poly_ infix separates these from ghasemy_2021_* tables built from another Ghasemy deposit. | 2026-10-09 | |
ghasemy_2021_poly_work_motivation |
Item numbering gaps are the deposit’s (items not retained); the file is the authors’ outlier-removed sample. The poly_ infix separates these from ghasemy_2021_* tables built from another Ghasemy deposit. | 2026-10-09 | |
gholami_2017_periodontal_knowledge |
cov_age is an age-band code, not age in years: 1 = “18-24”; 2 = “25-34”; 3 = “35-44”; 4 = “45>”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
ghosn_2021_refugee_return |
Conjoint table (irw_conjoint). choice: ‘Now tell me, which place in Syria would you choose to return to?’ (Place A / Place B; a Don’t know / No response option existed but those tasks are not in the deposit, so every stored task has one chosen place). Up to 5 tasks x 2 places, 4 attributes with levels from the questionnaire’s numbered options; profile order inferred from row order. 417 respondents in the file vs 406 reported in the article. | 2026-10-09 | |
giannoni_2025_firm_jobs |
Conjoint table (irw_conjoint_2). Two tasks of two hypothetical workers’ firms (7 binary attributes, order randomized per task); each of 11 political statements (anti-system attitudes and behaviours) is a forced choice of which worker (1 or 2) holds that attitude (choice_* columns, no opt-out). Level text is Portuguese as displayed; 389 file rows with no outcome omitted (1,374 respondents; paper reports 1,373); question stem paraphrased from the article. | 2026-10-11 | |
gidengil_2022_executive_aggrandizement |
Conjoint table (irw_conjoint). choice = preferred candidate of the pair (forced, no opt-out; wording paraphrased from the article). attr_checks holds the Parliament or courts version of the executive-aggrandizement position, randomized by respondent (trial_checks_version). Order of the last three attribute rows randomized per respondent (attrpos_). Article exclusions (attention check, 300-3,600 s, straightliners, non-citizens, tasks < 5 s) not applied; flags kept as cov_attention_pass, cov_duration_sec, cov_canadian_citizen, trial_task_time_sec. | 2026-10-09 | |
gift_2025_speech_scandals |
Conjoint table (irw_conjoint). Single-candidate text vignettes, 3 per respondent: choice = ‘would you ever consider voting for this candidate?’ 1 yes / 0 no (opt-out: rejecting is the outside option); rating_thermometer 0-100 (don’t know = NA); rating_objectionable 1 not at all - 7 most objectionable imaginable. Levels are the authors’ category labels; names inferred from Appendix Table A.4; remark attributes (not shown) in controls. YouGov weight kept; free text dropped. | 2026-10-09 | |
gilbert_meta_10 |
Some students took the reading self-concept (MMRP) survey twice; the original study team averaged their two records, giving 301 half-point responses (1.5 or 2.5) from 195 students in the source file. IRW’s build rounded these down (1.5 to 1, 2.5 to 2), the most conservative reading of the student’s standing, so up to 301 resp values (about 0.3%) are a floored average rather than a single answer. | #2765 | 2026-10-02 |
gilhernandez_2023_teacher_bias |
Conjoint table (irw_conjoint_2). Factorial survey: each of 1,717 Spanish pre-service teachers rated one student vignette from a full 2^7 design (student gender, ethnic origin, father’s SES, essay quality, essay cultural-capital sentence, subjects failed, behaviour). Ratings: essay grade (1-10, in rating), grade-retention recommendation, educational expectations and parental support (0-10); no opt-out. Attribute level text is the codebook’s English labels (paraphrased), not the Spanish wording shown; anchors of the 0-10 items are not documented. | 2026-10-14 | |
gilhernandez_2024_teacher_bias |
Conjoint table (irw_conjoint). Factorial vignette (full 2^7, one per respondent): rating = essay grade 1-10 (‘What grade from 1 to 10 (including decimal points) would you give to the essay…’), rating_retention 0-10 (0 = should never repeat 6th grade, 10 = should definitely repeat), rating_expectations 0-10 (likelihood of reaching the upper-secondary academic track), rating_parental_support 0-10 (paraphrase); decimals allowed. Level text mixes displayed Spanish stimulus (names, emails, essay sentence) and the article’s English descriptions. N = 1,717 as in the article. | 2026-10-09 | |
gimenez_espert_2026_aco |
Items were assigned to instruments by response format: IT1-IT12 are the only 1-5 block apart from the six violence items, matching the 12-item ACO-R. The affective items are negatively worded and are left as administered. | #2997 | 2026-10-14 |
gimenez_espert_2026_unipsico |
The deposit labels items IT1-IT74 with no scale assignment. The 53 items on the 0-4 format are eight UNIPSICO scales plus the 7-item reduced CESQT, but their order in the file is not recoverable (the paper order does not reproduce its reported alphas), so they ship as one table with the deposit item codes. Items IT21, IT22 and IT64 never use the full 0-4 range. | #2997 | 2026-10-14 |
glenk_2024_farm_support |
Conjoint table (irw_conjoint). Factorial survey: six single text vignettes per respondent from a 72-row orthogonal fold-over design, farm type randomized between respondents (trial_farm_type); four 1-11 ratings stored raw (acceptability of the payment change, fairness of the payment 1 unfairly low to 11 unfairly high, unreversed, happiness to be supplied (NA for ‘do not buy or eat’), willingness to lobby the MSP). No opt-out. Payment amount text is fixed by farm type, size and change. Task order inferred (display order not recorded). N 2,011 vs article 1,951 (missing covariates). | 2026-10-09 | |
goehring_2024_unilateral_policy |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (wave 1 of a two-wave study): President Obama or Trump, one of 15 policy topics (stored tags only; the full topic text is not in the deposit), and whether the president only stated a position, worked with Congress on a bill or signed an executive order (text from the article’s example, position-only level paraphrased). Two 1-7 approval ratings (handling of the topic, overall presidential approval; higher = more approval); no opt-out. Wave-2 follow-up vignette not built. | 2026-10-14 | |
goggin_2020_ideology_guess |
Conjoint table (irw_conjoint). One candidate per task, no choice: rating_conservative = guess of the candidate’s ideology (1 Conservative, 0 Liberal), rating_sure = sureness (1-4 from answer text), rating_favorability = 0-10 (wording not in the deposit). Same respondents as the JAH module of goggin_2020_party_guess; 177 of 1,000 have no ideology profiles saved and are dropped (823 kept). Task order inferred from row order. | 2026-10-08 | |
goggin_2020_party_guess |
Conjoint table (irw_conjoint). One candidate per task, no choice: rating_republican = guess of the candidate’s party (1 Republican, 0 Democrat), rating_sure = sureness (1 very unsure to 4 very sure, coded from the answer text), rating_favorability = 0 very unfavorable to 10 very favorable (wording not in the deposit). The three issue priorities are drawn without replacement. Two CCES team modules pooled (cov_cces_module); 2,000 respondents. | 2026-10-08 | |
goldberg_2023_citizen_forums |
Conjoint table (irw_conjoint). choice = ‘Which of the two scenarios do you prefer?’ (Citizen Forum A/B; translated), forced, no opt-out; rating = ‘In your view, how do you feel about Citizen Forum A/B?’ 1 = I don’t like it at all to 7 = I like it very much (translated). Attributes fully randomized; attribute order randomized per respondent with the issue always first (not recorded). Levels are the authors’ English labels of German text; no survey weight deposited although the article reports a weighted sample. | 2026-10-08 | |
goldberg_2025_minipublics_finland |
Conjoint table (irw_conjoint). choice: “which of the two scenarios do you prefer?” (forced, no opt-out). No rating in Finland (not collected). attr_output’s code-to-label mapping is inferred, not documented (indirect check in the script header). Attribute order randomized but not recorded. Trust and satisfaction covariates keep the source’s don’t-know code (997; 8 in Finland). | 2026-10-07 | |
goldberg_2025_minipublics_ireland |
Conjoint table (irw_conjoint). choice: “which of the two scenarios do you prefer?” (forced, no opt-out). rating: support for each forum, 1 = strongly oppose to 7 = strongly support. attr_output’s code-to-label mapping is inferred, not documented (indirect check in the script header). Attribute order randomized but not recorded. Trust and satisfaction covariates keep the source’s don’t-know code (997; 8 in Finland). | 2026-10-07 | |
goldberg_2025_minipublics_usa |
Conjoint table (irw_conjoint). choice: “which of the two scenarios do you prefer?” (forced, no opt-out). rating: support for each forum, 1 = strongly oppose to 7 = strongly support. attr_output’s code-to-label mapping is inferred, not documented (indirect check in the script header). Attribute order randomized but not recorded. Trust and satisfaction covariates keep the source’s don’t-know code (997; 8 in Finland). | 2026-10-07 | |
goldfarb_2026_aviation_fuel |
Conjoint table (irw_conjoint). Outcome: “Which of these two flights would you choose?” (Flight 1 / Flight 2), forced, no opt-out. Attribute order randomized per respondent and recorded (attrpos_). cov_survey_weight is the deposit’s weight. The fuel-type level text keeps the source’s spelling ‘Sutainable’. ZIP codes and Qualtrics IDs dropped, as were the omnibus survey’s unrelated items. | 2026-10-08 | |
gomm_2026_sustainable_products |
Conjoint table (irw_conjoint_2). choice = which of two versions of a product the respondent would buy (paraphrase; forced choice, no opt-out, 6 tasks); rating = 1-7 likelihood of buying each version (7 = very likely). Only price and environmental impact vary within a task; the product (6 types, assigned within a category arm) is trial_product. The authors’ 0/1 rating dummy and all scales dropped; 298 panel members without conjoint rows absent. | 2026-10-11 | |
gori_2020_cope |
The source deposit is the authors’ analysis file, in which missing item responses had been filled with the item mean. IRW removed those imputed cells (every non-integer value; each item had exactly one), so responses missing here were missing in the original survey, and the paper’s statistics, computed on the imputed file, will not reproduce exactly. | #2876 | 2026-10-06 |
gori_2020_dsq40 |
The source deposit is the authors’ analysis file, in which missing item responses had been filled with the item mean. IRW removed those imputed cells (every non-integer value; each item had exactly one), so responses missing here were missing in the original survey, and the paper’s statistics, computed on the imputed file, will not reproduce exactly. | #2876 | 2026-10-06 |
gori_2020_pss10 |
The source deposit is the authors’ analysis file, in which missing item responses had been filled with the item mean. IRW removed those imputed cells (every non-integer value; each item had exactly one), so responses missing here were missing in the original survey, and the paper’s statistics, computed on the imputed file, will not reproduce exactly. | #2876 | 2026-10-06 |
gori_2020_swls |
The source deposit is the authors’ analysis file, in which missing item responses had been filled with the item mean. IRW removed those imputed cells (every non-integer value; each item had exactly one), so responses missing here were missing in the original survey, and the paper’s statistics, computed on the imputed file, will not reproduce exactly. | #2876 | 2026-10-06 |
gosciak_2026_immigrant_admission |
Conjoint table (irw_conjoint). choice = ‘Which of these two immigrant profiles would you select for admission to the United States?’, one forced-choice task per respondent, no opt-out. The two profiles of a pair share each context value (different signals, e.g. Poland/Germany) and always differ in education; contexts were assigned by an adaptive bandit, then arms 7 and 10 were re-fielded as validation phases (trial_phase). Left/right position is inferred from the recorded choice; cov_garbage is the authors’ exclusion flag, and 121 repeat submissions by the same Prolific IDs are dropped. | 2026-10-08 | |
gothreau_2026_women_candidates |
Conjoint table (irw_conjoint). choice = “Which candidate would you support?”, forced (no opt-out), 7 tasks. Attribute text is the deposit’s English label set; respondents saw their own language, and origin is stored as the authors’ Born and Raised / Immigrated. Experience 0 and 10+ years and occupations lawyer and business had double randomization weight; about 86% of profiles are born and raised. Office (national/local) randomized except in US wave 2 (blank trial_office). 45 respondents without conjoint data dropped from the article’s 14,369. | 2026-10-07 | |
graham_2020_undemocratic_candidates |
Conjoint table (irw_conjoint). choice: ‘Which candidate do you prefer?’ (no opt-out); rating_turnout: ‘Would you vote in this election?’ (1 yes, 0 no), asked once per scenario and repeated on both candidates. Structured design: each respondent saw the same 13 party/policy-area scenarios, 9 pairing a generic with an undemocratic or scandal candidate; task follows scenario number, not display order. Tax-policy scenarios of respondents in states lacking an income, sales or corporate tax are dropped (state-specific wording, no deposited mapping); unanswered scenarios dropped. | 2026-10-09 | |
grahn_2025_tariffs_de |
Conjoint table (irw_conjoint). choice = preferred of two coalition agreements (paraphrase; forced choice, no opt-out); rating = likelihood item per agreement, 1 = ‘Gar nicht wahrscheinlich’ to 5 = ‘Sehr wahrscheinlich’ (wording not deposited). 3 tasks; German displayed text with emoji. Deposit N 3,864 vs article 3,994; sexual orientation, free text and panel IDs dropped. | 2026-10-09 | |
grahn_2025_tariffs_uk |
Conjoint table (irw_conjoint). choice = ‘Which candidate do you prefer?’ (forced choice, no opt-out); rating = ‘How likely is it that you would ever support each of the candidates?’ 1 = Not at all likely to 5 = Very likely. 5 tasks; both candidates of the respondent’s preferred party; level text as displayed with emoji. N = 1,500 as in the article; Prolific IDs and free text dropped. | 2026-10-09 | |
griffin_2020_candidate_income |
Conjoint table (irw_conjoint). Single-profile factorial vignette (name David/Denise Jones x income $3 million/$75,000/not stated, as Stata value labels); rating = vote for candidate 1 definitely not .. 4 definitely would; rating_represent 1 very well .. 4 very poorly; leadership/cares/honest/intelligent 1 extremely well .. 5 not well at all (raw, lower = better); wording is the .dta variable labels (paraphrase). Only the paper’s Experiment 1 is built; 10 respondents with no answer dropped (1,490 of 1,500); survey weight kept. | 2026-10-09 | |
grimmer_2017_credit_claiming |
Conjoint table (irw_conjoint). Factorial vignette (one press release per respondent, 5 factors); rating = approval of the legislator, 1-5 (lower = more approval per the authors’ coding; wording not deposited); rating_therm = 0-100 thermometer. 122 control-condition respondents without a vignette dropped; other post-treatment items lacking wording dropped. N = 952 treated of 1,074. | 2026-10-09 | |
gross_2026_cordon_dilemma |
Conjoint table (irw_conjoint). choice: which of two CDU candidates the respondent prefers (codebook ‘Candidate choice’; paraphrase), forced, no opt-out; rating: candidate thermometer -5..+5 as stored (higher = more positive). Levels are the stored German labels. Saxony, Thuringia and Western Germany samples pooled (cov_survey_region). 45 unanswered rounds dropped; panel ids dropped. | 2026-10-09 | |
gubitz_2022_incivility |
Conjoint table (irw_conjoint). Text vignette, 6 per respondent (task 1-6, profile 1). rating_uncivil: “How uncivil do you think the above scenario was?” 1 Not at all uncivil .. 5 Very uncivil; rating_newsworthy: “How newsworthy do you think the above scenario was?” 1 Not at all .. 5 Very newsworthy. Race and gender appear only through names (author’s coding in trial_ columns). Six PANAS emotion items dropped because which column is which emotion is undocumented; one scenario with no answers omitted. Respondents are White non-Latino only (author’s filter). | 2026-10-08 | |
guidi_2026_fdi_france |
Conjoint table (irw_conjoint). choice: should the government block the acquisition (1 = block, 0 = not; paraphrase of the article; single profile per task, so not blocking is the outside option). Attribute order randomized per task, not recorded; levels stored as the article’s English formulations (shown in French). Only screener-passing respondents deposited (1,223, matching the article); task order from row order. | 2026-10-08 | |
guidi_2026_fdi_germany |
Conjoint table (irw_conjoint). choice: should the government block the acquisition (1 = block, 0 = not; paraphrase of the article; single profile per task, so not blocking is the outside option). Attribute order randomized per task, not recorded; levels stored as the article’s English formulations (shown in German). Only screener-passing respondents deposited (1,327, matching the article); task order from row order. | 2026-10-08 | |
guidi_2026_fdi_italy |
Conjoint table (irw_conjoint). choice: should the government block the acquisition (1 = block, 0 = not; paraphrase of the article; single profile per task, so not blocking is the outside option). Attribute order randomized per task, not recorded; levels stored as the article’s English formulations (shown in Italian). Only screener-passing respondents deposited (1,183, matching the article); task order from row order. | 2026-10-08 | |
gulec_2023_pvs |
Item codes PKO1-6 are the deposit’s PKÖ1-6 with the diaeresis dropped. | 2026-10-11 | |
gunandran_2023_whoqol_bref |
The deposit’s domain totals equal plain sums of the stored items, so the negatively worded Q3, Q4 and Q26 are either stored already reversed or were never reversed; their item-rest correlations are about zero, so the direction cannot be determined from the data. Shipped as stored. | 2026-10-09 | |
gunderson_2025_judicial_appeals |
Conjoint table (irw_conjoint_2). One factorial vignette per respondent (US, Lucid): judicial candidate Mary Bailey / Michael Bailey x platform (criminal justice reform / tough on crime), 540 of 2,054 respondents (the rest saw a separate attorney experiment, not built). Eleven ratings stored raw: 1-4 items (vote, win chance, impartiality, crime-handling, reduce incarceration) have 1 = most positive option (direction inferred from the authors’ reversal; labels not in the deposit), the four performance items are 1-7 with 7 = very effective. Platform text paraphrased from the article. | 2026-10-14 | |
guul_2021_caseworker_creaming |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent, 2x2x2: the client is an expert or novice, diligent or lazy, male or female. Outcome rating_willingness is willingness to help, 0-10 (10 = very likely; higher = more willing); no opt-out. Level text is the stored value labels because the vignette wording is not in the deposit. 198 respondents without an outcome dropped (1,595 kept, the authors’ N); trial_cluster is the authors’ clustering variable. | 2026-10-14 | |
gwen_2025_interpersonal_comm |
The deposit gives no codebook: the four tables (interpersonal communication 13 items, self-esteem 2, social support 3, mental health 4) were split by the content of the statements, which head the columns in that order. Response-option labels were not published (1-5). | 2026-10-07 | |
haines_2023_loss_narrative_black |
Conjoint table (irw_conjoint_2). choice = which of two Democratic candidates the respondent would vote for (paraphrase), forced choice, no opt-out, 10 pairs; trial_narrative = 1 if the respondent first read the identity-politics loss narrative. Black respondents only (the white and Black samples are separate tables; race levels here: White / African American / Asian American / Hispanic). Attribute labels are the deposit’s text columns; tasks with no answer are omitted; N against the article not checked. | 2026-10-09 | |
haines_2023_loss_narrative_white |
Conjoint table (irw_conjoint_2). choice = which of two Democratic candidates the respondent would vote for (paraphrase), forced choice, no opt-out, 10 pairs; trial_narrative = 1 if the respondent first read the identity-politics loss narrative. White respondents only (the white and Black samples are separate tables; race levels here: White / Black / Asian / Hispanic). Attribute labels are the deposit’s text columns; tasks with no answer are omitted; N against the article not checked. | 2026-10-09 | |
hainmueller_2014_candidate |
Conjoint table (irw_conjoint). Outcomes: choice between two candidates (“which would you prefer for President”) and a 1-7 rating of each, recoded from the source’s 0-1 scale. The source has no task or profile column, so both are inferred from row order; 122 of 311 respondents have fewer than 6 tasks. No opt-out. MTurk worker IDs dropped. | 2026-10-07 | |
hainmueller_2014_immigrant |
Conjoint table (irw_conjoint). Outcomes: forced choice of which of two immigrants to admit, and a 1-7 rating of each (higher = more supportive of admission), the rating added from the same experiment’s fuller deposit (Hainmueller & Hopkins 2015, doi:10.7910/DVN/25505), matched row by row. No opt-out. Randomization was restricted: some professions appear only with some education levels, and escaping persecution only with some origins (22 combinations never occur), which AMCE estimates must account for. cov_survey_weight is the post-stratification weight. The same choices ship with the cjoint R package. | 2026-10-07 | |
hainmueller_2015_natur_forced |
Conjoint table (irw_conjoint). Forced choice of the applicant preferred for naturalization (choice = 1 - the source’s ‘rejected’); no opt-out. Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). The article analyses only respondents aged over 30 who voted in naturalization referendums (cov_voted_target = 1), weighted by cov_survey_weight. | 2026-10-08 | |
hainmueller_2015_natur_paired |
Conjoint table (irw_conjoint). Each applicant accepted or rejected separately; rating_reject stored raw (1 = reject, 0 = accept). Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). The article analyses only respondents aged over 30 who voted in naturalization referendums (cov_voted_target = 1), weighted by cov_survey_weight. | 2026-10-08 | |
hainmueller_2015_natur_paired_vignette |
Conjoint table (irw_conjoint). Each applicant accepted or rejected separately; rating_reject stored raw (1 = reject, 0 = accept). Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). The article analyses only respondents aged over 30 who voted in naturalization referendums (cov_voted_target = 1), weighted by cov_survey_weight. | 2026-10-08 | |
hainmueller_2015_natur_single |
Conjoint table (irw_conjoint). Accept or reject the applicant (choice = 1 accepted, from the source’s ‘rejected’); rejecting is the opt-out. Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). The article analyses only respondents aged over 30 who voted in naturalization referendums (cov_voted_target = 1), weighted by cov_survey_weight. | 2026-10-08 | |
hainmueller_2015_natur_single_vignette |
Conjoint table (irw_conjoint). Accept or reject the applicant (choice = 1 accepted, from the source’s ‘rejected’); rejecting is the opt-out. Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). The article analyses only respondents aged over 30 who voted in naturalization referendums (cov_voted_target = 1), weighted by cov_survey_weight. | 2026-10-08 | |
hainmueller_2015_natur_students |
Conjoint table (irw_conjoint). Forced choice of the applicant preferred for naturalization (choice = 1 - the source’s ‘rejected’); no opt-out. Level text is the authors’ short English labels; profile 1/2 is file order, not screen position; the undocumented irregular task 10 (regular == 0) is dropped, as in the authors’ code. Observed restrictions (German/Austrian applicants never ‘Adequate’ German; age 21 never 29 years since arrival) and unequal level weights (Italy, Turkey ~26%). | 2026-10-08 | |
halim_2025_digital_finance |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_behavior |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_literacy |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_self_efficacy |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_skills |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_stress |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
halim_2025_fin_wellbeing |
No codebook: the construct of each five-item block is read from its prefix against the constructs the record names (FLC literacy, FSK skills, DFC digital finance, FSS stress, FSE self-efficacy, FBR behaviour, FWB well-being). | 2026-10-09 | |
hamzawi_2025_party_valence |
Conjoint table (irw_conjoint). Outcome: forced choice of which of two hypothetical parties’ candidate (district vote, trial_context (S/M)MD) or party (PR vote) is preferable; skipping was allowed and the 524 skipped tasks (2.8%) are omitted, so 1,834 of 1,866 respondents remain. Tasks 1-5 are always the district block although block order was randomized for half the sample (not recorded), so task is not the display order across blocks. Levels are the authors’ English labels; randomization was restricted to realistic combinations (e.g. MPs vs ministers, committee chairs, bills), so level shares are unequal. Attribute order was randomized per respondent and is kept as attrpos_*. | 2026-10-08 | |
han_2026_fairness_packages |
Conjoint table (irw_conjoint). choice: “어느 정책안이 한국 사회를 더 공정하게 만든다고 생각하십니까?” (which policy package would make Korean society fairer), A / B, forced, no opt-out; choice_support: a second A/B pick about the same pair, wording not deposited. 6 pairs per respondent, 3 attributes with 4 levels each, varied independently (codebook). Level text from the data labels, which differ slightly from the codebook’s Korean mock-up. | 2026-10-08 | |
han_2026_party_newcomers |
Conjoint table (irw_conjoint_2). 2x2 factorial vignette, one per respondent (South Korea, 2026 survey of 2,000 public + 1,000 party members, pooled with cov_sample): politician’s previous party (same bloc as Party A / opposing bloc) x organizational contribution (recently joined / sustained participation). Seven 7-point ratings (1 = strongly disagree, 7 = strongly agree; fairness 1 = not at all fair): accept as member, campaign, party office, local nomination, national nomination, leadership, fairness to existing members. Korean text as displayed; no article DOI. | 2026-10-14 | |
hankinson_2024_housing_compensation |
Conjoint table (irw_conjoint_2). Single-profile proposals (Boston, up to 5 per respondent): housing development with affordable units or 100% market-rate, then a compensation offer (amount $50-\(10,000; private cash or public park investment). choice = support (1) or oppose (0) the proposal with compensation (opt_out yes: single-profile accept/reject); rating_support_before_compensation = support without compensation, 0-1 (asked before the compensation was shown). Distance is not randomized (trial_distance_km). 229 rows with missing affordable-housing level dropped; 531 respondents. | | 2026-10-14 | | `hankla_2022_bihar_candidates` | Conjoint table (irw_conjoint). choice: 'If these two candidates were running against each other for MLA, and the election were held today, which would you vote for?' (forced choice, no opt-out); candidates shown as drawings with surname, title and an appeal read aloud; party (BJP vs RJD) shown only in tasks 3-4, '(not shown)' elsewhere. Task/profile inferred from row blocks; the authors' binary StrongPref recode, 18 incomplete or off-pattern tasks and one empty respondent dropped (1,999 of ~2,000 respondents). | | 2026-10-09 | | `hansen_2023_teacher_extra_role` | Conjoint table (irw_conjoint). Four dilemmas, one student profile each (task = dilemma, trial_dilemma), rating 1-10 from '1: [would not], helt sikkert' to '10: [would], helt sikkert' (higher = more willing to go beyond the job); Danish level text as displayed. 287 teachers with no rating dropped (1,507 kept, as in the paper); gender option 'Andet/oensker ikke at svare' set to NA. | | 2026-10-09 | | `hansen_2023_warmth` | Conjoint table (irw_conjoint_2). 2x2 factorial vignette, one per respondent (Denmark, YouGov panel, 2021): municipality's decision on a nursing-home application (rejected/approved) x caseworker warmth (low/high). Ratings: trust in caseworker and in municipality (0-1, author's recode of 0-10, higher = more trust) and inclination to complain (1 = very inclined, 7 = very disinclined, 88 = don't know). Level text paraphrased from the article because the vignette appendix is not deposited. No respondent id in the file (id = row number). | | 2026-10-14 | | `hare_2024_trophy_hunting` | Conjoint table (irw_conjoint). Factorial vignette (2 x 2 x 3, one per respondent): rating = how acceptable it would be for the hunt described to take place, 1 = Very unacceptable ... 7 = Very acceptable (numbered from the deposited answer text in scale order). Level text paraphrases the manipulated vignette language from the article. N = 1,192 as in the article; authors removed 'I don't know' answers, non-binary/undisclosed gender and speeders before deposit. | | 2026-10-09 | | `hargrave_2023_speech_style` | Conjoint table (irw_conjoint_2). Ratings 1 strongly disagree to 5 strongly agree (agree or disagree that the MP seems likeable / competent; that the argument is emotional, aggressive or evidence-based, asked only for the style shown); 'Don't know' is missing and 220 of 5,028 tasks with no answer at all are omitted (1,628 respondents have a row). Styles and policies are drawn without repetition across a respondent's three tasks. Speech texts are in the paper's appendix, not the deposit; levels are the design factors with the authors' labels. MP gender comes from the first name (crosswalk, name signal); the paper's 5,028 observations are the pre-omission count. | | 2026-10-11 | | `hariyoko_2024_tea_preferences` | Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent scored 12 fixed tea-drink cards (flavour, flavour intensity, pack type) on a 1-5 scale stored raw; the deposit gives no question wording or scale labels. Fixed 3x2x2 full factorial, task = card number in column order (display order not documented). One covariate: purchase frequency (Indonesian text). Exactly 100 respondents. | | 2026-10-12 | | `harris_2023_neighbor_cooperation` | Conjoint table (irw_conjoint). Single neighbour vignette; ratings kept raw: rating = 'Say you had loaned this neighbor some money, would you trust them to pay it back?' (1 No, 2 Yes), rating_donations / rating_follow_headman / rating_share_well 1 Very likely to 4 Not at all likely (higher = less likely), rating_way_of_life 1 Agree / 2 Disagree; don't know and refusals NA. Only the non-co-ethnic arm (6,209 of 12,362) is kept because the deposit did not save the ethnic group shown to the 6,153 co-ethnic-arm respondents, so the co-ethnic contrast cannot be estimated (the arm was randomized, so the rest is a random subset); the title text's pronoun follows the neighbour's gender; 21 respondents with no answer omitted. | | 2026-10-08 | | `harsgor_2023_covid_aid_palestinians` | Conjoint table (irw_conjoint). Forced choice between two policy profiles per task (`selected`), no opt-out; question wording not in the deposit. Levels are the authors' English labels; the display language is not documented. Observed restriction: Medical Aid never shown with Palestinian illness Deterioration and Monitor never with Improvement, so those levels are rarer. 334 unanswered tasks dropped (1,586 of 1,589 respondents remain). | | 2026-10-08 | | `hartman_2020_refugee_hosting` | Conjoint table (irw_conjoint). Three paired tasks shown as pictographs; choice = which of two refugee families the respondent would host (paraphrase; forced choice, no opt-out). Levels are the authors' Table 1 labels; ethnicity is relative (co-ethnic / not co-ethnic); co-ethnic non-co-religious profiles were excluded for respondents from religiously homogeneous groups. task/profile inferred from row order; 2 respondents and 4 tasks with missing attributes dropped; N = 1,089 (manuscript 1,230). | | 2026-10-09 | | `hartmann_2023_covid_restrictions` | Conjoint table (irw_conjoint). Outcomes: forced choice of the preferred of two COVID-19 policy proposals (no opt-out), a 0-10 support rating of each ('definitiv dagegen'-'definitiv dafuer'), 0-10 trust in government asked about one randomly drawn proposal per task, and 0-10 vaccination likelihood asked only of unvaccinated respondents (wording paraphrased). The two proposals of a task always differ and share the severity scenario; trust is NA for 141 respondents with a programming-error flag; policy levels are short German design labels. | | 2026-10-08 | | `hartmann_2026_disaster_relief` | Conjoint table (irw_conjoint). choice is the candidate the respondent supported in each of 6 pairs (exact wording not deposited); exactly one profile chosen per pair, no opt-out. Attribute text is the author's short English labels, not the wording respondents saw; effort and effectiveness levels are crossed freely and restrictions are not documented. trial_disaster_prime is the respondent's arm in a disaster prime shown before the conjoint; unlabelled covariate codes and village are dropped. | | 2026-10-07 | | `hassell_2020_journalist_coverage` | Conjoint table (irw_conjoint_2). Journalists chose which of two candidate announcements for governor to cover in person (no opt-out), two pairs each, seven attributes. The file has no respondent id, so id/task/profile are inferred from blocks of four rows; 897 respondents answered (the paper's N = 3,276 candidate rows), and 83 pairs with a blank occupation were dropped, leaving 882 respondents. Further outcomes (readerscover, editorscover) have undocumented wording and are not kept. | | 2026-10-11 | | `hassell_2023_electability` | Conjoint table (irw_conjoint). 8 pairs of primary candidates per respondent; choice_electable = candidate identified as more electable, choice_vote = preferred candidate (paraphrase from codebook; questionnaire not deposited), forced choice, no documented opt-out. Ideology levels depend on the primary scenario (Democratic: very liberal/liberal/moderate; Republican: moderate/conservative/very conservative), district attributes are per pair; skipped pairs NA (authors' 0), 48 respondents without answers dropped; Qualtrics/panel ids and free text dropped; paper N not checked. | | 2026-10-09 | | `haugom_2021_ski_resort` | Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each skier rated 18 ski-day scenarios (8 attributes) on a 0-100 scale, stored raw (the question wording and anchors are not in the deposit; the price effect in the data is negative, so higher = more favourable). Fixed designs in ten questionnaire versions (trial_version); rows with a missing rating are omitted, so 375 of 399 respondents appear; price currency and units of temperature and queue time are not documented. | | 2026-10-12 | | `hayashi_2024_nuclear_waste_site` | Conjoint table (irw_conjoint). choice: which of Construction Plan A and B the respondent would accept (forced, no opt-out); rating: willingness to accept each plan, 1 (absolutely not willing) to 7 (very strongly willing); wording and levels are the codebook's English translation of a Japanese survey. Attribute order randomized per respondent (attrpos_*); no documented restrictions. Municipality of residence dropped; one implausible age (3) set to NA, 13 respondents aged 14-17 kept. | | 2026-10-08 | | `heermann_2022_covid_aid_2020` | Conjoint table (irw_conjoint_2). Two single-profile text vignettes per respondent (medical aid, financial aid) with three randomized attributes, rated 1-7 for support (1 = not support at all, 7 = strongly support); no opt-out. Attribute level text taken from the article appendix (A, L); the deposit stores 0/1 dummies, so the printed variant is the dummy-1 level and the opposite level is paraphrased. Task numbers follow the vignette, not the order shown (not recorded). | | 2026-10-14 | | `heermann_2022_covid_aid_2021` | Conjoint table (irw_conjoint_2). Two single-profile text vignettes per respondent (medical aid, financial aid to an EU country) with five randomized binary attributes, rated 1-7 for support (1 = not support at all, 7 = strongly support); no opt-out. Level text from the article appendix; the deposit stores 0/1 dummies whose direction is read from the printed vignette and the authors' figure labels (opposite levels paraphrased). | | 2026-10-14 | | `hegewald_2026_urban_rural` | Conjoint table (irw_conjoint). Single-profile design: rating_rural is the judgement that the person "typically lives in a rural area" (1) rather than an urban area (0); rating_sureness is how sure (1 = very sure to 4 = very unsure); rating is a 0-100 feeling thermometer. Levels were uniform and independent; attribute text is the English master with a [country] placeholder, while respondents saw their national language. Countries pooled (cov_country, Greece as GR); the author's sureness-weighted outcome dropped. | | 2026-10-07 | | `heinrich_2020_nasty_aid` | Conjoint table (irw_conjoint). rating: 'Please express your support for each aid package by checking the buttons.' 9-point, 1 = Oppose, 5 = Indifferent, 9 = Support, both packages rated in each of 4 tasks; no choice question. Remedial aid exists only for the four nasty-regime issues (design restriction); its displayed wording is not deposited (amount as codebook value, '(not shown)' when 0, inferred). The authors' attention/time exclusions (to n=2,093) are not applied; the needed covariates are kept. | | 2026-10-09 | | `heinze_2025_sarpanch_vignette` | Conjoint table (irw_conjoint). Factorial vignette (2 gender x 2 caste, equal probabilities), one president per respondent; three 0/1 ratings: president most active in village decisions (1) vs someone else (0), would resign (1) vs resist (0) when elites ask, would likely face backlash (1) vs not (0). Gender and caste were each signalled by a name plus a category word; the table stores the category words. 2 respondents with no condition and 6 with no outcome dropped; N 2,360 vs 2,347 in the article. | | 2026-10-08 | | `heinzel_2024_aid_reforms` | Conjoint table (irw_conjoint). Article Experiment 3. rating: 'Please indicate to what extent you support giving foreign aid through each aid agency' 1 (strongly oppose) to 10 (strongly support), one rating per agency, one task of two agencies; no choice, no opt-out. Each attribute is a statement shown with Yes/No, independently randomized. Survey weight kept (cov_survey_weight). N = 2,058 matches the article. | | 2026-10-09 | | `heinzel_2024_aid_transparency` | Conjoint table (irw_conjoint). Article Experiment 2. rating: 'Please indicate to what extent you support giving foreign aid through each aid agency' 1 (strongly oppose) to 10 (strongly support), one rating per agency, one task of two agencies; no choice, no opt-out. Each attribute is a statement shown with Yes/No, independently randomized. Survey weight kept (cov_survey_weight). N = 2,058 matches the article. | | 2026-10-09 | | `heinzel_2024_world_bank_dpl` | Conjoint table (irw_conjoint). Paired profiles without forced choice, 7 tasks. rating_approval: 'In your personal opinion, how likely is each project to get board approval?' and rating_impact: 'In your personal opinion, how likely is each project to have a positive impact on development outcomes in the recipient country?', each 1 = Very unlikely to 10 = Very likely. Eight binary features independently randomized; levels as in Appendix Table A2. Survey weight kept (weights_final); alternative robustness weights dropped. | | 2026-10-09 | | `heinzel_2025_imf_conditionality` | Conjoint table (irw_conjoint). Each respondent rated Loan A and Loan B (8 Yes/No conditions) for support (1 strongly oppose - 6 strongly support), willingness to pay more taxes and to endure spending cuts (1 strongly unwilling - 6 strongly willing); no choice question. Three countries pooled with cov_country and YouGov weights (cov_survey_weight); 2,643 respondents in the file vs 2,694 reported. | | 2026-10-09 | | `heinzel_2025_io_earmarking` | Conjoint table (irw_conjoint). Paired contributions, 7 tasks; choice = forced choice between Contribution 1 and 2 (wording not reported, no opt-out); rating_amount = minimum worthwhile grant (million USD) and rating_overhead = overhead to charge (%), both open-text answers stored as the authors' numeric cleaning; rating_refuse = 1 if the respondent wrote 'No' (reject the contribution). Six binary features independently randomized; only respondents who completed all 7 pairs are in the deposit (280, as in the paper). | | 2026-10-09 | | `helbling_2020_climate_migrants` | Conjoint table (irw_conjoint_2). Single-vignette factorial survey: choice = yes to 'Do you think Germany should let him immigrate?' (single accept/reject, opt_out yes). Attribute levels are the codebook's short cell labels, not the vignette wording; id is the row number; one respondent without an answer dropped (1,101 of 1,102); the second vignette block (messages) is not included. | | 2026-10-11 | | `helbling_2023_immigration_policy` | Conjoint table (irw_conjoint_2). choice = which of the two immigration policy packages the respondent prefers (paraphrase), forced choice, no opt-out, 3 contests. Attribute text is the authors' English factor labels of a German-language survey; profile numbers follow file order, not screen position; unanswered contests are omitted; samples (pretest, main halves, extra batch) pooled as by the authors, with cov_sample. N against the article not checked. | | 2026-10-09 | | `helbling_2023_integration_policy` | Conjoint table (irw_conjoint_2). choice = which of the two integration policy packages the respondent prefers (paraphrase), forced choice, no opt-out, 3 contests. Attribute text is the authors' English factor labels of a German-language survey; profile numbers follow file order, not screen position; unanswered contests are omitted; samples (pretest, main halves, extra batch) pooled as by the authors, with cov_sample. N against the article not checked. | | 2026-10-09 | | `helbling_2023_naturalization_policy` | Conjoint table (irw_conjoint_2). choice = which of the two naturalization policy packages the respondent prefers (paraphrase), forced choice, no opt-out, 3 contests. Attribute text is the authors' English factor labels of a German-language survey; profile numbers follow file order, not screen position; unanswered contests are omitted; samples (pretest, main halves, extra batch) pooled as by the authors, with cov_sample. N against the article not checked. | | 2026-10-09 | | `helbling_2023_numbers_rights_s2` | Conjoint table (irw_conjoint_2). One policy vignette per row; rating = 'Inwieweit wären Sie mit einer solchen Zuwanderungspolitik einverstanden?' 1 (Voll und ganz) - 5 (Überhaupt nicht), stored raw so lower = more agreement; Weiß nicht = NA. Levels are the displayed fragments (e.g. 'zu veringern.', 'in Bezug auf Bildung verschärfen', 'Rechte ausweiten, damit sie stärker vom Wohlfahrtsstaat profitieren können.'). Waves 14-23 of study12.dta; no respondent id (row number); N against the article not checked. | | 2026-10-09 | | `helbling_2023_numbers_rights_s3` | Conjoint table (irw_conjoint_2). rating = support for the proposed policy, answer text mapped as by the authors: 1 voll und ganz ablehnen ... 6 Voll und ganz unterstützen; Weiß nicht = NA (wording not in the deposit). Two vignettes per respondent: task 1 region x numbers, task 2 the same plus requirements and rights (absent clauses are '(not shown)'; task 2 shows the numbers verb as 'erhöht'/'verringert'); trial_task1_echo = the respondent's task-1 stance echoed in the task-2 text. Region levels drawn about 2:1. Only attention-check passers saw it; N against the article not checked. | | 2026-10-09 | | `hemesath_2024_chatbot_citizens` | Conjoint table (irw_conjoint). Forced choice between two chatbot profiles (paraphrase: which chatbot do you prefer), exactly one chosen per task, no opt-out; 8 attributes with 3 levels each, in German. Sample: citizens; trial_prime records a priming module (Control/Duty/Service) shown before the conjoint. The -5..+5 relative-preference slider is dropped because its wording and anchors are undocumented. Randomization restrictions and attribute order undocumented; all level pairs occur. | | 2026-10-08 | | `hemesath_2024_chatbot_officers` | Conjoint table (irw_conjoint). Forced choice between two chatbot profiles (paraphrase: which chatbot do you prefer), exactly one chosen per task, no opt-out; 8 attributes with 3 levels each, in German. Sample: municipal front desk officers. The -5..+5 relative-preference slider is dropped because its wording and anchors are undocumented. Randomization restrictions and attribute order undocumented; all level pairs occur. | | 2026-10-08 | | `henderson_2014_party_guess` | Conjoint table (irw_conjoint). rating_party: "If you had to guess, do you think this candidate is a Democrat or Republican?" (1 Democrat, 2 Republican); rating_sure: "How sure are you about the candidate's party?" (1 very sure .. 4 very unsure); rating: favorability 0 very unfavorable .. 10 very favorable, 'Not sure' (about 19%) set NA. One profile per task, four tasks; issue priorities 1-3 never repeat within a profile. Level text from value labels (display table not deposited). 1,000 respondents; team weight in cov_survey_weight. | | 2026-10-08 | | `hernandez_martinez_2025_math_attitudes` | The AFE (ids 1-3629) and AFC (ids 100001-102449) halves are one sample split at random by the authors; cov_sample marks the half. The deposit's age column mixes category codes with years and is not shipped. | | 2026-10-11 | | `higashijima_2025_election_timing_2021` | Conjoint table (irw_conjoint). choice = the preferred of two hypothetical election-calling scenarios (`selected`, exactly one per task; question wording not in the deposit). 8 attributes, Japanese text shown; attribute order not recorded. Remaining term 0 occurs only with reason 'full term' (任期満了のため) and vice versa. Counts not checked against the article (not retrievable). | | 2026-10-07 | | `higashijima_2025_election_timing_2022` | Conjoint table (irw_conjoint). choice = the preferred of two hypothetical election-calling scenarios (`selected`, exactly one per task; question wording not in the deposit). 8 attributes, Japanese text shown; attribute order not recorded. Remaining term 0 occurs only with how-decided 'term expired' (任期満了のため) and vice versa; reason 'full term' only with remaining term 0. The authors' code also forbids term 0 with other reasons, but 2,179 such profiles occur in the data (flagged). Counts not checked against the article (not retrievable). | | 2026-10-07 | | `higashijima_2026_muslim_immigrants` | Conjoint table (irw_conjoint). choice = which of Immigrant 1 / Immigrant 2 was chosen (forced choice; wording not deposited); rating = 1-7 rating of each immigrant (wording and anchors not deposited; higher = more favourable, inferred); rating_sim = a second 1-7 rating (ENGQ51, 'sim', meaning undocumented). Restrictions: high-skill professions only with university/postgraduate education; Russian ethnicity only with Russian-speaking language levels. Levels are the authors' English labels. All 3,000 respondents kept; the article analyses Muslim respondents. | | 2026-10-07 | | `hillenbrand_2025_asylum_reform` | Conjoint table (irw_conjoint). choice: 'If you had to choose a policy mix for Germany, which of the two options would you prefer?' (forced, no opt-out); rating: 'And how would you rate the two options?' 1 very negative - 7 very positive (English translation). 7 tasks, 4 attributes, all combinations allowed, attribute order randomized per respondent (attrpos_*); attribute text is the paper's English translation. All 2,211 respondents kept; the paper's main results use attention passers (cov_attention_pass = 1). | | 2026-10-09 | | `himmelstein-number_series-2025` | The source's answer key for NS_6 (200, 198, 192, 174, ___) is 165, but the differences (2, 6, 18) imply 120; about 1% of respondents gave 165. IRW keeps the source's scoring, so resp for NS_6 follows the 165 key. | [#2513](https://github.com/ben-domingue/irw/issues/2513) | 2026-09-28 | | `hitlan_2025_ders16` | The record instructs "select cas_08=0": 17 respondents flagged for failed attention checks and one with the flag missing are dropped from all three hitlan_2025 tables. | | 2026-10-11 | | `hobolt_2021_brexit_bbc_dg` | Conjoint table (irw_conjoint). choice: 'Which of the two applicants would you prefer as the next Director General of the BBC?' (forced, no opt-out), 5 pairs. Names drawn without replacement (no name repeats within a respondent); other attributes uniform and independent; attribute rows shuffled once per respondent (attrpos_). 1,635 respondents, all tasks answered; weight kept. | | 2026-10-09 | | `hobolt_2021_brexit_lodger` | Conjoint table (irw_conjoint). choice: 'Which of the two lodgers would you prefer?' (forced, no opt-out), 5 pairs. Names drawn without replacement (no name repeats within a respondent); other attributes uniform and independent; attribute rows shuffled once per respondent (attrpos_). 1,669 respondents, all tasks answered; weight kept. | | 2026-10-09 | | `hobolt_2022_brexit_negotiations` | Conjoint table (irw_conjoint). choice: 'Which of these two outcomes do you prefer?' (forced, no opt-out); rating_acceptable: 1 completely acceptable to 4 completely unacceptable, raw (lower = more acceptable); rating_respects_referendum: 1 if the respondent said this outcome (or both) would respect the referendum result, 0 otherwise, Don't know = NA; rating_likely: likelihood the chosen outcome happens, 1 very likely to 4 not at all likely, chosen profile only. 5 random pairs; a fixed sixth pair is dropped. Attribute rows shuffled once per respondent. Weight kept. | | 2026-10-09 | | `hobolt_2024_authoritarian_reactions` | Conjoint table (irw_conjoint). choice: respondents choose one of the two candidates (forced; exact wording not deposited). rating: each candidate on a 7-point scale, restored from the authors' 0-1 rescaling (rating2 * 6 + 1); higher = more favourable (chosen profiles rate higher), anchors unknown. Attribute order randomized but not recorded; no restrictions documented. All 4,012 respondents, matching the article. | | 2026-10-07 | | `hoellerbauer_2023_civil_society_orgs` | Conjoint table (irw_conjoint). choice_meeting = which of two organizations (A/B) the respondent would rather attend meetings of (paraphrase of the abstract), choice_scandal = organization picked in the scandal question (wording not deposited; picking is unfavourable, the author reverses it), both forced A/B, no opt-out. Level text from the codebook in English, display language undocumented; restrictions and attribute order undocumented. 13 profiles with an unsaved attribute value have their task dropped. | | 2026-10-07 | | `hoffmann_2025_immigrant_sexuality` | Conjoint table (irw_conjoint). choice = "which of these two immigrants would you personally prefer to see admitted to the United States?" and choice_values = "which ... do you think has greater shared values with the United States?" (forced choice, no opt-out); rating and rating_values are 1-7 ratings of admission (1 = absolutely not admit, 7 = definitely admit) and of shared values (1 = none, 7 = a great deal). Levels are the phrases embedded in written descriptions; no randomization restrictions are documented. 1,621 respondents after the authors' filters (article: 1,650). | | 2026-10-07 | | `holland_2024_migrants_col` | Forced choice, no opt-out: which migrant the respondent "would prefer to stay in their city and receive housing and employment assistance" (the article's wording); level text is the authors' English labels. 29 unanswered tasks and 51 task-3 pairs missing a level dropped; 1,005 in the file vs "1,000" in the article. Names, phones, addresses, IP addresses and GPS dropped. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `holland_2024_migrants_ven` | Same design and outcome as the Colombian table (forced choice, no opt-out, authors' English labels). 44 unanswered tasks dropped; 1,612 respondents in the file match the article. Names, phones, addresses, IP addresses and GPS dropped. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `holliday_2026_policy_municipal` | Conjoint table (irw_conjoint). Forced choice of the preferred candidate (no opt-out), "Given this choice, which federal House of Representatives candidate would you prefer?" in tasks 1-5 and the same for a city council candidate in tasks 6-10 (trial_office). Each task shows 4 policies, the same 4 for both candidates, drawn from a pool; other policy attributes are (not shown) and the row order of the 4 is not kept. Party labels shown to a random half of respondents (trial_show_party). cov_baseline_<policy> is the respondent's own stance (1 = affirmative), which the authors use to code agreement; cov_survey_weight is the raked weight. Respondents are those left after the authors' speeder and attention-check drops. | | 2026-10-07 | | `holliday_2026_policy_state_2021` | Conjoint table (irw_conjoint). Forced choice of the preferred candidate (no opt-out), "Given this choice, which federal House of Representatives candidate would you prefer?" in tasks 1-5 and the same for a state assembly candidate in tasks 6-10 (trial_office). Each task shows 4 policies, the same 4 for both candidates, drawn from a pool; other policy attributes are (not shown) and the row order of the 4 is not kept. No party labels. cov_baseline_<policy> is the respondent's own stance (1 = affirmative), which the authors use to code agreement; cov_survey_weight is the raked weight. Respondents are those left after the authors' speeder and attention-check drops. | | 2026-10-07 | | `holliday_2026_policy_state_2022` | Conjoint table (irw_conjoint). Forced choice of the preferred candidate (no opt-out), "Given this choice, which federal House of Representatives candidate would you prefer?" in tasks 1-5 and the same for a state assembly candidate in tasks 6-10 (trial_office). Each task shows 4 policies, the same 4 for both candidates, drawn from a pool; other policy attributes are (not shown) and the row order of the 4 is not kept. Candidates carry party labels (attr_party). cov_baseline_<policy> is the respondent's own stance (1 = affirmative), which the authors use to code agreement; cov_survey_weight is the raked weight. Respondents are those left after the authors' speeder and attention-check drops. | | 2026-10-07 | | `holloway_2025_indigenous_candidates` | Conjoint table (irw_conjoint_2). Single-profile factorial: each respondent saw one photographed prospective candidate (4 ethnicity/skin-tone groups x man/woman, one photograph per cell) and gave 8 trait ratings (agree scale 1 Strongly disagree - 5 Strongly agree) and an attractiveness rating (1-9); no choice and no opt-out. The scale-order flag (normal/reversed) is kept as trial_scale_order; whether reversed values were recoded by the authors is not stated. Survey 1 (stimulus pre-test) is not built; N against the paper not checked beyond the readme's ~4,000. | | 2026-10-11 | | `holzknecht_2022_listening` | treat varies by task within student: 153 students heard both multiple-choice tasks twice and the note-form tasks once, 153 the reverse. For a twice-heard task resp is the answer after the second listening; the first-listening answers are not shipped. | | 2026-10-09 | | `horiuchi_2018_japan_manifestos` | Conjoint table (irw_conjoint). choice: "Suppose that the following two parties were nominating candidates in this general election. Which party would you support? Even if you are not entirely sure, please indicate which of the two you would dare to support." (codebook translation of the Japanese), forced choice, no opt-out; unanswered tasks omitted. Attribute order randomized once per respondent (attrpos_, constant across a respondent's tasks); restrictions not documented, level shares near uniform. Sample filter as in the authors' code (fielding window, consent); 1,951 respondents, article N not checked. | | 2026-10-07 | | `horiuchi_2020_japan_politicians` | Conjoint table (irw_conjoint). choice is a forced choice of which of two people is more desirable as a member of the given house and tier (wording in Japanese; house and tier vary by block, see trial_house and trial_tier), no opt-out. Attribute text is the Japanese displayed, with prefecture-specific hometown, occupation and parent levels; attribute order was randomized per respondent (attrpos_). Restriction observed: 'never in office' experience appears only with zero previous terms. Qualtrics IDs, free-text and attitude items dropped; the authors' entropy-balancing weights are not deposited. | | 2026-10-07 | | `horiuchi_2022_sdb_candidates` | Conjoint table (irw_conjoint). rating = 'How likely or unlikely do you think you would be to vote for each candidate from the [Democratic/Republican] Party instead of the candidate from the [Republican/Democratic] Party in each of the two elections?' (party filled per respondent; Very Unlikely = 1 to Very Likely = 7). The two candidates in a pair are not competing against each other. Arms: fully randomized vs constrained (only Scandal differs within a pair); in both arms exactly one candidate per pair has the scandal. Attribute order randomized per respondent and recorded. Respondents who failed the attention checks are kept and flagged (cov_attention_pass_1 = Q4.1, which the authors use to drop respondents; cov_attention_pass_2 = Q36.1). | | 2026-10-07 | | `horiuchi_2022_sdb_shoes` | Conjoint table (irw_conjoint). rating = 'How likely are you to purchase shoe 1?' (asked for shoe 1 and shoe 2; Very Unlikely = 1 to Very Likely = 7). Four arms (trial_design): in 'random' arms all attributes vary, in 'constrained' arms only the focal attribute (Eco-Friendly Materials for treat, Gel Cushioning for control) differs between the two shoes; in every arm the focal attribute always differs within a pair. Attribute order randomized per respondent and recorded. One MTurk worker who answered twice dropped. | | 2026-10-07 | | `horiuchi_2024_f35` | Conjoint table (irw_conjoint). choice: which of two F-35 deployment plans (計画案A / 計画案B) the respondent supports more, forced, no opt-out. 8 pairs, the 8th repeating the 1st with the same left/right order (trial_repeat_of); 5 attributes, attribute order randomized per respondent. National and Okinawa oversample pooled; authors' sample filters not applied but recoverable from cov_ columns; non-consenting and blank respondents omitted. | | 2026-10-08 | | `horiuchi_2024_osprey` | Conjoint table (irw_conjoint). choice: which of two Osprey deployment plans (計画案A / 計画案B) the respondent supports more, forced ("if you cannot say clearly, choose the one you would rather support"), no opt-out. 10 pairs; 6 attributes, attribute order randomized per respondent. National and Okinawa oversample pooled; authors' sample filters (reCAPTCHA score, Okinawa residence) not applied but recoverable from cov_ columns; the authors analyse tasks 1-7 only. | | 2026-10-08 | | `horne_2018_bridewealth_chop` | Conjoint table (irw_conjoint_2). Single text vignette per respondent, 3 x 2 between subjects: bridewealth none/some/all by whether the man paid chop money; level text is paraphrased from the codebook (vignette wording not deposited). Eight 1-10 ratings: how right or wrong the man's/woman's family and men/women in the community will say the woman is (1 very wrong) and how much they will encourage the man to leave her (1 strongly discourage); no opt-out. The same 2,400 respondents also answered a second vignette (horne_2018_bridewealth_hiv); conditions follow the authors' versnumber recode, not the o-columns. | | 2026-10-14 | | `horne_2018_bridewealth_hiv` | Conjoint table (irw_conjoint_2). Single text vignette per respondent, 3 x 2 between subjects: bridewealth none/some/all by whether the man is HIV positive or negative; level text is paraphrased from the codebook (vignette wording not deposited). Eight 1-10 ratings: how right or wrong the man's/woman's family and men/women in the community will say the woman is (1 very wrong) and how much they will encourage the man to leave her (1 strongly discourage); no opt-out. The same 2,400 respondents also answered the chop-money vignette (horne_2018_bridewealth_chop); conditions follow the authors' versnumber recode, not the o-columns. | | 2026-10-14 | | `horne_2021_energy_rebound` | Conjoint table (irw_conjoint_2). One text vignette per respondent: electricity source (utility/solar) x household energy use (low/high); 16 ratings of the family stored raw on 1-11 (buttons labelled 0-10; rating = approval). Question order (knowledge items before or after moral items) is trial_question_order. One Qualtrics block labelled high-use shows low-use text and is coded as displayed. No respondent id: id = row order; 13 respondents with no answers omitted. | | 2026-10-14 | | `horne_2021_privacy_norms_prolific` | Conjoint table (irw_conjoint_2). Four ratings on 1-10 scales after a household energy-app vignette (others' approval of the provider analyzing and selling data, expected invasion by the provider, trust in the provider, willingness to use the app; paper Table 2 wording); two attributes (app popularity, other providers' data practices) in a 2x2 between-subject design, one vignette per respondent; no choice or opt-out. Study 1 of the paper, analysed separately from Study 2. N = 403 as in the paper. | | 2026-10-11 | | `horne_2021_privacy_norms_yougov` | Conjoint table (irw_conjoint_2). Four ratings on 1-10 scales after a household energy-app vignette (others' approval of the provider analyzing and selling data, expected invasion by the provider, trust in the provider, willingness to use the app; paper Table 2 wording); two attributes (app popularity, other providers' data practices) in a 2x2 between-subject design, one vignette per respondent; no choice or opt-out. Study 2 of the paper (weighted YouGov sample), analysed separately from Study 1. N = 1,200 as in the paper. | | 2026-10-11 | | `horvath_2022_contact_tracing_apps` | Conjoint table (irw_conjoint). choice = 'Which app would you be more likely to install on your phone?' (Mobile App 1 / Mobile App 2), forced choice, no opt-out. Restriction observed: a decentralised app always has 'Not stored', 'Notify users directly', and no contacts or location uploaded, and those storage/purpose levels never occur with a central database. Attribute order randomized and recorded (attrpos_). Preview/pre-launch responses and screened-out respondents dropped as by the authors; N = 1,504 matches the article. | | 2026-10-07 | | `hrbkova_2024_conversation_partners` | Conjoint table (irw_conjoint). rating is a 0-100 feeling thermometer for each stranger; choice is which of the two the respondent would prefer to talk to (forced, no opt-out); wording paraphrased from the article (Czech wording not deposited). Unrestricted design, attribute order fixed. 321 tasks with unsaved attribute text and 14 with no answer dropped; 1,032 respondents and 10,250 rows as in the article. Qualtrics IDs re-keyed; attribute and covariate text in Czech. | | 2026-10-08 | | `huber_2017_dating_profiles` | Conjoint table (irw_conjoint_2). Ten fictitious online-dating profiles, one per task, each rated on six questions (interest in dating, attractiveness, long-term interest, would respond, shares values, would be friends; 4-5 point scales, higher = more favourable; question wordings are the authors' labels, not in the deposit). No choice or opt-out. Age and height levels depend on the profile's sex; ideology '--' means no ideology listed. The Easton-Holbein single-factor same-party studies in the deposit are not included. | | 2026-10-11 | | `huber_2020_car_policy_instruments` | Conjoint table (irw_conjoint). rating = support (1) or reject (0) each proposal; choice = which of the two proposals (initiative, counter-proposal) the respondent prefers, forced, one per task (wording paraphrased from the appendix example). Two attributes (instrument, implementation year) plus a between-respondent frame (trial_frame). Level shares unequal (observed). N 2,034. | | 2026-10-09 | | `huebscher_2018_austerity_de` | Conjoint table (irw_conjoint_2). Germany: five screens of two policy packages (one conservative, one leftist party; small/large cuts to health, pensions, education, unemployment benefits, infrastructure; tax increase none/across the board/for the wealthy only); choice = would vote for this package (one of the two, or neither: opt-out). Levels are codebook labels, not the on-screen wording; vote question wording not in the deposit. Rows with no package or vote dropped; cov_survey_weight is NA on about a fifth of rows. | | 2026-10-11 | | `huebscher_2018_austerity_es` | Conjoint table (irw_conjoint_2). Spain: five screens of two policy packages (one conservative, one leftist party; small/large cuts to health, pensions, education, unemployment benefits, infrastructure; tax increase none/across the board/for the wealthy only); choice = would vote for this package (one of the two, or neither: opt-out). Levels are codebook labels, not the on-screen wording; vote question wording not in the deposit. Rows with no package or vote dropped; cov_survey_weight is NA on about a fifth of rows. | | 2026-10-11 | | `huebscher_2018_austerity_it` | Conjoint table (irw_conjoint_2). Italy: five screens of two policy packages (one conservative, one leftist party; small/large cuts to health, pensions, education, unemployment benefits, infrastructure; tax increase none/across the board/for the wealthy only); choice = would vote for this package (one of the two, or neither: opt-out). Levels are codebook labels, not the on-screen wording; vote question wording not in the deposit. Rows with no package or vote dropped; cov_survey_weight is NA on about a fifth of rows. | | 2026-10-11 | | `huebscher_2018_austerity_pt` | Conjoint table (irw_conjoint_2). Portugal: five screens of two policy packages (one conservative, one leftist party; small/large cuts to health, pensions, education, unemployment benefits, infrastructure; tax increase none/across the board/for the wealthy only); choice = would vote for this package (one of the two, or neither: opt-out). Levels are codebook labels, not the on-screen wording; vote question wording not in the deposit. Rows with no package or vote dropped; cov_survey_weight is NA on about a fifth of rows. | | 2026-10-11 | | `huebscher_2018_austerity_uk` | Conjoint table (irw_conjoint_2). UK: five screens of two policy packages (one conservative, one leftist party; small/large cuts to health, pensions, education, unemployment benefits, infrastructure; tax increase none/across the board/for the wealthy only); choice = would vote for this package (one of the two, or neither: opt-out). Levels are codebook labels, not the on-screen wording; vote question wording not in the deposit. Rows with no package or vote dropped; cov_survey_weight is NA on about a fifth of rows. | | 2026-10-11 | | `huidobro_2026_coalition_partner` | Outcome: "If you could choose between two partners to form a government coalition with the following two leaders, which one would you choose?" forced choice; four 1-5 agreement ratings (similar preferences, easy communication, capacity to govern, trustworthy) on task 2 only. Task and profile are inferred from row order (checked), so profile may not be screen position. The paper reports 979 mayors and 3,324 observations; the author's log uses 905 respondents and 3,526 rows, which this table reproduces; only grouped covariates are kept. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `hultinrosenberg_2022_voting_rights` | Conjoint table (irw_conjoint). Outcomes: choice of one of two hypothetical persons (question wording not deposited; the authors treat it as which should have voting rights) and a 1-7 rating of each (the authors code 5-7 as granting and 1-3 as denying the vote; anchors not deposited). 22-43 respondents per pair skipped the choice (choice NA on both profiles, ratings kept); 29 profile ratings are missing. Level text is the authors' value labels. The 19 respondents Qualtrics did not flag as good completers are dropped, as in the authors' analysis (N = 980). | | 2026-10-08 | | `hutchinson_2023_household_arrangements` | Conjoint table (irw_conjoint). choice = which household arrangement the respondent would personally prefer to be in (paraphrase; forced choice, no opt-out); rating = 'How FAIR do you think Couple A's/B's situation is?' 1 = very unfair to 5 = very fair. Attribute levels are the phrases inserted into a vignette whose template is not deposited. 8 tasks whose levels were not saved and 24 tasks with no choice dropped (1,938 -> 1,935 respondents); the article reports nearly 2,000 parents. | | 2026-10-07 | | `huwyler_2025_side_income` | Conjoint table (irw_conjoint_2). One Tweet image per respondent; rating_trust = trustworthiness of the MP and rating_vote = whether the respondent can imagine voting for the MP, both 0-10 (wording paraphrased from the article; question order randomized, trial_item_order). Restrictions: only 12 of 16 board type x seats x income combinations were shown (no pro bono company seats, no 5 seats with 20% income), crossed with MP gender. English level descriptions stored; IP, GPS and free text dropped; 16,241 respondents vs 14,100+ complete cases in the article. | | 2026-10-09 | | `ibbett_2024_pa_compliance_scenario` | Conjoint table (irw_conjoint_2). Factorial scenario, one per respondent: peers' compliance (high/low) x enforcer corruption (high/low); rating is willingness to follow protected-area rules, raw 1 very willing to 5 very unwilling (higher = less willing), rating_attitude is agreement that the rules are fair (1 agree to 5 disagree). Indonesia and Tanzania pooled (cov_country); 10 rows with no willingness answer omitted. Factor wording is paraphrased from the article. N is 454 against the paper's 432 complete-case analysis. | | 2026-10-14 | | `ibbett_2024_pa_sanction_fairness` | Conjoint table (irw_conjoint_2). Factorial survey: each respondent rated the fairness of 8 of 32 vignettes (full 4x2x2x2: crime, offender origin, offender power, sanction) on a 3-point scale as stored (1 Unfair, 2 Neutral, 3 Fair; the article's 5-point scale condensed by the authors); no opt-out. Indonesia and Tanzania are pooled (cov_country); the crime levels differ by country (logging Indonesia only, grazing Tanzania only). Attribute level text taken from the article (Table 1); the deposit stores short tags. Task is row order; the paper's exclusions of incomplete or non-varying respondents are not applied, so N differs from the paper's analysed sample. | | 2026-10-14 | | `igarashi_2022_skilled_immigrant` | Conjoint table (irw_conjoint_2). One text vignette per respondent (Japan, 2020): skill of a Chinese applicant (education and former job) x extra information (none, wants a convenience-store job, chronic asthma and health insurance, teen shoplifting arrest); choice = 1 if the respondent would approve his visa, 0 if not (single-profile accept/reject, opt-out). Attribute level text taken from the English vignettes in the authors' RIETI discussion paper appendix (B.2); the deposit stores group codes only, so levels are paraphrased from that translation. 'No extra information' is coded (not shown). | | 2026-10-14 | | `igarashi_2025_discrimination_judgement` | Conjoint table (irw_conjoint). Single-vignette rating: how likely the described scenario is to be considered discriminatory, 1 (almost certainly not discriminatory) to 6 (almost certainly discriminatory); no choice, no opt-out. Each target attribute (gender, nationality, sexuality, education) appears once per respondent in random order and the displayed phrases are fixed by target x group status, so combinations are restricted; attr_ text is the article's English translation of the Japanese phrases; between-respondent definition arm in trial_definition; open-ended reasons dropped; 4,000 respondents as in the paper. | | 2026-10-09 | | `igarashi_2025_minority_candidates` | Conjoint table (irw_conjoint_2). rating_vote_willingness (Q13.5) and rating_favorability (Q13.4) are 5-point and stored raw; the authors recode them as 6 - x so that higher = more willing/favourable, so raw 1 is the most willing/favourable end (inferred, answer labels and question wording are not in the deposit). Levels are the Japanese texts shown; name and origin are linked (American name only with American origin, Chinese name only with Chinese origin). Cov_attention_pass follows the authors' three manipulation checks; N = 3,000 as in the article. | | 2026-10-11 | | `iilonga_2023_algebra_word_problems` | Each of six word problems is scored on four Polya domains (item Q<k>_D<j>; item_family = Q<k>): D1 understanding, D2 planning, D4 looking back scored 0-2, D3 carrying out the plan 0-4. The rubric itself is not published; these ranges are the observed ones. One Q4_D4 score of 3 (the only 3 on any D4 item) was dropped as an entry error. Learners who did not attempt a problem keep the source's zero scores. | | 2026-10-07 | | `incerti_2021_china_japan_cn` | Conjoint table (irw_conjoint). Three forced choices between two Japan-China policy packages: 'please compare the two sets of policy proposals and choose which proposal you prefer between the two' (appendix English version; respondents saw Chinese); no opt-out. Resource development by Japan only never shown with Chinese sovereignty and by China only never with Japanese sovereignty. Per-profile score and 'favours' items left out (wording not deposited). 3,639 respondents (21,834 rows) under the authors' sample rules. | | 2026-10-09 | | `incerti_2021_china_japan_jp` | Conjoint table (irw_conjoint). Three forced choices between two Japan-China policy packages: 'please compare the two sets of policy proposals and choose which proposal you prefer between the two' (appendix English version; respondents saw Japanese); no opt-out. Resource development by Japan only never shown with Chinese sovereignty and by China only never with Japanese sovereignty. Per-profile score and 'favours' items left out (wording not deposited). 3,331 respondents (19,986 rows) under the authors' sample rules. | | 2026-10-09 | | `inouye_2026_sympathies_2023` | Conjoint table (irw_conjoint). choice: "Which side do you sympathize more with: Group A or Group B?" forced, no opt-out. The two groups always differ on every binary attribute (mirror images); task 11 repeats task 1 with sides swapped (trial_repeat_of = 1), used by the authors for reliability correction. All 1,253 panel completes kept; cov_in_authors_sample marks the authors' 1,198 (their kept-id list; the deposited fraud-flag rule alone does not reproduce it). IP, geolocation and panel ids dropped. | | 2026-10-07 | | `inouye_2026_sympathies_2025` | Conjoint table (irw_conjoint). choice: "Which side do you sympathize more with: Group A or Group B?" forced, no opt-out; choice_similar_palestinian / choice_similar_israeli: which group is more similar to actual Palestinian groups (pairs 1-5) / the actual Israeli government (pairs 6-10), asked later about the same pairs, kept as asked (the authors flip the Palestinian item). The two groups always differ on every binary attribute (mirror images); task 11 repeats task 1 with sides swapped (trial_repeat_of = 1), used by the authors for reliability correction. All 1,104 panel completes kept; cov_in_authors_sample marks the authors' 1,001 (their kept-id list; the deposited fraud-flag rule alone does not reproduce it). IP, geolocation and panel ids dropped. | | 2026-10-07 | | `ishima_2026_perceived_ideology` | Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical Japanese politicians (7 attributes, 8 tasks, fixed attribute order); forced choice of the politician perceived as more left / liberal / progressive, the wording assigned between respondents (trial_group); no opt-out. Level text is the authors' English codes, not the Japanese text shown; surname, IP and geolocation dropped. | | 2026-10-11 | | `islam_2026_ethical_distance_canada` | Conjoint table (irw_conjoint). Six candidate pairs per respondent, one per ethical issue (trial_issue); the issue position is attr_ethical_position. choice: "If you had to choose between them, which of these two candidates would you vote for?" (forced, no opt-out); rating: 0-100 rating of each candidate, wording not deposited, stored raw. Task numbers follow the deposit's column order, not a recorded display order. Canadian French-language respondents' levels in French. Panel IDs re-keyed; free-text answers dropped. | | 2026-10-08 | | `islam_2026_ethical_distance_us` | Conjoint table (irw_conjoint). Six candidate pairs per respondent, one per ethical issue (trial_issue); the issue position is attr_ethical_position. choice: "If you had to choose between them, which of these two candidates would you vote for?" (forced, no opt-out); rating: 0-100 rating of each candidate, wording not deposited, stored raw. Task numbers follow the deposit's column order, not a recorded display order. Panel IDs re-keyed; free-text answers dropped. | | 2026-10-08 | | `izquierdo_lopez_2026_eosp` | Each pupil (id) was rated once by one teacher (rater); 125 teachers rated 160 pupils. | | 2026-10-11 | | `jager_2023_security_liberties` | Conjoint table (irw_conjoint). One text vignette per respondent; rating = support for the surveillance measure (1 do not support it at all - 10 strongly support it), plus rating_efficacy, rating_restrict, rating_concern_personal, rating_concern_society (1-10; wording paraphrased from the preregistration). No choice, no opt-out; full factorial with equal-probability assignment; attribute levels are the author's labels because the German vignette text is not deposited; control levels are (not shown). N = 2,045 as in the article; post-treatment fear item dropped. | | 2026-10-09 | | `jamal_2022_fdi_tunisia` | Conjoint table (irw_conjoint). choice = "Which investment do you prefer most? Investment A or Investment B?" (forced); rating = support for each investment, 1-7, reversed from the source so 7 = very strongly support (source 1 = very strongly support, 7 = very strongly oppose). Industry, who the jobs are for, skill and prayer-facility sentences were sometimes omitted, a randomized level coded (not shown). Attribute text is the English stored text; the field language is not documented. Qualtrics ResponseIds dropped; one empty respondent and respondents with no outcome dropped (1,497 vs 1,502 reported). | | 2026-10-07 | | `jankowski_2020_hiring_civil_servants` | Conjoint table (irw_conjoint). choice: "Wenn Sie sich zwischen den beiden Bewerber/innen entscheiden müssten, wen würden Sie in den öffentlichen Dienst einstellen?" (which of two applicants would you hire into the public service), forced choice, no opt-out. Level text is the German text displayed (raw Qualtrics export); attribute row order randomized per respondent, kept as attrpos_*. Respondents' e-mail addresses in the raw export dropped. 234 respondents (the authors' analysis set; 7 did not finish the survey but completed all tasks). Level shares unequal for family origin (Germany ~50%) and disability (yes ~20%); no restriction documented. | | 2026-10-08 | | `jankowski_2020_hiring_private` | Conjoint table (irw_conjoint). choice: "Wenn Sie sich zwischen den beiden Bewerber/innen entscheiden müssten, wen würden Sie in den öffentlichen Dienst einstellen?" (which of two applicants would you hire into the public service), forced choice, no opt-out. Level text is the authors' English labels (German text shown, not deposited for this sample; 'x1 - ' style prefixes on None removed); attribute order not recorded. 639 respondents. Level shares unequal for family origin (Germany ~50%) and disability (yes ~20%); no restriction documented. | | 2026-10-08 | | `jankowski_2020_hiring_students` | Conjoint table (irw_conjoint). choice: "Wenn Sie sich zwischen den beiden Bewerber/innen entscheiden müssten, wen würden Sie in den öffentlichen Dienst einstellen?" (which of two applicants would you hire into the public service), forced choice, no opt-out. Level text is the authors' English labels (German text shown, not deposited for this sample; 'x1 - ' style prefixes on None removed); attribute order not recorded. 401 respondents. Level shares unequal for family origin (Germany ~50%) and disability (yes ~20%); no restriction documented. | | 2026-10-08 | | `jankowski_2023_strategic_voting` | Conjoint table (irw_conjoint_2). Factorial vignette (Germany, 2021): one description of Campact's strategic-voting campaign per respondent, with a justification clause and a Green-party-support clause each randomly shown or '(not shown)' (2x2); rating = evaluation of the campaign, 1 'very bad' to 10 'very good'. Level text translated from the author's preprint. Both factors describe the evaluated campaign, so it is included although the author calls them treatments. | | 2026-10-14 | | `jankowski_2024_list_ranking_austria` | Conjoint table (irw_conjoint). Two-stage design with one task of nine aspirant profiles: rating_deselected = ticked under "Wen würden Sie nicht nominieren?" (exactly three of nine; 1 = deselected), rating_list_position = list position 1-6 assigned to each kept profile in stage 2 (1 = top; NA for deselected). Each respondent saw only one of the two dual-candidacy attributes (regional vs state/federal list); the other is (not shown). Attribute order randomized per respondent and recorded. Qualtrics metadata, free text and the years respondents joined their party or executive dropped. | | 2026-10-08 | | `jarbin_2020_mfq_parent` | Parent report; id is the child. One 10-year-old had both self-report forms and keeps the child form in jarbin_2020_mfq_self. | | 2026-10-10 | | `jeanlouis_2025_plankton_dmv` | Conjoint table (irw_conjoint). DCE: choice = "Please select your preferred scenario" among Scenario 1, Scenario 2 and a business-as-usual scenario that displays its levels (profile 3; no separate opt-out). Fixed design of 16 cards answered by everyone, row order differs by card (attrpos_). Five countries pooled, English level text from the English questionnaire. Answer codes 1-3 inferred as the three columns. | | 2026-10-09 | | `jeannet_2026_migration_policy` | Outcomes: forced choice between Policy A and Policy B (no opt-out) and a 1-7 support rating for each policy (1 = extremely oppose, 7 = extremely support). Attribute text is the English codebook master with "[pipe: COUNTRY]" for the respondent's country (shown translated); dimension order was randomized per respondent but not recorded, and level frequencies vary by card though they balance overall. Six countries pooled with cov_country; weight in cov_survey_weight. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `jenke_2021_eyetracking_candidates` | Conjoint table (irw_conjoint). choice = which candidate the participant would vote for for President (forced choice among 2 or 3 candidates, no opt-out). Six within-subject blocks (trial_block: 5/8/11 attributes x 2/3 candidates; trial_block_iter = order); attributes left out of a block are '(not shown)', attribute rows randomized per subject-block (attrpos_). Race, military service and religion levels have unequal design weights. 122 participants as in the article; 11 trials without eye-tracking data flagged trial_fixation_data = 0. | | 2026-10-08 | | `jensen_2014_incentive_credit` | Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: a US state's tax incentive package (greater, or equal or less, than other states') x whether the state receives a 1,000-job plant; rating = effect on the likelihood of voting for the governor, 1-5 with 5 = much more likely (authors' inverted coding, stored raw; 3 = vote choice not altered). Exact vignette wording from the article; 1,974 respondents = the paper's N. | | 2026-10-14 | | `jensen_2021_city_development_plans` | "For each comparison, please indicate which of the two plans you prefer" (forced choice between Development Plan 1 and 2, no opt-out, 5 comparisons). Attributes fully randomized; dimension order random per respondent but not recorded; levels naming the city or state were filled with the respondent's metro area. YouGov weight kept; county/ZIP, free text and other survey items dropped; N = 7,800 matches the paper. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `jeune_2024_research_use` | Conjoint table (irw_conjoint). Factorial vignette survey (French text); rating = 'À quel point vous semble-t-il probable que Mx. A utilise la recherche qui l'intéresse dans ces conditions ?' slider -5 (extrêmement improbable) to 5 (extrêmement probable); no choice. Fixed blocked design of 16 vignettes; display order randomized but not saved (task = questionnaire position). 440 respondents = paper's 100 pilot (cov_pilot) + 340 main. | | 2026-10-09 | | `jones_2020_cme_complexity` | Raters who were not logged in to Zooniverse (rater_logged_in = 0, 67,558 decisions) are identified per browser session, so one person may appear as several raters. The two sessions show the same 1,111 images, the second after brightness equalisation; analyse them separately or model session. | [#2692](https://github.com/ben-domingue/irw/issues/2692) | 2026-10-01 | | `jones_2022_gay_teacher` | Conjoint table (irw_conjoint). Factorial vignette (one news story per respondent, 6 randomized factors, full factorial). rating: "If you were deciding this case, would you" definitely side with the school .. definitely side with the gay teacher, 4 points stored as the author's 0-1 recode (0, 1/3, 2/3, 1); rating_similar: 7-point slider 'Not at all similar' .. 'Extremely similar' stored as k/7; rating_lgb_jobs: support for laws protecting LGB people against job discrimination, 4 points recoded 0-1 (a general policy item asked after the story). No opt-out. Levels are the author's labels, not the story text. Straight respondents only; the four survey waves are pooled and not identified; the author's LGB-rights index is dropped. | | 2026-10-09 | | `jones_2025_reform_legitimacy_s2` | Conjoint table (irw_conjoint). Single-profile text vignettes of an unnamed authoritarian country (five tasks); five ratings, 1 'Very unfavorable' to 7 'Very favorable', toward the country, trading with it, cutting off relations, a boycott and visiting (paraphrase; for cut/boycott higher = more hostile). Levels are the fragments inserted into the (undeposited) vignette template; expert and critics sentences are '(not shown)' in no-reform vignettes. N = 1,891. | | 2026-10-09 | | `jones_2025_reform_legitimacy_s3` | Conjoint table (irw_conjoint). Single-profile text vignettes of an unnamed authoritarian country (four tasks); five ratings, 1 'Very unfavorable' to 7 'Very favorable', toward the country, trading with it, cutting off relations, a boycott and visiting (paraphrase; for cut/boycott higher = more hostile). Levels are the fragments inserted into the (undeposited) vignette template; expert and critics sentences are '(not shown)' in no-reform vignettes. N = 1,317 (124 non-raters dropped; Saudi Arabia task and region fragment left out). | | 2026-10-09 | | `joo_2026_descriptive_rep` | Conjoint table (irw_conjoint). Six candidate pairs; choice = candidate the respondent would vote for, with an 'I would not vote' opt-out (1,056 tasks, all from 176 respondents, have no chosen profile); choice_concerns = candidate who addresses the respondent's concerns (one per task); rating_contact, rating_donate, rating_trust = 1-4 per candidate, wording/anchors/direction not deposited, stored raw. 96 rows with no outcome dropped; N = 1,043 with outcomes of 1,051 (article n = 1,051). | | 2026-10-09 | | `jordan_2020_burnout` | 169 of the 539 respondents are excluded from Q12_8, Q12_9 and Q12_10. In the source file their answers to those items carry the next block's response labels, which looks like a column shift in the survey export, so which answer belongs to which item cannot be recovered. Their answers to the other items are kept, so those items have 539 respondents and these have 370. | [#2098](https://github.com/ben-domingue/irw/issues/2098) | 2026-09-30 | | `jordan_2020_pss10` | 169 of the 539 respondents are excluded from Q13_8, Q13_9 and Q13_10. In the source file their answers to those items carry the next block's response labels, which looks like a column shift in the survey export, so which answer belongs to which item cannot be recovered. Their answers to the other items are kept, so those items have 539 respondents and these have 370. | [#2098](https://github.com/ben-domingue/irw/issues/2098) | 2026-09-30 | | `jordan_2020_resilience` | 169 of the 539 respondents are excluded from Q14_4, Q14_5 and Q14_6. In the source file their answers to those items carry the next block's response labels, which looks like a column shift in the survey export, so which answer belongs to which item cannot be recovered. Their answers to the other items are kept, so those items have 539 respondents and these have 370. | [#2098](https://github.com/ben-domingue/irw/issues/2098) | 2026-09-30 | | `jost_2024_zambia_participation` | Conjoint table (irw_conjoint). Factorial vignette, one per respondent, read by phone. rating: likelihood of spending a day campaigning / attending the meeting, 1 very likely to 4 not likely at all (raw; lower = more willing); rating_leader_sanction and rating_community_sanction: likelihood of being treated better or worse depending on participation (same 1-4); rating_enjoy: 1 very much to 4 not at all; rating_authority_support: would the authority support this candidate, 1 yes, 2 no. Don't know/refuse = NA. Co-ethnicity and co-partisanship are relative levels (the piped-in group/party not saved). 11 respondents dropped (missing levels or no answers). | | 2026-10-09 | | `jung_2024_value_conjoint` | Conjoint table (irw_conjoint_2). Six single-profile rated people per respondent (no choice, no opt-out): favorability 1-7 (1 = very positive, so lower is more favourable), neighbor 1-5 (1 = very happy to have as a neighbor, lower more favourable) and house 1-5 (1 = not comfortable at all to ... look after the respondent's house, higher more favourable), all raw; 98 'skipped on web' set to NA. Value statement text is the Stata label, completed from the online appendix for 37 cut-off labels; question wordings for neighbor and house are paraphrases; N (863) not checked against the paper. | | 2026-10-11 | | `jurado_2022_brexit_deal_germany` | Conjoint table (irw_conjoint). choice: "Which Brexit agreement would you choose if the decision were yours?" (forced, no opt-out). Two cross-sectional waves pooled, wave in trial_wave. Labels are the authors' English versions of the German text. Level frequencies are unequal (randomization weights undocumented). profile assumes the source's ab = 0 is agreement A. | | 2026-10-07 | | `jurado_2022_brexit_deal_spain` | Conjoint table (irw_conjoint). choice: "Which Brexit agreement would you choose if the decision were yours?" (forced, no opt-out). Two cross-sectional waves pooled, wave in trial_wave. Labels are the authors' English versions of the Spanish text. Level frequencies are unequal (randomization weights undocumented). profile assumes the source's ab = 0 is agreement A. | | 2026-10-07 | | `kacovsky_2022_imi` | 1 = the highest feeling, 7 = the lowest (reversed from the usual IMI direction), as stored. | | 2026-10-10 | | `kage_2019_elect_politician` | Conjoint table (irw_conjoint). choice = which of the two candidates the respondent would personally prefer to see elected to the House of Representatives (forced choice, no opt-out); rating = an evaluation of each candidate, 1-7, wording and anchors not documented (stored raw). 9 attributes, author English labels; attribute order randomized per task, not recorded. Weight kept; N 1,611 as in the article. | | 2026-10-09 | | `kage_2019_promote_politician` | Conjoint table (irw_conjoint). choice = which of the two politicians the respondent would personally prefer to see promoted as LDP leader (paraphrase; forced choice, no opt-out); rating = an evaluation of each politician, 1-7, wording and anchors not documented (stored raw). 6 attributes, author English labels; attribute order randomized per task, not recorded. Weight kept; N 1,611 as in the article. | | 2026-10-09 | | `kahraman_2026_bai` | BAI stored 1-4 (the 0-3 form shifted by one). | | 2026-10-10 | | `kahraman_2026_zbi` | The deposit has 19 ZBI columns (ZBI-1..ZBI-19), not the 22-item form. | | 2026-10-10 | | `kamalov_2026_emigrant_aid` | Conjoint table (irw_conjoint). choice = which of two hypothetical Russian individuals the respondent preferred to help (article paraphrase; verbatim wording not deposited), forced choice, no opt-out, 2 tasks. Attributes independently randomized (article); level text is the authors' English labels, respondents saw Russian. Authors' derived binaries and Russian-text fields dropped; N = 2,036 matches the article. | | 2026-10-07 | | `kamwendo_2021_detergent_ranking` | Conjoint table (irw_conjoint_2). Ranking of ten fixed detergent profiles (5 attributes) in a single task: rating = the rank given to each profile, 1 = most preferred to 10 = least preferred, stored raw (not a choice column). Levels were read from the poster image shipped with the deposit; the sheet columns are read as rank per profile (inferred), and the 13 rows that are not a full permutation of 1-10 are dropped. Fixed design, all respondents ranked the same 10 profiles; the respondent sample is not described. | | 2026-10-12 | | `kang_2024_police_priority_exp1` | Conjoint table (irw_conjoint). choice = which of two cases the investigator would prioritize (the authors' 'case prioritization'; verbatim Korean wording not deposited), forced choice between pairs; task is inferred from profile order. Attribute levels are the authors' English coding labels, not the Korean text shown. Occupation levels are unequal (accuser occupation unemployed on 14% of profiles vs low-SES 44%, high-SES 42%). cov_gender is the investigator's gender (female/male). Unanswered pairs omitted. | | 2026-10-07 | | `kang_2024_police_priority_exp2` | Conjoint table (irw_conjoint). choice = which of two cases the investigator would prioritize (the authors' 'case prioritization'; verbatim Korean wording not deposited), forced choice between pairs; task is inferred from profile order. Attribute levels are the authors' English coding labels, not the Korean text shown. Occupation levels are unequal (accuser and accused occupation unemployed on about 8% of profiles each). cov_gender is the investigator's gender (female/male). Unanswered pairs omitted. | | 2026-10-07 | | `kantorowicz_2025_copyright_uk_pros` | Conjoint table (irw_conjoint). choice: which of Contract A and Contract B the respondent would accept (paraphrase; forced choice, no opt-out). Five 2-level attributes randomly set each task; attribute row order randomized per respondent (not deposited). Task and profile are inferred from row order (task order assumed). Deposited sample = the authors' analysis sample after attention-check and completion exclusions; N matches the article. | | 2026-10-07 | | `kantorowicz_2025_copyright_uk_public` | Conjoint table (irw_conjoint). choice: which of Contract A and Contract B the respondent would accept (paraphrase; forced choice, no opt-out). Five 2-level attributes randomly set each task; attribute row order randomized per respondent (not deposited). Task and profile are inferred from row order (task order assumed). Deposited sample = the authors' analysis sample after attention-check and completion exclusions; N matches the article. | | 2026-10-07 | | `kantorowicz_2025_copyright_us_pros` | Conjoint table (irw_conjoint). choice: which of Contract A and Contract B the respondent would accept (paraphrase; forced choice, no opt-out). Five 2-level attributes randomly set each task; attribute row order randomized per respondent (not deposited). Task and profile are inferred from row order (task order assumed). Deposited sample = the authors' analysis sample after attention-check and completion exclusions; N matches the article. | | 2026-10-07 | | `kantorowicz_2026_rights_computers` | Conjoint table (irw_conjoint). choice: which of two laptop computers the respondent chose (wording not deposited), forced choice, one chosen in every task. Up to 6 tasks; task and profile inferred from the deposit's row layout (no id columns). Privacy levels are the authors' short labels. Attribute order was randomized but not kept. 2,238 respondents; derived covariate dummies dropped. | | 2026-10-08 | | `kantorowicz_2026_rights_laundry` | Conjoint table (irw_conjoint). choice: which of two laundry machines the respondent chose (wording not deposited), forced choice, one chosen in every task. Up to 6 tasks; task and profile inferred from the deposit's row layout (no id columns). Privacy levels are the authors' short labels. Attribute order was randomized but not kept. 2,239 respondents; derived covariate dummies dropped. | | 2026-10-08 | | `kanwal_2024_csr` | Four respondents were dropped from the deposit's 349: two pairs of adjacent rows were identical on all 45 items while reporting different demographics, so neither row of a pair could be trusted. The 26 items span four CSR dimensions (employees, community, environment, customers). | | 2026-10-04 | | `kanwal_2024_goc` | Four respondents were dropped from the deposit's 349: two pairs of adjacent rows were identical on all 45 items while reporting different demographics, so neither row of a pair could be trusted. | | 2026-10-04 | | `kanwal_2024_gsv` | Four respondents were dropped from the deposit's 349: two pairs of adjacent rows were identical on all 45 items while reporting different demographics, so neither row of a pair could be trusted. | | 2026-10-04 | | `kanwal_2024_wpeb` | Four respondents were dropped from the deposit's 349: two pairs of adjacent rows were identical on all 45 items while reporting different demographics, so neither row of a pair could be trusted. | | 2026-10-04 | | `kao_2021_covid_compliance` | Conjoint table (irw_conjoint). Single vignette per respondent (task = profile = 1). choice: "Would you be likely to comply with the instructions from {Authority}?" Yes = 1 / No = 0 (accept/reject, so opt_out = yes). Ratings: others' compliance 1 None .. 4 Most; whether the authority has the right to ask, whether the action lowers one's risk, whether the authority would monitor (1 Yes, 0 No). Don't know/refuse -> NA. The religious-leader authority arm (1,510 respondents) is excluded because the piped level text was not saved; 6 respondents with no outcome answer omitted. | | 2026-10-08 | | `kao_2023_committee_gender` | Conjoint table (irw_conjoint_2). One audio vignette per respondent (committee all-male or gender-balanced x decision supported or rejected, domestic-violence penalties; 2 x 2); ratings 1-4: rightness of the decision for all citizens, women and men, trust in the committee, fairness of the process and likely public acceptance. No opt-out. Jordan respondents who heard the other issue area are excluded (issue text not recoverable); 4,501 respondents; the preprint reports 4,754 overall. | | 2026-10-11 | | `kao_2024_tax_reform` | Conjoint table (irw_conjoint). choice: which of two tax-reform proposals the respondent prefers (paraphrase: 'choose which is most preferred'), forced, no opt-out; 6 tasks, 4 attributes (tax type, rate 1-20%, government services, institutional political influence), levels as the authors' value labels (respondents saw Chinese). 57 respondents with no choices omitted; the article's screened sample = cov_elite_t10 non-missing. Taiwan shadow-case file not built (tasks not recoverable). | | 2026-10-09 | | `kao_2025_new_residents` | Three forced choices with no opt-out on the same pairs: admission, national health insurance, mother tongue (authors' labels; exact wording unconfirmed). Every attribute has a "No Info" level, and Ukraine appears only with ethnicity "No Info". Rows with any outcome kept (2,075 respondents vs the authors' complete-case 2,013). | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `karlsson_2023_climate_risk` | The six items are three targets (self, other people, non-human life) each asked in two wordings; the paired items (1/4, 2/5, 3/6) correlate .93 to .98 but are not duplicates. | [#2513](https://github.com/ben-domingue/irw/issues/2513) | 2026-09-28 | | `karpowitz_2023_masculinity_2016` | Conjoint table (irw_conjoint_2). Paired choice of a woman vs a man state-delegate candidate with self-descriptions; choice = the vote (forced, no opt-out); rating_likeable / rating_competent are 1-4 per candidate (direction read from the authors' difference scores, 4 = extremely well). Profile 1 = woman, 2 = man because screen order was not saved; the woman's five profile types and gender-appeal sentence and a respondent-level equality prime (trial_) are randomized, the man is always Masculine. Respondents with a missing vote dropped (none). | | 2026-10-11 | | `karpowitz_2023_masculinity_2018` | Conjoint table (irw_conjoint_2). Paired choice of a woman vs a man state-delegate candidate with self-descriptions; choice = the vote (forced, no opt-out); rating_likeable / rating_competent are 1-4 per candidate (direction read from the authors' difference scores, 4 = extremely well). Profile 1 = woman, 2 = man because screen order was not saved; both candidates' four self-presentation types are crossed, with unequal cell sizes and unsaved wording variants. Respondents with a missing vote dropped (559). | | 2026-10-11 | | `karsen_2025_mooc_motivation` | Eleven prefix blocks (SE, SJ, SD, SI, MC, GA, GS, CC, IM, EM, MCO) shipped as one table because the record documents none of the codes. | | 2026-10-10 | | `kasparek_2025_audit` | Scored exactly as the deposit codes it: audit_1 uses 0, 1, 2, 3 and 5 (the codebook gives '5 = 4 or more times a week'; 4 is unused), audit_2-8 0-4, audit_9-10 0/2/4. audit_2-10 are blank for the 156 respondents who answered 'Never' to audit_1. | | 2026-10-07 | | `kassas_2021_covid_mitigation` | The deposit's variable label for 'cleaning' repeats the 'delivery' question (a label slip); the item ships under its column name without item text. | | 2026-10-11 | | `kato_2025_candidates_jan22` | Conjoint table (irw_conjoint_2). choice = 'more desirable as a Single Member District member of the House of Representatives' (3 preference tasks), choice_expect = 'more likely to win' (3 expectation tasks; translated), paired forced choice, no opt-out; block order randomized (trial_block_order), task is display order. Party type (ruling/opposition) follows from party. Dropped 1 respondent with unassigned attributes, unanswered tasks and Qualtrics IDs; row counts (10,606 / 10,612) match the article. | | 2026-10-09 | | `kato_2025_candidates_mar22` | Conjoint table (irw_conjoint_2). choice = 'more desirable' (4 preference tasks), choice_expect = 'more likely to be elected' (4 expectation tasks; translated), for the House of Representatives or a Municipal Council (trial_election_level); paired forced choice, no opt-out; block order randomized, task is display order. A 35-year-old is never shown with 5 terms (observed). Dropped 2 respondents with unassigned attributes and unanswered tasks; 38,222 rows match the authors' script but the article's per-arm counts (37,828 in total) do not. | | 2026-10-09 | | `kawakami_2026_visa_england` | Conjoint table (irw_conjoint). choice: "Should this person receive a work visa to come to the UK?" yes = 1, no = 0; single-profile accept/reject, so no-visa is the opt-out. 8 profiles per respondent, 5 attributes in a fixed-order table; restrictions and level probabilities not documented. Article says 646 respondents; deposit has 643. | | 2026-10-08 | | `kawakami_2026_visa_japan` | Conjoint table (irw_conjoint). choice: "この申請者は日本に来るための就労ビザを発給されるべきでしょうか？" yes = 1, no = 0; single-profile accept/reject, so no-visa is the opt-out. 4 profiles per respondent, 5 attributes in a fixed-order table; restrictions and level probabilities not documented. Level text is English (authors' labels); respondents saw Japanese. 27 unanswered tasks omitted. | | 2026-10-08 | | `kennedy_2022_policy_algorithms` | Conjoint table (irw_conjoint). Five pairs of recidivism-forecasting algorithms; choice = which algorithm profile the respondent chose (Choice 1 / Choice 2, forced; codebook paraphrase); rating = rating of each algorithm, 1 'Definitely Would Not Use' to 7 'Definitely Would Use'. Three samples pooled as in the paper (cov_sample: MTurk 2018, Lucid March 2019, Lucid April 2019). Tasks without answers dropped (30 MTurk respondents have only tasks 1-3; 70 Lucid respondents answered none): 1,750 respondents. | | 2026-10-09 | | `kennedy_2023_hr_partner_s1` | Conjoint table (irw_conjoint_2). Five tasks of two hypothetical countries; choice = preferred of the pair (forced, no opt-out), rating_acceptable = 0/1 judgement of each profile. The deposit stores dummy-coded levels with short tags and no codebook, so attribute names and levels are the stored stems and tags, not displayed text; question wording is not documented. Tasks with a missing choice were dropped. | | 2026-10-14 | | `kennedy_2023_hr_partner_s2` | Conjoint table (irw_conjoint_2). Five tasks of two hypothetical countries; choice = preferred of the pair (forced, no opt-out), rating_acceptable = 0/1 judgement of each profile. The deposit stores dummy-coded levels with short tags and no codebook, so attribute names and levels are the stored stems and tags, not displayed text; question wording is not documented. Tasks with a missing choice were dropped. | | 2026-10-14 | | `kennedy_2025_judicial_candidates` | choice = "Which candidate do you prefer?" (forced, no opt-out; tasks 4 and 7 carry a leftover "Which job offer do you prefer?" export label); rating = "How qualified is Candidate 1/2 to be a state high court justice?" 4-point, reversed from source so 4 = most qualified (as the authors did; anchor wording not in deposit). 15 tasks from 2 respondents lack a choice; attribute order randomized per respondent, no documented restrictions. IP, location, Prolific/Qualtrics IDs, free text and timers dropped; article N not checked. | [#2887](https://github.com/ben-domingue/irw/issues/2887) | 2026-10-07 | | `kenwick_2022_military_advisers` | Conjoint table (irw_conjoint_2). One vignette per respondent (president stays out / engages / engagement succeeds / fails, after consulting a military or a civilian adviser; 2 x 4 between-subject); rating = approval of the decision, 1 = strongly disapprove to 4 = strongly approve, plus trust in president and adviser (0/1) and credit or blame (-1 adviser, 0 both, 1 president, asked only in the Success or Failure cells). No opt-out. The deposit has no questionnaire, so level text is the codebook's condition names (paraphrase); N = 2,394 respondents, all consented. | | 2026-10-11 | | `kenwick_2024_border_militarization` | Conjoint table (irw_conjoint). Survey 1 (conjoint only). choice: Policy A vs Policy B (forced; wording not deposited, paraphrase). Attribute text from the embedded display strings; bots and attention-check failures were removed by the authors before deposit; 776 respondents. | | 2026-10-09 | | `kenwick_2024_border_treatments` | Conjoint table (irw_conjoint). Survey 2: respondents randomized to one of four status retention/loss vignettes (trial_arm) before the conjoint. choice: Policy A vs Policy B (forced; wording not deposited). Unanswered tasks omitted (3,228 of 3,240 respondents kept); authors' manipulation and attention checks kept as cov_ flags; free text not included. | | 2026-10-09 | | `kermen_2019_anxiety` | The record names the State-Trait Anxiety Inventory; the file's 20-item anxiety block (sum 'kaygi') is coded 0-3, not the STAI's usual 1-4, and the deposit does not say which STAI form. Items shipped as coded. | | 2026-10-07 | | `kermen_2019_attentional_control` | Items e1, e2, e3, e6, e7, e8, e11, e12, e15, e16, e20 (suffix y, the ACS's reverse-keyed items) are stored reversed by the source: the deposit's scale sum is the plain sum of the stored values and these items correlate positively with the rest. Two typo cells (e9 = 22, e18 = 32) dropped. | | 2026-10-07 | | `kertzer_2021_resolve` | Conjoint table (irw_conjoint). Eight paired country tasks; choice = which country the respondent sees as more likely to stand firm (forced, no opt-out); rating = estimated likelihood the country will stand firm, 0-100 (paraphrase; 5 missing). The United States as a profile is always a democracy with a powerful military and never faces itself (observed); the paper's main analyses drop tasks involving the US, kept here. Entropy-balancing survey weight kept; Qualtrics IDs re-keyed; N = 2,003. | | 2026-10-09 | | `khan_2018_bidr_im` | IM1-IM19 value labels run 1 = 'Very True' to 5 = 'Not True', the reverse of SDE (1 = 'Not True'); codes are kept as stored. The deposit has 19 IM columns, not the BIDR's 20. | | 2026-10-06 | | `khan_2023_inspiration` | No codebook: block membership was recovered by regressing the deposit's composite means on the items (exact equal-weight fits). | | 2026-10-09 | | `khan_2023_perceived_motivation` | No codebook: block membership was recovered by regressing the deposit's composite means on the items (exact equal-weight fits). | | 2026-10-09 | | `khan_2023_perceived_purpose` | No codebook: block membership was recovered by regressing the deposit's composite means on the items (exact equal-weight fits). | | 2026-10-09 | | `kim_2019_trade_policy_firms` | Conjoint table (irw_conjoint). choice: which of two trade policies the firm would prefer ("indique cuál es la política que su empresa preferiría"), forced choice, no opt-out. Up to 5 tasks; 5 attributes with Spanish level text, attribute order randomized per respondent (attrpos_*). 8 of 1,596 tasks dropped (7 profiles not matchable to the display record, 1 with no choice). 325 respondents; the paper uses the 214 tradable-sector firms. | | 2026-10-08 | | `kim_2022_immigration_covid` | Conjoint table (irw_conjoint). Outcomes: forced choice between two immigrants (the authors' 'supporting admission'; question wording not deposited) and rating_entry, a per-profile entry decision stored in Korean and coded here 1 = entry banned, 2 = entry allowed with a mandatory 14-day quarantine, 3 = entry allowed. Profile position was not recorded, so profile numbers are arbitrary within a task. Attribute levels are the authors' English labels, including their typos ('volumne', 'communitation'). Randomization restrictions and attribute order are undocumented. | | 2026-10-08 | | `kim_2024_carbon_tax_korea` | Conjoint table (irw_conjoint). Five paired carbon-tax proposals; choice = which proposal ('탄소세 정책안 A/B') the respondent chose (forced, no opt-out; wording not deposited). Korean level text. Only the no-information control arm is in the deposit (845 respondents); the paper's information-treatment arm is not. No covariates. | | 2026-10-09 | | `kim_2024_class_gender_uk` | Conjoint table (irw_conjoint). Paired MPs (respondent's party), 3 tasks; choice: 'which of the two MPs would you prefer to have as your MP?' plus 14 further forced picks (relate, represent five groups, advocate on seven issues and on the respondent's top issue); no opt-out, skipped items NA. Five attributes, order randomized per respondent. Only respondents who answered at least one conjoint question (975 of 992 attention-check passers). | | 2026-10-09 | | `kim_2024_class_gender_us` | Conjoint table (irw_conjoint). Paired candidates (respondent's party primary), 3 tasks; choice: 'which of the two candidates would you be more likely to support in this [party]'s congressional primary?' plus 14 further forced picks (relate, represent five groups, advocate on seven issues and on the respondent's top issue); no opt-out, skipped items NA. Five attributes, order randomized per respondent. Only respondents who answered at least one conjoint question (865 of 939 attention-check passers). | | 2026-10-09 | | `kim_2025_isi` | resp is stored 1 to 5, as in the deposit (1 = not at all ... 5 = very severe). The published Insomnia Severity Index scores each item 0 to 4. To get the standard 0-28 total and use its cutoffs, subtract 1 from each of the seven items (ISI_1_1, ISI_1_2, ISI_1_3, ISI_2 to ISI_5) before summing. | [#2100](https://github.com/ben-domingue/irw/issues/2100) | 2026-09-30 | | `kim_2025_karrot_attitude` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. The attitude/behaviour split is not confirmed by the data: q45 and q50 correlate slightly more with the behaviour items, so it rests on the article's order and block sizes alone. | | 2026-10-09 | | `kim_2025_karrot_behav_intent` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. The attitude/behaviour split is not confirmed by the data: q45 and q50 correlate slightly more with the behaviour items, so it rests on the article's order and block sizes alone. | | 2026-10-09 | | `kim_2025_karrot_econ_motiv` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. | | 2026-10-09 | | `kim_2025_karrot_env_motiv` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. | | 2026-10-09 | | `kim_2025_karrot_interact_motiv` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. | | 2026-10-09 | | `kim_2025_karrot_reput_motiv` | The deposit has no codebook and its q-numbers do not match the article's questionnaire numbering. The 20 Likert columns were assigned to the article's Table 1 constructs in order (economic 3, environmental 5, interaction 3, reputation 2, attitude 4, behaviour 3); the correlation matrix confirms the first four blocks. | | 2026-10-09 | | `kiratli_2024_intervention_turkey` | Conjoint table (irw_conjoint). choice: "Sizce Türkiye hangi kriz için askeri birlik göndermelidir?" between two military-operation proposals, 6 tasks per respondent; forced choice, no neither option. 6 attributes, fully randomized per the article. Turkish respondents saw Turkish text; levels stored in English from the article's Table 1. Qualtrics IDs re-keyed; no survey weight; N matches the article. | | 2026-10-08 | | `kiratli_2024_intervention_us` | Conjoint table (irw_conjoint). choice: "If you had to choose only one, which of these two military operations would you approve?" between two military-operation proposals, 6 tasks per respondent; forced choice, no neither option. 6 attributes, fully randomized per the article. Qualtrics IDs re-keyed; no survey weight; N matches the article. | | 2026-10-08 | | `kiratli_2025_alliance_cohesion` | Conjoint table (irw_conjoint_2). Five single-profile vignettes per respondent (task = file row order, inferred). rating = agreement with the proposed US response 0-10; rating_alliance_favorability 0-10; rating_contribution 1 = increase, 2 = keep the same, 3 = decrease US contributions (stored raw; lower = more favourable). No choice or opt-out. Act and response levels use the article's template wording mapped from the deposit's short labels. 1,993 respondents are kept; cov_attention_pass = 1 for the 1,502 the article analyses. | | 2026-10-11 | | `kirkland_2018_candidates_mturk` | Conjoint table (irw_conjoint). choice = 'Which of these two candidates do you prefer?' (forced choice, no opt-out); rating = 'On a scale from 0 to 100, how competent do you think these candidates would be as mayor?' (0-100, higher = more competent). Attributes fully randomized; whether party is shown is randomized per election (trial_election; attr_party = '(not shown)' when not shown). Profile order is inferred from row order; attribute order was randomized and not recorded. MTurk worker ID dropped and respondents re-keyed; one worker with two sets of contests and 4 contests left with one row by the authors' na.omit dropped. | | 2026-10-07 | | `kirkland_2018_candidates_yougov` | Conjoint table (irw_conjoint). choice = 'Which of these two candidates do you prefer?' (forced choice, no opt-out); rating = 'On a scale from 0 to 100, how competent do you think these candidates would be as mayor?' (0-100, higher = more competent). Competence is missing for some candidates (127 respondents never rated). Attributes fully randomized; whether party is shown is randomized per election (trial_election; attr_party = '(not shown)' when not shown). Profile order is inferred from row order; attribute order was randomized and not recorded. Up to 5 contests per respondent (some fewer in the deposit). | | 2026-10-07 | | `kjoller_2025_council_positions` | Conjoint table (irw_conjoint). choice = "Forestil dig, at du tilbydes følgende to poster som medlem af byrådet. ... Hvilken post vil du vælge?" (Mulighed A/B, forced, no opt-out). Levels are the Danish text shown; the position wording varies with the respondent's municipality type (Borgmester/Rådmand/Udvalgsformand). Simple randomization; position always first, the other three attributes in a per-respondent random order (attrpos_*). 2,141 respondents incl. partial completers; cov_in_authors_sample = the authors' complete-and-on-time filter, 1,939 respondents vs 1,938 in the article. | | 2026-10-07 | | `kobayashi_2021_hk_demands_wave1` | Conjoint table (irw_conjoint). choice = which of two government-response packages the respondent preferred (wording not deposited; paraphrase), forced choice, one chosen per pair. Attribute levels are the authors' English labels, not necessarily the text shown; randomization restrictions undocumented. Wave 1 of 2 (pre-NSL); cov_party_id is the respondent's political camp. Qualtrics IDs re-keyed. | | 2026-10-07 | | `kobayashi_2021_hk_demands_wave2` | Conjoint table (irw_conjoint). choice = which of two government-response packages the respondent preferred (wording not deposited; paraphrase), forced choice, one chosen per pair. Attribute levels are the authors' English labels, not necessarily the text shown; randomization restrictions undocumented. Wave 2 of 2 (post-NSL); cov_party_id is the respondent's political camp. Qualtrics IDs re-keyed. | | 2026-10-07 | | `kodapanakkal_2020_banking` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `kodapanakkal_2020_citizen_scores` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `kodapanakkal_2020_crime_prevention` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `kodapanakkal_2020_employment` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `kodapanakkal_2020_healthcare` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `kodapanakkal_2020_investigations` | Conjoint table (irw_conjoint). choice = "If you have to make a choice between these versions of the technology, which one would you choose?" (version A / version B / Neither of these: opt-out, both 0); rating = moral evaluation of each version, 0 (morally unacceptable) to 100 (morally acceptable). Fixed design of 12 pairs shown in random order (task = pair number, display order not recorded); trial_framing = between-subjects description (Brand New / New but Used / Used). Respondent ids are per technology; persons cannot be linked across the six tables. | | 2026-10-09 | | `koelln_2023_party_legitimacy` | Conjoint table (irw_conjoint_2). Between-subject vignette, one profile per respondent: a fictitious Danish party's ideology (5 levels) x democratic stance (2), plus a control coded (not shown) on both. Six ratings: three legitimacy items stored 0-1 (higher = more legitimacy) and three 1-7 / 1-5 perception items (threat and ideology reversed in the source); wording not in the deposit; levels are the authors' labels, not the displayed text. | | 2026-10-11 | | `kolanska_stronka_2026_climate_anxiety` | The deposit has no codebook; the three tables (CA_, GSES_, COPE_) were identified from the record title and the sum columns. A handful of isolated out-of-range entries were dropped (one respondent's 0s on five CA and three GSES items, GSES_10 = 5, COPE_1 = 4 twice, COPE_24 = 8). | | 2026-10-07 | | `kolcava_2021_coregulation` | Conjoint table (irw_conjoint_2). choice = referendum vote between Proposal A and B; choice_responsiveness = which design is more likely to yield a policy most citizens want; choice_effectiveness = which is more likely to reach the sector's target (forced choices, no opt-out, 5 pairs). Two green-economy contexts (finance, pesticides) pooled in trial_context with the sector named neutrally in the decision attribute; the separate vignette experiment is not built; N against the article not checked. | | 2026-10-11 | | `kolcava_2022_supply_chain_belgium` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (EN 246, BE 797, FR 957 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_canada` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (EN 1942, FR 59 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,001 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_france` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (FR 1993, EN 7 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_germany` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (DE 1953, EN 47 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_italy` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (IT 1986, EN 14 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_japan` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (JA 1996, EN 4 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_korea` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (KO 1736, EN 264 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_netherlands` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (NL 1912, EN 88 respondents by Qualtrics language), so each level has several spellings. The Dutch middle strictness level as stored repeats "geen verplichte informatie" (no mandatory information), a translation error in the instrument. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_spain` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (ES 1966, EN 34 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_switzerland` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (DE 1298, IT 49, FR 539, EN 114 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_uk` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,000 matches the deposit. | | 2026-10-08 | | `kolcava_2022_supply_chain_us` | Conjoint table (irw_conjoint). Five paired tasks of supply-chain law proposals (company size covered, reporting strictness, enforcement), fixed attribute order. choice: "Suppose you had to decide between the two proposals in a vote today. Would you rather support Proposal A or Proposal B?" (forced, no opt-out); rating: "If you were presented Proposal A [B] today, would you be opposed or in favour of it?" 1 = in favour, 0 = opposed (source 1/2). trial_info_treatment = randomized information block before the conjoint (Control / Con Market / Pro Market). Level text is in the respondent's survey language (EN 1997, ES 5 respondents by Qualtrics language), so each level has several spellings. The separate post-conjoint probe experiment (ExpB) and Qualtrics IDs are dropped; N 2,002 matches the deposit. | | 2026-10-08 | | `koliev_2026_unfccc_reform` | Conjoint table (irw_conjoint). choice = "Which proposal do you prefer (Proposal 1 or Proposal 2)? Please choose one.", forced choice, no opt-out; 6 tasks, attribute row order randomized per respondent (attrpos_). The status-quo marker " (SQ)" stored in the data was not displayed and is stripped; the free-text 'other group' answer and Qualtrics ids are dropped. N = 151 matches the paper; the state-representative presidency AMCEs (0.010, 0.011) reproduce. | | 2026-10-09 | | `koob_2026_protest_dimensions` | Conjoint table (irw_conjoint). Four forced choices on the same pairs of protests, names from the authors' files (exact wording not deposited): choice_sympathy (more sympathy), choice_policy (support for the protesters' policy demands), choice_disorder (more of a public disorder), choice_deserving (protesters more deserving); exactly one profile chosen per task. Attribute levels are the displayed sentence fragments; attribute order fixed. Q3.1 and race kept as unlabelled codes. Respondent count not checked against the paywalled article. | | 2026-10-09 | | `KoreanNursing_Park_2017` | The item codes (X0, X1, X1.1, X0.1, ... X0.39) carry no meaning: they were generated from one examinee's 0/1 answers when the headerless source file was read with a header row. They are kept so existing references still work. The codes follow the exam's question order: the first code, X0, is question 1, and the last, X0.39, is question 50 (the published item text uses the same codes). | [#2096](https://github.com/ben-domingue/irw/issues/2096) | 2026-09-30 | | `krajewski_2025_criminal_opportunities` | Conjoint table (irw_conjoint). Outcome: choice of one of two criminal opportunities (Stata label "Criminal opportunity chosen"; question wording not deposited); forced choice, no opt-out. Task and profile are inferred from row order; 18 unanswered tasks omitted. Attribute text comes from Stata value labels; randomization restrictions undocumented. The deposit has 1,000 respondents, the article reports 1,023. | | 2026-10-07 | | `krakowski_2023_informal_justice` | Conjoint table (irw_conjoint). Factorial vignette, 2 per respondent (task 1-2, profile 1). Outcomes: "how do you think he/she/they tried to resolve this problem?" and "how do you think people should resolve this issue?", each with 3 options (state authority / a religious cleric (.dta label: a local imam) / by mediation according to tradition and Kanun), stored as 0/1 indicators per option (rating_tried_*, rating_should_*). Levels are the English questionnaire wording parsed from the stored Albanian text; resources and community wording depends on the dispute and each respondent heard one civil and one criminal dispute. Per-dispute Ns and Table A1 coefficients reproduce. | | 2026-10-08 | | `krakowski_2023_violence` | Conjoint table (irw_conjoint). Factorial vignette, 3 per respondent (task 1-3, each event once, order randomized; profile 1). rating: "what do you think of his decision to ...? His decision was:" 1 Completely expected .. 5 Completely unexpected (stored raw, higher = more unexpected). Job wording depends on the event. This experiment is not analysed in the article; the questionnaire dates the protest 2016 while the stored Albanian text says 2014. | | 2026-10-08 | | `kramer_2023_efct` | EFCT halves A (efct_1-84) and B (efct_85-168) were counterbalanced, so for any one item about half the respondents answered it before training (wave 1) and half after (wave 2). | | 2026-10-04 | | `kramon_2025_democratic_innovations` | Conjoint table (irw_conjoint). Two pairs of district decision processes; choice = 'Which of these processes would you prefer?' (forced); rating_fair / rating_democratic = 'Thinking of process A [B], how fair / democratic would you say it is?' 1-7 (7 = extremely fair / democratic). Levels are the author's short labels; level shares unequal (status quo 35%, no endorsement 43%) though described as independently randomized. 1,043 respondents as in the paper. | | 2026-10-09 | | `kreps_2020_covid_vaccine` | Conjoint table (irw_conjoint). Respondents could choose vaccine A, B or neither: 2,071 of 9,855 tasks are opt-outs, with choice = 0 on both profiles. rating is the 1-7 likelihood of taking each vaccine (7 = extremely likely). FDA attribute text shortened; consent item and the authors' derived dummies dropped. | | 2026-10-07 | | `kreps_2021_vaccine_incentives` | Conjoint table (irw_conjoint). Seven single vaccine profiles per respondent; choice = would choose to get this vaccine (1) vs not be vaccinated (0), a single-profile accept/reject (opt_out yes); rating = 'How likely or unlikely would you be to get the vaccine described above?' 1 extremely likely to 7 extremely unlikely, raw (lower = more likely). Attribute order randomized once per respondent; restrictions not documented; EUA level text truncated in the deposit. N = 1,096. | | 2026-10-09 | | `kreps_2023_ai_trust` | Conjoint table (irw_conjoint). Single-profile ratings: rating_support ("Do you support the use of AI under these circumstances?") and rating_trust ("Do you trust the use of AI under these circumstances?"), both 1-5 (1 = strongly disagree, 5 = strongly agree); no choice, no opt-out. Five tasks per respondent in the data although the article says four. Platform IDs, ZIP code, unlabelled Lucid profile codes and the separate follow-up mediation experiment dropped. | | 2026-10-08 | | `krewson_2021_judicial_philosophy` | Conjoint table (irw_conjoint). choice is a forced choice between two nominees (24 unanswered tasks blank) and rating a 1-5 rating of each, wording and anchors not deposited (5 = favourable inferred from agreement with choice). Randomization restricted: party constrains ideology (no Democratic conservative or Republican liberal), and race and law-school levels have unequal weights. Task and profile inferred from row blocks; YouGov IDs re-keyed; cov_survey_weight is the YouGov weight. | | 2026-10-07 | | `krishnarajan_2022_pledges_dk` | Conjoint table (irw_conjoint). One statement per task, no choice: rating = 'Would you consider this an election pledge?' 1 = yes, 0 = no; rating_accountable = should the sender be held accountable (1 yes / 0 no); rating_priority, rating_can_act, rating_cannot_check on 1-5 scales (higher = more agreement / more able). 'Don't know' is NA. Statements without a numeric goal, time horizon or condition store '(not shown)' for that clause. The deposit has 1,800 in the deposit (appendix C1: 2,031); the paper does not say what removed the difference. | | 2026-10-08 | | `krishnarajan_2022_pledges_uk` | Conjoint table (irw_conjoint). One statement per task, no choice: rating = 'Would you consider this an election pledge?' 1 = yes, 0 = no; rating_accountable = should the sender be held accountable (1 yes / 0 no); rating_priority, rating_can_act, rating_cannot_check on 1-5 scales (higher = more agreement / more able). 'Don't know' is NA. Statements without a numeric goal, time horizon or condition store '(not shown)' for that clause. The deposit has 2,050 in the deposit (appendix C1: 2,214); the paper does not say what removed the difference. | | 2026-10-08 | | `krishnarajan_2022_pledges_us` | Conjoint table (irw_conjoint). One statement per task, no choice: rating = 'Would you consider this an election pledge?' 1 = yes, 0 = no; rating_accountable = should the sender be held accountable (1 yes / 0 no); rating_priority, rating_can_act, rating_cannot_check on 1-5 scales (higher = more agreement / more able). 'Don't know' is NA. Statements without a numeric goal, time horizon or condition store '(not shown)' for that clause. The deposit has 1,745 in the deposit (appendix C1: 2,032); the paper does not say what removed the difference. | | 2026-10-08 | | `kruse_2025_expected_contribution` | Conjoint table (irw_conjoint). Three profiles per task: choice = which participant the respondent thought would give the most to the common pool (forced, one of three); rating = expected contribution 0-10 in units of DKK 100. Between-respondent information conditions (trial_information) add socio-economic, cultural-value, norm-compliance, all, or placebo attributes; attributes not shown are (not shown). Danish-named profiles were born in Denmark 90% of the time, Middle Eastern-named 50%. attr_name_origin is the author's grouping of attr_name. All profiles are men. cov_survey_weight is the YouGov weight. | | 2026-10-07 | | `kuecuekbalaban_2017_selftest` | Conjoint table (irw_conjoint_2). Factorial survey, 6 text vignettes per respondent (the application purpose is the vignette number), rating = intention to use the test, a 1-100 visual analogue scale (1 = certainly not, 100 = most certainly), no choice. Six attributes; the level codes carry no labels in the deposit and were mapped to the English text of Supplementary Table S2 by matching per-level means (inferred but unique). Analysis/feedback levels depend on the test setting (restrictions). N = 208 respondents; the 18 other outcome items per vignette are not named in the deposit and are dropped. | | 2026-10-11 | | `kustov_2026_populism_mobilization` | Conjoint table (irw_conjoint_2). choice = candidate chosen in the pair; pref = 0 (abstain, neither candidate) is the opt-out, 4,516 of 28,016 tasks; 8 tasks of 2 candidates, 5 attributes (job, office, and three statements: pluralism, moralism, policy, each drawn from a bank of 32/32/64 sentences). The pluralism statement is blank on 3.0% of candidates, only with non-populist moralism, and is coded (not shown) as an inference (the deposit is silent). The vote question wording is not in the deposit; respondent count (3,502) matches the authors' replication notebook. | | 2026-10-11 | | `kyrkopoulou_2026_immigrants_greece` | Conjoint table (irw_conjoint). Forced choice between Individual A and B (question wording not in the deposit), no opt-out, 5 tasks; Greek attribute text as exported (feminine/masculine forms are separate levels); gender and skin tone stored as the authors' English codes, apparently shown by photo. Observed restriction: 'very good Greek & English' only with residence in Greece without a permit. All 2,007 respondents kept (authors' Study 2 drops 3 with no weight); cov_survey_weight = deposited raking weight weight_new. | | 2026-10-09 | | `landgrave_2025_pac_contributions` | Conjoint table (irw_conjoint). One candidate vignette per respondent (2 x 3 x 2: PAC contributions, prior experience, gender); rating_vote 'Vote likelihood', rating_donate 'Donation likelihood', rating_trust 'Candidate trustworthiness' (paraphrase; wording not deposited), 7-point stored raw with 1 = Very likely/trustworthy and 7 = Very unlikely/untrustworthy. No opt-out; candidate party is matched to respondent party (trial_candidate_party); skipped answers NA; zip codes and panel ids dropped; N = 1,000 as in the authors' log. | | 2026-10-09 | | `lara_2024_equality_mobility` | Conjoint table (irw_conjoint). choice: forced choice between hypothetical Society A and Society B (wording not deposited; paraphrase), no opt-out; 4 tasks, 4 attributes each drawn from 9 values (\) incomes for bottom 20%/middle 60%/top 20%, % upward mobility); screens also showed the computed average income and 90/10 ratio (not stored). trial_framing distant/nearby society (80/20, per respondent). Source A/B column labels are swapped against the README; positions inferred. Full sample of 1,458 kept with cov_preferred_sample for the article’s 1,249 and both raked weights. | 2026-10-09 | |
larsen_2024_apartment_opposition |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: building type x floor area x height. rating is support for the project (1 definitely against .. 7 definitely for, 8 = don’t know); four further 1-5 agree/disagree beliefs (6 = don’t know) are stored raw. Height wording differs for retail premises; the home-price belief item is missing for 3,303 respondents. Levels are Danish text. | 2026-10-11 | |
larsen_2025_welfare_chauvinism |
Conjoint table (irw_conjoint). Four paired tasks; choice = the social-assistance reform proposal the respondent preferred (paraphrase; wording not deposited), forced choice, no opt-out. Levels are the author’s English labels of a Danish instrument (from the replication Rmd); attribute row order randomized per respondent (attrpos_); survey weight kept. N = 1,287 (not checked against the article). | 2026-10-09 | |
larsen_2026_local_preservation |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (Denmark, register sample, 2023): project type (9) x size (3) x height (4) x distance band to the respondent’s home (0-10 km in 10 bands, ‘(not shown)’ for the 10% control with neither map nor text), plus map and distance-in-text arms. Seven ratings: support (1-7, 8 = don’t know) and six agreement items (1-5, 6 = don’t know); home-value and rent items are NA outside owners/renters. Attribute text taken from the article’s Table 1; the deposit stores short tags. Survey weight kept. | 2026-10-14 | |
laterzo_2023_crime_candidates |
Conjoint table (irw_conjoint). Five forced choices between two candidates (wording not deposited; codebook: whether the profile was chosen), 6 binary attributes as the codebook’s English wording. Task/profile rebuilt from row blocks; on-screen position is not recorded, so profile numbers are arbitrary and outcome-independent (sorted by attribute text); 100 Argentine tasks with missing levels dropped; Argentina and Brazil pooled (cov_country); N = 2,688 (paper analyses 2,301 after an ideology filter). | 2026-10-09 | |
laterzotingley_2024_crime_national |
Conjoint table (irw_conjoint). Paired candidate choice, five tasks, forced choice (exactly one chosen per pair); question wording not deposited. Pairs rebuilt from the stacked file; screen position and task order not recorded (profile numbering is outcome-independent but arbitrary). Level text = codebook labels, displayed wording undocumented. National sample, 1,499 respondents. | 2026-10-09 | |
laterzotingley_2024_crime_students |
Conjoint table (irw_conjoint). Paired candidate choice, five tasks, forced choice (exactly one chosen per pair); question wording not deposited. Pairs rebuilt from the stacked file; screen position and task order not recorded (profile numbering is outcome-independent but arbitrary). Level text = codebook labels, displayed wording undocumented. Student sample, 713 respondents. | 2026-10-09 | |
laustsen_2023_candidate_names |
Conjoint table (irw_conjoint_2). One candidate description per respondent; rating = ‘hvad du synes om [navn] som politiker’, 0 = synes meget dårligt om, 10 = synes virkelig godt om, stored raw (authors rescale to 0-1). The 2 x 2 sex x ethnicity design is carried by four names, so attr_pronoun is fixed by attr_name; party piped from the respondent’s own vote intention. 1,116 respondents, as in the article. | 2026-10-09 | |
leal_2022_cyber_attribution |
Conjoint table (irw_conjoint). Five pairs of cyberattack scenarios shown side by side, no choice asked; rating_confidence: ‘how confident are you in the intelligence community’s assessment of who was responsible for each attack?’ 0 (not confident)-100 (complete confidence); rating_severity: 0 (not severe)-100 (very severe); rating_support_*: Would support (1) / Would not support (0) for six measures against foreign perpetrators or six sentences for US perpetrators (display logic, NA on the other kind). Perpetrator type x origin restricted (observed: terrorists only from Iran/Russia/US, government agencies never from the US). Authors’ sample: 1,233 after consistency and straight-lining drops; derived dummies dropped. | 2026-10-09 | |
lee_2022_ameritrust |
Measured after the polarization-perception manipulation; treat = 1 for either article condition, cov_condition names the arm. | 2026-10-10 | |
lee_2022_perpol |
Measured after the polarization-perception manipulation (treat); perceived polarization is the manipulated construct. | 2026-10-10 | |
lee_2024_skilled_immig_probability |
Conjoint table (irw_conjoint). choice: forced choice between two policy schemes (attitudes toward highly skilled immigrants), no opt-out, refusals NA; rating: 0-10 rating of each scheme (dictionary says 1-10; anchors not deposited). Profile order inferred from row order; nationality ‘mainland Chinese only’ never paired with a 40,000/year quota (observed); attribute text in Chinese; survey question wording and the paper’s N not verified. | 2026-10-09 | |
lee_2024_skilled_immig_volunteer |
Conjoint table (irw_conjoint). choice: forced choice between two policy schemes (attitudes toward highly skilled immigrants), no opt-out, refusals NA; rating: 0-10 rating of each scheme (dictionary says 1-10; anchors not deposited). Profile order inferred from row order; nationality ‘mainland Chinese only’ never paired with a 40,000/year quota (observed); attribute text in Chinese; survey question wording and the paper’s N not verified. | 2026-10-09 | |
lee_2025_constituent_feedback |
Conjoint table (irw_conjoint). Each official rated 2 feedback messages on each of 3 local issues: rating_informative (1 Not informative - 4 Very informative) and rating_influential (1 Not influential - 4 Very influential), wording paraphrased; no choice. Position is the same for both messages of an issue and constituent/argument levels are issue-specific; the order of issues is not recorded (task = issue). 573 officials with ratings (149 with no saved attributes dropped). | 2026-10-09 | |
lee_2025_immigration_nimby |
Conjoint table (irw_conjoint_2). choice = ‘If you had to choose between them, which of these two immigrants should be given priority to come to the United States to live?’ (paired forced choice, no opt-out, 8 pairs). Restrictions: high-skill jobs only with two years of college or more; persecution reason only for Iraq, Somalia, Sudan. Task and profile inferred from row order (verified one choice per pair); 8 respondents have 5-7 pairs; the deposit’s respondent-covariate file cannot be linked and is not used. N = 1,311 matches the article. | 2026-10-09 | |
leininger_2026_name_recognition |
Conjoint table (irw_conjoint_2). Paired design (Germany): 5 tasks of 2 candidates varying surname (40), party, gender, occupation (white/blue collar) and residence (city/town-village); choice = candidate chosen (no opt-out; NA in 627 tasks that have only ratings) plus 1-5 ratings of trustworthiness, local knowledge, district representativeness and election chances. Occupation and residence are stored as the authors’ tags (displayed wording not deposited); outcome wording is from the codebook only. Panel id and postal code dropped. | 2026-10-14 | |
leis_2025_english_motivation |
The paper gives the subscale item counts (6/6/3/5) but not which item is which, so the 20 items are one table; item codes q01-q20 follow the sheet’s column order (each code’s stem is in the script header and the item text). The sheet’s four trailing summary rows (Average, SD) are dropped. | 2026-10-09 | |
lelkes_2021_candidate_ideology_party |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: a fictitious candidate whose issue positions (none, moderate, extreme; conservative for a Republican, liberal for a Democrat) and party (shown or not) are randomized, as an in-party or out-party candidate (trial_candidate_relation). Outcome: feeling thermometer 0-100, higher = warmer; no opt-out. Ideology level text is paraphrased from the article appendix. 83 rows with no recorded condition dropped; 5,703 respondents vs 5,798 in the paper. | 2026-10-14 | |
lemi_2020_multiracial_candidates |
Conjoint table (irw_conjoint). choice = which of the two candidates the respondent would vote for (paraphrase; wording not in the deposit), forced choice, no opt-out, 10 tasks. Sample restricted by the author to single-race White, Black, Asian and Hispanic respondents (cov_race); 6 tasks without an answer are omitted. Covariate labels from the author’s do-file. | 2026-10-08 | |
lemke_2021_research_metrics |
Conjoint table (irw_conjoint). Two-step ranking of 3 fictitious publications per set: choice = picked first (no opt-out), rating_rank = 1 picked first, 2 picked second, 3 remaining (1 = most preferred; question wording not deposited). Fixed chained design of 20 sets, same for all; screen position not recorded. Set 1 or 2 is absent for 110 of 224 respondents (undocumented); a practice set and one task where the deposit’s two files disagree are dropped; 224 respondents vs 247 participating in the abstract. | 2026-10-09 | |
leon_rubio_2025_cdpe_antecedents |
The deposit codebook says reverse-keyed items are already inverted in the file, but the six items the scoring key marks as inverse (CDPE4, 6, 13, 18, 39, 56) correlate negatively with their own subscale siblings (e.g. CDPE56 ‘tasks are well planned’ vs CDPE35 ‘orders are vague and ambiguous’, r = -0.43), so they are stored as worded. Nothing is reversed in IRW. itemcov_subscale follows the deposit’s scoring key, with two typos fixed by matching wording (the first ‘45’ under Sobrecarga is item 46; ‘176’ is item 156). Header CDPE03 is item 93. | 2026-10-07 | |
leslie_2023_hypertension_care |
Conjoint table (irw_conjoint). Discrete choice experiment: “Please tell us which of the two facilities you would prefer to go to for your care.” between Health facility A and B, no opt-out (exactly one chosen per set). Fixed blocked design (35 sets in 5 versions plus one common set, trial_version); task is the design’s choice-set id because display order was not recorded. Level text from the published choice card; respondents saw English or Kannada. 1,085 respondents = the article; 24 skipped cards are absent. | 2026-10-08 | |
letki_2025_bottle_recycling |
Conjoint table (irw_conjoint_2). Factorial vignette (5 x 2 x 3, full factorial), one per respondent: rating = ‘How likely is it that you will be using these facilities to dispose of your glass bottles?’ on a 4-point card, 1 = very likely … 4 = not likely at all (lower = more likely; ‘Don’t know’ and ‘Refusal’ removed). Fourteen countries pooled (cov_country) on the English master text; the translated wording is not in the deposit. Survey weight kept; id = row of the source file. N = 21,435 respondents of 22,039. | 2026-10-11 | |
letki_2025_donation |
Conjoint table (irw_conjoint_2). Factorial vignette (3 x 4 x 3, full factorial), one per respondent: rating = ‘How likely is it that you will make THIS DONATION?’ on a 4-point card, 1 = very likely … 4 = not likely at all (lower = more likely; ‘Don’t know’ and ‘Refusal’ removed). Fourteen countries pooled (cov_country) on the English master text; the translated wording is not in the deposit. Survey weight kept; id = row of the source file. N = 21,346 respondents of 22,039. | 2026-10-11 | |
letki_2025_park_cleanup |
Conjoint table (irw_conjoint_2). Factorial vignette (3 x 3 x 3, full factorial), one per respondent: rating = ‘How likely is it that YOU will join this initiative to clean up a local park?’ on a 4-point card, 1 = very likely … 4 = not likely at all (lower = more likely; ‘Don’t know’ and ‘Refusal’ removed). Fourteen countries pooled (cov_country) on the English master text; the translated wording is not in the deposit. The incentive attribute is ‘(not shown)’ when omitted. Survey weight kept; id = row of the source file. N = 21,561 respondents of 22,039. | 2026-10-11 | |
letki_2025_tax_evasion |
Conjoint table (irw_conjoint_2). Factorial vignette (6 x 4 x 2 (penalty and desk levels can be omitted), full factorial), one per respondent: rating = ‘How likely it is that you will not declare this income to the Tax Office?’ on a 4-point card, 1 = very likely … 4 = not likely at all (lower = more likely; ‘Don’t know’ and ‘Refusal’ removed). Fourteen countries pooled (cov_country) on the English master text; the translated wording is not in the deposit. The two attributes shown as ‘–’ in the master questionnaire are ‘(not shown)’ when omitted. Survey weight kept; id = row of the source file. N = 19,304 respondents of 22,039. | 2026-10-11 | |
leuffen_2023_vaccine_solidarity |
Conjoint table (irw_conjoint_2). rating = Selection, support for sending vaccines to the described country, 7-point from 1 ‘I do not support at all’ to 7 ‘I fully support’ (paper wording; the deposit has no value labels for it); one vignette per respondent, no choice or opt-out. Attribute levels are the English DDI value labels (respondents saw German); the randomized information treatment is trial_info_treatment. The paper’s n = 4,022 (1,974 treatment, 2,048 control) reproduces. | 2026-10-11 | |
levis_2026_carbon_messenger |
Conjoint table (irw_conjoint_2). Seven single-vignette tasks per respondent (cost, policy sector, messenger, quoted wording; the 7th repeats the 1st, trial_repeat_of = 1); rating = support for the policy, integers 1-6 (higher = more support), no opt-out. Level text is the authors’ English codes for German/Polish vignettes; six per-vignette belief ratings are not included. | 2026-10-11 | |
levy_2022_perpetrator_candidates |
Conjoint table (irw_conjoint). choice: ‘¿Si tuviera que elegir, por cuál de los dos candidatos votaría?’ forced choice (NA where the task was skipped); rating: likelihood of voting for candidate A/B, 1 muy improbable to 5 muy probable. Levels in Spanish as displayed (from the codebook). Attribute order fixed. Rows with no answer omitted (1,528 of 1,589 respondents remain); the author’s rescaled rating dropped. | 2026-10-09 | |
li_2017_fdi_china |
Conjoint table (irw_conjoint). choice = “Of these two projects, which one do you prefer?” (forced, no opt-out); rating = support for each project if your city were evaluating it, 1 = no support at all to 7 = complete support (English translation of the Chinese questionnaire). Levels are the Chinese text shown; attribute order randomized per task (attrpos_*). Waves 1 (Nov 2014) and 2 (Oct 2015) pooled with trial_wave as the authors do. Task and profile are inferred from the stacked file layout; 4 tasks with unsaved attribute text dropped; 2,846 respondents (2,224 + 622 as reported). | 2026-10-07 | |
li_2021_china_intervention |
Conjoint table (irw_conjoint_2). One vignette per respondent (waves 1 and 2 pooled, trial_wave); rating = approval of how the Chinese government handled the situation, 1 strongly disapprove … 5 strongly approve (authors’ coding; wave 1 collapses a 7-point branching answer, kept in rating_7pt). Wave-2 direction follows the authors’ code but is contradicted by the questionnaire option order, so treat wave 2 with care. No choice or opt-out; id = row number; 3 rows with no answer omitted; location, income and free text dropped. | 2026-10-11 | |
li_2021_china_peacekeeping |
Conjoint table (irw_conjoint_2). One vignette per respondent; rating = evaluation of the Chinese government’s handling, ordered 1 disapprove, 2 neutral, 3 approve (the source stores 1 approve, 2 disapprove, 3 neutral and was recoded here; the strength follow-up questions are not kept). No choice or opt-out; id = row number; attribute codes decoded from the authors’ do-file; location, income and free text dropped. N 2,114 as in the paper’s wave 3. | 2026-10-11 | |
li_2024_sad |
resp is the raw true/false answer (0/1), not the scored SAD. To reproduce the published SAD total (mean 13.68, SD 7.47, r = 0.195 with expressive suppression), score sad_2, 5, 8, 10, 11, 13, 14, 16, 18, 20, 21, 23, 24 and 26 as 1 - resp and sum all 28. The supplement lists the other 14 items as reverse-scored, and following it gives the mirror score, 28 minus that total, with the correlations’ signs flipped. | #2119 | 2026-09-30 |
li_2024_spa |
resp is stored as answered (1 to 5), not reverse-scored. Items spa_1, spa_3, spa_7, spa_11, spa_13 and spa_14 are reverse-keyed; score them as 6 - resp before summing. The study’s supplement (s002.docx) gives this keying, and it reproduces the published mean (43.21), SD (10.52) and alpha (0.877). Summing the raw values gives SD 9.00 and alpha 0.805. | #2114 | 2026-09-30 |
li_2026_psmus |
Three pairs of rows are identical on all 23 deposited columns (demographics, items and composites); one row of each pair was dropped, so N is 785 rather than the paper’s 788. | 2026-10-04 | |
lichtin_2024_public_transport |
Conjoint table (irw_conjoint). One factorial policy proposal per respondent; choice = ‘Imagine you had to decide today only on this proposal in a popular vote, would you vote for or against it?’ (1 = vote for; voting against is the opt-out), rating = ‘Please indicate, how much do you support or oppose this proposal?’ 1 strongly oppose to 7 strongly support. All 54 designs kept (the article analyses 12 ‘plausible’ ones, n = 1,634); two instruction arms in trial_info_arm; raking weight kept; municipal accessibility score dropped. | 2026-10-09 | |
lichtin_2024_road_pricing |
Conjoint table (irw_conjoint_2). Choice between Proposal A and B in a referendum (no opt-out) and rating of each proposal on a 5-point favor/oppose scale stored raw (1 = strongly favor, 5 = strongly oppose; ‘Don’t know’ = NA). 5 tasks x 2 proposals, 5 attributes; levels were stored in the respondent’s language (de/fr/it/en) and are given here in English; the information-vignette arm is trial_info_treatment. Tasks without a recorded choice (6%) were dropped, leaving 3,359 of 3,448 respondents; the paper’s N was not checked. | 2026-10-11 | |
lindgren_2026_fuel_rationing |
Conjoint table (irw_conjoint). Five paired fuel-rationing proposals; choice = .dta ‘Forced choice’ (one proposal per round, no opt-out; wording not deposited, the authors analyse its complement as rejection); rating = 1 Strongly against … 5 Strongly in favor. Levels are English labels of a Swedish survey. All 3,247 kept; the authors’ analytic n of 3,175 excludes speeders and attention-check failures (flags kept); survey weight kept. | 2026-10-09 | |
lindgren_2026_meat_rationing |
Conjoint table (irw_conjoint). Five paired meat-rationing proposals; choice = .dta ‘Forced choice’ (one proposal per round, no opt-out; wording not deposited, the authors analyse its complement as rejection); rating = 1 Strongly against … 5 Strongly in favor. Levels are English labels of a Swedish survey. All 3,247 kept; the authors’ analytic n of 3,175 excludes speeders and attention-check failures (flags kept); survey weight kept. | 2026-10-09 | |
lipps_2025_border_measures |
Conjoint table (irw_conjoint). choice: forced choice of the border plan the respondent would more likely support (wording paraphrased in the article), no opt-out; rating: each plan on an 11-point (0-10) scale, wording and anchors not given. Profile position was not deposited, so profile numbers are arbitrary. Level text is the article’s English description of the German levels. No randomization restrictions stated. Respondent distance-to-border variables dropped (identifying). | 2026-10-07 | |
littmann_2026_terror_women_candidates |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent would prefer as the leading candidate for their party in the upcoming parliamentary election (article’s English; forced, no opt-out). Half the respondents first read that the terrorist threat to Denmark is significant (trial_terror_threat = Yes). Origin and education levels were weighted to match Danish candidates (Danmark ~79% of profiles). Source respondent IDs re-keyed. | 2026-10-08 | |
liu_2019_cadre_selection |
Conjoint table (irw_conjoint). choice: candidate selected (one per answered pair; wording not deposited); rating_qualified / rating_leadership / rating_implementation: 1-5 ratings of being qualified and suitable, leadership quality and task implementation (anchors not deposited; ratings correlate positively with selection). Restrictions and attribute order undocumented. 332 respondents in the deposit, 33 with no answers omitted. | 2026-10-07 | |
liu_2023_program_evaluation |
Conjoint table (irw_conjoint). Forced choice between two programs plus a 0-100 support slider per program; exact wording not deposited. No opt-out. All 1,154 respondents kept with check-pass flags; the authors analyse 885. Low-trust prime in trial_low_trust_prime; attribute order per respondent in attrpos_*. | 2026-10-07 | |
liu_2024_partisan_collaboration |
Conjoint table (irw_conjoint). choice is the official’s pick of which of two collaborative programs to adopt (wording not deposited; paraphrase), forced choice, no opt-out. Attribute levels are the author’s short labels (e.g. 750:250, Republicans), not the displayed wording. 43 of 772 officials answered fewer than 3 pairs (missing pairs absent, task numbers kept). City identifiers and city census figures dropped; party, gender, education and tenure kept as undocumented codes (_code). | 2026-10-08 | |
liu_2026_competitive_goals |
Item-to-dimension membership is the deposit’s Data_Dictionary_clean.xlsx; every deposited composite equals the mean of its dictionary items (checked). | 2026-10-09 | |
liu_2026_empathy |
Item-to-dimension membership is the deposit’s Data_Dictionary_clean.xlsx; every deposited composite equals the mean of its dictionary items (checked). | 2026-10-09 | |
liu_2026_moral_disengagement |
Item-to-dimension membership is the deposit’s Data_Dictionary_clean.xlsx; every deposited composite equals the mean of its dictionary items (checked). | 2026-10-09 | |
liu_2026_self_control |
Item-to-dimension membership is the deposit’s Data_Dictionary_clean.xlsx; every deposited composite equals the mean of its dictionary items (checked). | 2026-10-09 | |
liu_2026_social_alienation |
Item-to-dimension membership is the deposit’s Data_Dictionary_clean.xlsx; every deposited composite equals the mean of its dictionary items (checked). | 2026-10-09 | |
liu_2026_taiwan_ballots |
Conjoint table (irw_conjoint). Study 1 (Taiwan). Three picks per pair of ballots: choice_fair, choice_democratic, choice_efficient (‘which of the two ballots is more fair / democratic / efficient’, paraphrase from variable names; wording not deposited); forced pick, unanswered picks NA. 7 fixed ballot designs crossing language (4) and arrangement (stacked/separated; monolingual ballot has none, ‘(not shown)’). Levels are the authors’ English descriptions of Chinese ballot images. Demographic codes without codebook dropped. | 2026-10-09 | |
lobo_2023_parties_be |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw Dutch or French). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
lobo_2023_parties_de |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw German). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
lobo_2023_parties_es |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw Spanish). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
lobo_2023_parties_gr |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw Greek). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
lobo_2023_parties_ie |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw English). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
lobo_2023_parties_pt |
Conjoint table (irw_conjoint). choice: which of two hypothetical parties the respondent would vote for in national legislative elections (forced, no opt-out); rating: probability of voting for each party, 0 (definitely would not) to 10 (definitely would). 2 tasks, attribute order randomized per respondent; level text is the codebook’s English wording (respondents saw Portuguese). Respondents with blank attribute values dropped. One of six country tables (lobo_2023_parties_*). | 2026-10-08 | |
loewen_2019_depression_incumbent_ca |
Conjoint table (irw_conjoint). One task per respondent: choice = vote between the two candidates; choice_prepared/choice_trust/choice_character = which candidate is most prepared/trustworthy/of best character (all forced, no opt-out; wording paraphrased). Attribute levels are the text fragments piped into the vignette; the health manipulation applies only to the first candidate (second = (not shown)); parties always opposite within a pair. Empty D military sentence (375 of 779) stored as (not shown) (inferred). | 2026-10-09 | |
loewen_2019_depression_incumbent_us |
Conjoint table (irw_conjoint). One task per respondent: choice = vote between the two candidates; choice_prepared/choice_trust/choice_character = which candidate is most prepared/trustworthy/of best character (all forced, no opt-out; wording paraphrased). Attribute levels are the text fragments piped into the vignette; the health manipulation applies only to the first candidate (second = (not shown)); parties always opposite within a pair. | 2026-10-09 | |
loewen_2019_depression_open_ca |
Conjoint table (irw_conjoint). One task per respondent: choice = vote between the two candidates; choice_prepared/choice_trust/choice_character = which candidate is most prepared/trustworthy/of best character (all forced, no opt-out; wording paraphrased). Attribute levels are the text fragments piped into the vignette; the health manipulation applies only to the first candidate (second = (not shown)); parties always opposite within a pair. | 2026-10-09 | |
loewen_2019_depression_open_us |
Conjoint table (irw_conjoint). One task per respondent: choice = vote between the two candidates; choice_prepared/choice_trust/choice_character = which candidate is most prepared/trustworthy/of best character (all forced, no opt-out; wording paraphrased). Attribute levels are the text fragments piped into the vignette; the health manipulation applies only to the first candidate (second = (not shown)); parties always opposite within a pair. US leave text mapped from the Canadian wording (inferred). | 2026-10-09 | |
lomholt_2025_teacher_expectations |
Conjoint table (irw_conjoint_2). Factorial vignette: Danish teachers rate a described student (ethnicity, gender, parental involvement, engagement, subject, instructional context, mode of learning engagement) on the likelihood of being in the top 25% academically (rating, 1 = very unlikely to 6 = very likely) and on how far other teachers would agree (rating_others_agree, codebook 1-4 but codes 1-5 in the data). One vignette per task, up to 12 per teacher, no opt-out. Attribute level text is the codebook’s value labels (the deposit does not hold the wording shown); task order is inferred from row order; 102 of 403 respondents did not complete the survey. | 2026-10-14 | |
lopezgarcia_2023_mx_military_proposal |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: ‘a recent proposal [made by the PAN / the PRI] calls for the national government to take the military off the streets and find other ways to reduce crime [in collaboration with municipal authorities / the private sector / citizens and community organisations]’, control clauses ‘(not shown)’. rating_agree 1-5 (5 = strongly agree). Balanced 3x4 design. Level text translated from the questionnaire. | 2026-10-14 | |
lopezgarcia_2023_mx_police_recruits |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: ‘authorities hire more [female] municipal police officers [who previously worked as soldiers / in private security / were self-defence force members / human rights activists / migrants in the US]’, control clauses ‘(not shown)’. rating_faith: faith they would catch the criminals if you were a victim (1 = not at all, 4 = a lot); rating_abuse_likelihood: how likely they would commit bribery, torture, discrimination, abuse or false arrest (1 = not likely, 4 = very likely). Balanced 2x6 design. Level text translated from the questionnaire. | 2026-10-14 | |
lopezgarcia_2023_mx_security_officers |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: ‘to reduce and control crime [in the most violent municipalities of the country], it is necessary to have more security officers on the streets [and five add-on programmes, from youth training to curfews and zero tolerance]’, control clauses ‘(not shown)’. rating_agree 1-5 (5 = strongly agree). Balanced 2x6 design. Level text translated from the questionnaire. | 2026-10-14 | |
lopezgarcia_2023_mx_tax_earmark |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: ‘paying more taxes [VAT / income tax] than you do now to increase public spending [in health / roads / security]’, each clause randomized with a left-out control (‘(not shown)’). rating_agree = agreement with paying more taxes, 1 = strongly disagree, 5 = strongly agree. Balanced 3x4 design. Level text translated from the questionnaire. | 2026-10-14 | |
lopezgarcia_2025_haiti_property_tax |
Conjoint table (irw_conjoint). choice: “Considering the tax proposals below. If you had to choose which proposal, would you prefer: A or B?” (pre-analysis-plan wording; forced, no opt-out). Level text is the English pre-analysis-plan text mapped from the deposit’s word codes; respondents likely saw Haitian Creole. Attribute order randomized per task (attrpos_). RIWI weight kept as cov_survey_weight; two empty rows dropped; vignette experiments 2-3 not included. | 2026-10-08 | |
lopezortega_2025_adoption_couples |
Conjoint table (irw_conjoint_2). Paired design (Catalonia, 2022): 5 tasks of 2 couples applying to adopt; attributes sexuality (two-men, two-women or man-woman couple, shown by names), open/closed relationship, age difference, education, family values, reason for adopting; choice = the couple that should be given a child first (forced). Sexuality and age difference are stored as the authors’ categories (displayed as partner names and age pairs, listed in the level text); other level text is the supplement’s, translated from Catalan. 822 tasks without a recorded choice omitted; CEO weight kept as cov_survey_weight. | 2026-10-14 | |
lopezortega_2025_curriculum_de |
Conjoint table (irw_conjoint). choice = which of two curriculum-reform proposals the respondent supports (forced choice; exact wording not deposited). Faces are AI-generated images stored by file id, with the authors’ coding of gender, age and ethnicity (attr_photo_*); the LGBT+/trans marker (a visual emoji cue) is stored as the authors’ code. The foreign-country attribute is the country name as displayed; its framing (opposition/endorsement) is not deposited. Faces, names, countries, reasoning and proposals were not to repeat within a respondent. The deposit has 1,263 respondents; the article reports 1,358 for this country. | 2026-10-07 | |
lopezortega_2025_curriculum_nl |
Conjoint table (irw_conjoint). choice = which of two curriculum-reform proposals the respondent supports (forced choice; exact wording not deposited). Faces are AI-generated images stored by file id, with the authors’ coding of gender, age and ethnicity (attr_photo_*); the LGBT+/trans marker (a visual emoji cue) is stored as the authors’ code. The foreign-country attribute is the country name as displayed; its framing (opposition/endorsement) is not deposited. Faces, names, countries, reasoning and proposals were not to repeat within a respondent. The deposit has 1,140 respondents; the article reports 1,169 for this country. | 2026-10-07 | |
lsbq_maleki_2025_persian_comprehension |
Items are 0-10 self-rating sliders (decimals are real slider positions). One response of 75 on 16.1.S, outside the 0-10 range and probably a typed 7.5, is excluded rather than recoded. | #1700 | 2026-10-01 |
lu_2025_property_tax_lab |
Conjoint table (irw_conjoint). Lab-in-the-field conjoint: forced choice between two property tax policies (no opt-out) and a 1-5 rating of each (variable rank; wording and anchors not deposited). Level text is the deposit’s terse English value labels. Separate design from the online conjoint (six attributes). | 2026-10-08 | |
lu_2025_property_tax_online |
Conjoint table (irw_conjoint). Forced choice between two property tax policies (no opt-out) and a 1-5 rating of each (variable rank; wording and anchors not deposited). Level text is the deposit’s terse English value labels. SSI and Qualtrics panels pooled as in the authors’ combined model (trial_panel); 5 respondents have fewer than 4 tasks. | 2026-10-08 | |
lublin_2023_candidate_pairs |
Conjoint table (irw_conjoint_2). Five pairs (A-E) of hypothetical congressional candidates with 7 attributes; choice = preferred candidate, ‘No Preference’ answers are an opt-out (both profiles 0; 1,959 of 12,515 tasks). Every pair has one Democrat and one Republican and one incumbent and one challenger; ethnicity is drawn conditionally on the respondent’s group. Level text is the deposit’s value labels; the question wording is only a truncated variable label (paraphrase); profile order (Candidate A = first) is inferred from the answer labels. | 2026-10-14 | |
lublin_2023_candidate_vignette |
Conjoint table (irw_conjoint_2). One text vignette per respondent of a US congressional candidate: ethnic identity bundle (surname, parents’ origin, roots, pan-ethnic identity; 9 levels) x career background (2); three 0-4 agree-disagree ratings (strongly disagree 0 … strongly agree 4; rating = ‘understands the issues that matter to me’). Ethnicity is drawn conditionally on the respondent’s own group, so its level shares depend on the respondent. Don’t know answers set to NA; 193 respondents with all three missing omitted. | 2026-10-14 | |
lueders_2025_relocation |
Conjoint table (irw_conjoint). Four forced choices between two destination cities (choice; wording paraphrased, no opt-out; 67 unanswered pairs kept as NA where ratings exist) plus ten per-city 0/1 judgements (rating_*, e.g. fear discrimination, government protects immigrants), five asked in each of two randomized question versions (cov_question_version). Attribute row order randomized per respondent (attrpos_). 1,164 respondents with an outcome (the authors’ d.full.conjoint, 1,165; completes 1,112); survey weight kept where the deposit has one. | 2026-10-09 | |
lupu_2024_laws_of_war_fr |
Conjoint table (irw_conjoint_2). One vignette per respondent (2x2x2x2 between-subject: rebel aims, government type, rebel compliance with the laws of war, government compliance); rating_government and rating_rebels = approval on a 5-point scale, 1 strongly disapprove … 5 strongly approve (question wording is not in the deposit). Levels are the authors’ condition names, not the vignette text; the order of describing rebels and government was randomized (trial_actor_order). No choice or opt-out; id = row number; rows answering neither item are omitted. Separate table per country. | 2026-10-11 | |
lupu_2024_laws_of_war_uk |
Conjoint table (irw_conjoint_2). One vignette per respondent (2x2x2x2 between-subject: rebel aims, government type, rebel compliance with the laws of war, government compliance); rating_government and rating_rebels = approval on a 5-point scale, 1 strongly disapprove … 5 strongly approve (question wording is not in the deposit). Levels are the authors’ condition names, not the vignette text; the order of describing rebels and government was randomized (trial_actor_order). No choice or opt-out; id = row number; rows answering neither item are omitted. Separate table per country. | 2026-10-11 | |
lupu_2024_laws_of_war_us |
Conjoint table (irw_conjoint_2). One vignette per respondent (2x2x2x2 between-subject: rebel aims, government type, rebel compliance with the laws of war, government compliance); rating_government and rating_rebels = approval on a 5-point scale, 1 strongly disapprove … 5 strongly approve (question wording is not in the deposit). Levels are the authors’ condition names, not the vignette text; the order of describing rebels and government was randomized (trial_actor_order). No choice or opt-out; id = row number; rows answering neither item are omitted. Separate table per country. | 2026-10-11 | |
lushenko_2025_ai_trust_cadets |
Conjoint table (irw_conjoint). rating: “Do you trust partnering with this AI-enhanced military technology when it is used under these conditions?” 1 (strong distrust) .. 5 (strong trust), raw; single profile per task, 9 tasks, fixed attribute order. Reliability scenario shown before and after the conjoint dropped (attributes not exported). 3 of 474 respondents dropped (attribute levels not saved). Covariates are unlabelled codes; masked IP/location/name columns and the free-text item dropped. | 2026-10-08 | |
lust_2024_female_candidates |
Conjoint table (irw_conjoint). Each respondent rated one candidate; rating = vote for the candidate and rating_* = nine further judgements of the same candidate, all 0-10 (anchors and exact wording not deposited; names paraphrase the authors’ variables; 98/99 don’t know/refused set to NA). Four binary attributes; level text is the authors’ value labels, not the displayed vignette prose. Saudi Arabia and Yemen pilot respondents dropped as by the authors; 507 respondents with no outcome omitted. | 2026-10-07 | |
lutz_2026_responsibility_sharing |
Conjoint table (irw_conjoint_2). choice = ‘Which of the two models of European migration policy do you favour?’ (forced choice, no opt-out, 2 pairs); rating = ‘How would you rate Model A [B]?’, 1 strongly disagree to 7 strongly agree (higher = more favourable). Choice is NA for some respondents’ tasks (ratings kept). Six countries pooled as the authors do; N against the article not checked. | 2026-10-11 | |
luypaert_2024_party_manifestos |
Conjoint table (irw_conjoint). Forced choice between two party manifestos (selected), no opt-out; question wording not in the deposit (authors: vote intention). Tasks 1-5 social democratic, 6-10 liberal manifestos (trial_manifesto). Levels uniform and unconstrained; attribute order randomized per respondent (attrpos_). The deposited manifesto ratings are not included: their scale and direction are undocumented. N = 2,119 vs 2,082 stated in the deposit. |
2026-10-08 | |
lyu_2024_ate |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_nfa |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_org |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_pbc |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_pvs |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_rtp |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_sef |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_sei |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
lyu_2024_sun |
Three pairs of adjacent rows are identical on all 42 items and all six demographics (differing only in a generated ‘Random’ variable); one row of each pair was dropped, so N is 681 rather than the paper’s 684. | 2026-10-05 | |
mader_2024_eu_military_costs |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (25 EU countries pooled): an EU military crisis intervention with or without a soldiers-at-risk clause and with national, pooled or no financing information; rating = support on a 5-point scale stored as 0-1 (higher = more support; wording not in the deposit). Levels are the authors’ labels; clauses left out are (not shown). Respondents with no answer are omitted. | 2026-10-11 | |
mader_2024_eu_procurement_costs |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (25 EU countries pooled): joint European defence procurement of aircraft with or without a job-cuts clause and with national, pooled or no financing information; rating = support on a 5-point scale stored as 0-1 (higher = more support; wording not in the deposit). Levels are the authors’ labels; clauses left out are (not shown). Respondents with no answer are omitted. | 2026-10-11 | |
madrid_2022_religious_candidates |
Conjoint table (irw_conjoint_2). Ten tasks of two hypothetical congressional candidates, six attributes (party, ideology, occupation, race, religion, sex); choice = candidate preferred as the respondent’s representative (forced choice, no opt-out). Attribute row order was randomized and is recorded (attrpos_). 538 export rows, 508 answered; the choice question is truncated in the export, so wording is a paraphrase; task order follows the authors’ script. | 2026-10-11 | |
maghbouleh_2022_mena_classification |
Conjoint table (irw_conjoint_2). Five pairs of fictitious immigrants (or U.S.-born citizens, trial_nativity_arm) per respondent, each profile classified with ‘How would you classify these immigrants? (please choose the single best category for each individual)’ into White / Black / Native American / Hispanic / MENA / Asian, stored as six 0/1 rating_ columns; no choice between profiles. Ancestry levels weighted to equalize broad categories; attribute order randomized per respondent (attrpos_); skin colour shown as a hand image (stored as ‘Hand image 1-10’). 64 unanswered profiles dropped; IP, zip code and Prolific IDs dropped. N = 1,077 answering (paper 333 + 329 + 417; the 4 White respondents the paper excludes are not flagged in the deposit and are kept). | 2026-10-09 | |
magistro_2024_automation_fairness |
Conjoint table (irw_conjoint). Single-profile design: rating = agreement that ‘The company’s decision to introduce the new technology is fair’, rating_ceo = would make the same decision as CEO, both 1 = strongly disagree to 5 = strongly agree (don’t know missing). Only the specific-information arm (2/3 of 8,033) saw tables; price levels are conditional on the product (same / 20% / 50% cheaper) and the low-skilled count never stays at 200, so estimate within those restrictions; wages are in local currency. 20 respondents whose attribute values were not saved are dropped; the policy items (cov_policy_*) were asked after the tables. | 2026-10-07 | |
magni_2021_lgbt_candidates_nz |
Conjoint table (irw_conjoint). choice: ‘Which of these two candidates would you be more likely to vote for?’, forced, no opt-out; rating: 1-7 per candidate, wording and anchors not deposited (stored raw; correlates positively with being chosen). Six tasks, 8 fully randomized attributes, row order randomized per task (attrpos_). Level shares unequal (Straight 2/3, Healthy 3/7, White 2/3, Not religious 1/2). Mechanism items dropped (undocumented 0 code); tasks without a recorded choice dropped. Six tasks in the file vs five in the article. | 2026-10-09 | |
magni_2021_lgbt_candidates_uk |
Conjoint table (irw_conjoint). choice: ‘Which of these two candidates would you be more likely to vote for?’, forced, no opt-out; rating: 1-7 per candidate, wording and anchors not deposited (stored raw; correlates positively with being chosen). Five tasks, 8 fully randomized attributes, row order randomized per task (attrpos_). Level shares unequal (Straight 2/3, Healthy 3/7, White 1/2, Not religious 1/2). Mechanism items dropped (undocumented 0 code); tasks without a recorded choice dropped. | 2026-10-09 | |
magni_2021_lgbt_candidates_us |
Conjoint table (irw_conjoint). choice: ‘Which of these two candidates would you be more likely to vote for?’, forced, no opt-out; rating: 1-7 per candidate, wording and anchors not deposited (stored raw; correlates positively with being chosen). Five tasks, 8 fully randomized attributes, row order randomized per task (attrpos_). Level shares unequal (Straight 2/3, Healthy 3/7; other attributes near-uniform). Mechanism items dropped (undocumented 0 code); tasks without a recorded choice dropped. | 2026-10-09 | |
maier_2026_supranational_de |
Conjoint table (irw_conjoint_2). Paired forced choice between two supranational governance arrangements (4 attributes, text in German), plus a raw 1-7 rating of each profile. Question wording, rating anchors and direction are not in the deposit; covariates kept as undocumented codes; IP, coordinates and name/email columns dropped. Germany sample, 1,027 respondents; sister table maier_2026_supranational_pl. | 2026-10-12 | |
maier_2026_supranational_pl |
Conjoint table (irw_conjoint_2). Paired forced choice between two supranational governance arrangements (4 attributes, text in Polish), plus a raw 1-7 rating of each profile. Question wording, rating anchors and direction are not in the deposit; covariates kept as undocumented codes; IP, coordinates and name/email columns dropped. Poland sample, 1,051 respondents; sister table maier_2026_supranational_de. | 2026-10-12 | |
makovi_2021_environmental_concern |
air_polluted (local air pollution is an issue) is printed in the survey’s Environmental Concern Scale block but is not part of the authors’ 7-item environmental concern score; consider whether to model it with escale_1..7. escale_6 (threats are exaggerated) is reverse-worded and not reversed. | 2026-10-04 | |
makovi_2021_symbolic_racism |
Response formats differ by item: rscale_4 has 4 printed options (1-4) while the other ten have 5, and items 4, 6, 7 and 11 use their own verbal anchors rather than agree/disagree. Items 3, 8 and 9 are reverse-worded relative to the rest and are not reversed in this table. | 2026-10-04 | |
malesky_2020_vietnam_na_2016 |
Conjoint table (irw_conjoint). One pair of hypothetical National Assembly candidates per respondent; choice = “Which candidate would you vote for?” (forced; 1,147 don’t-know/refusals omitted). Attribute text is the file’s value labels, Vietnamese for gender and province and English for the rest; the candidate’s province is by design the respondent’s own or Hanoi. cov_survey_weight is the PAPI sampling weight used in the authors’ svyset (which also post-stratifies by province population, cov_province_population). Village and commune identifiers and free text dropped. | 2026-10-08 | |
malhotra_2017_immigrant_skill |
Conjoint table (irw_conjoint). Factorial vignette, one hypothetical immigrant per respondent. choice: binary admission question (Q1; 1 = admit, 0 = not admit; single-profile accept/reject, so opt-out); rating_admit_scale: admission scale 1-7 (7 = most in favour); rating_how_many_more and rating_admit_ten_thousand: 1-5, stored raw with LOW = most in favour (the authors reverse them). Wording not deposited (paraphrase). Education and profession are bundled, and the prevalence sentence is fixed by country and skill (‘(not shown)’ in the no-information cells). Survey weight kept. | 2026-10-09 | |
malik_2018_social_motivation |
Item codes do not follow the questionnaire’s numbering. social_1 and social_2 are SC1 and SC2; social_3 is the source’s 4th item (its column is named CS4 in the deposit); social_4 to social_14 are SC5 to SC15; social_15 is SC3, “I am satisfied with my personal life issues”, which was missing until 2026-09-30. | #2127 | 2026-09-30 |
malik_2025_afghan_refugees |
“I want you to pick which one you would prefer to stay in Pakistan”: forced choice; refused or don’t-know rounds omitted, and 714 of 3,500 respondents skipped all three rounds. One occupation level is the respondent’s own household-head occupation, filled in from an earlier answer. Attribute order randomized but not recorded; English labels, while interviews were in Urdu/Pashto. | #2887 | 2026-10-07 |
manekin_2019_disputed_territory |
Conjoint table (irw_conjoint). choice = forced choice between two policy packages (wording not deposited), no opt-out, one task per respondent with profile order inferred from row order and verified against the recorded choice; rating = the authors’ ‘Conjoint rank policy (1-7)’, wording and direction undocumented, stored raw. Three prohibited level pairs (security x budgets, security x economy) per the authors’ design code; levels are the authors’ English (survey in Hebrew). The deposit’s second (phase-2) conjoint is not built: numeric codes without displayed text. | 2026-10-09 | |
marble_2025_run_for_office |
Conjoint table (irw_conjoint). Two ratings per election scenario, two scenarios per screen, 3 screens: rating = “How interested would you be in running in each election?” (1 Not interested - 5 Very interested) and rating_winprob = “If you were to run, how likely is it that you would win in each election?” (1 Very unlikely - 5 Very likely), coded from the option text; question order and option order randomized (cov_question_order, cov_answer_order). Opponent ideology was restricted to the side opposed to the respondent’s self-placement; attribute order randomized per respondent (attrpos_). 745 respondents vs 734 responses in the preanalysis plan. | 2026-10-09 | |
maredia_2020_bean_seed_choice |
Conjoint table (irw_conjoint). Discrete choice experiment, real and incentivized: ‘choose which of the two seeds you would purchase if you were at the market to buy seed. If you wouldn’t buy either then choose ninguna’ (enumerator script); ninguna = opt-out (both 0). Seed shown only as the plot symbol (circle = certified, square = QDS, triangle = recycled, blind to farmers). Fixed 12-pair design in four display orders. 2 farmers without a choice set and 2 invalid answers dropped. Price barely affects choice. | 2026-10-08 | |
mares_2020_lesser_evil |
Conjoint table (irw_conjoint). choice = which candidate the respondent would vote for (authors’ ‘Electoral Choice’; question wording not deposited), with a third, undocumented response coded 3 in 257 of 2,510 tasks treated as an opt-out (both profiles 0). Restriction: every pair shows one incumbent mayor and one candidate without mayoral experience. Attribute text is the appendix’s English version. N = 502 as in the paper. | 2026-10-08 | |
margalit_2022_stock_flow |
1-7 scale from “absolutely not admit [definitely deport]” to “definitely admit [allow to stay]”, higher = more favourable; no choice question. Each respondent saw one framing (trial_frame: Flows, Stocks, Stocks-deport). Levels are the authors’ short labels; no covariates because the respondent file cannot be linked. | #2887 | 2026-10-07 |
markgraf_2024_credit_welfare |
Conjoint table (irw_conjoint). choice = which of two policy packages the respondent preferred (forced choice; wording not in the deposit). Five three-level attributes (mortgage and credit-card interest rates, social assistance, unemployment insurance, income tax), every level pair present. Profile number is not screen position (the deposit has no position column and its rows are shuffled). N = 1,286 as in the codebook. | 2026-10-08 | |
martin_2021_corruption |
Conjoint table (irw_conjoint). choice: “which official would you rather see prosecuted and punished?” (forced, no opt-out). rating: seriousness of the corruption, 1 = not at all to 5 = extremely serious; higher means more severe, the same direction as choice. Levels were shown as icons explained in Lusoga/Luganda; attr_ holds the English captions. GPS and interviewer identifiers dropped. | 2026-10-07 | |
martin_2021_minority_candidates |
Conjoint table (irw_conjoint). One task of two candidate vignettes per respondent; choice = ‘Which of these candidates would you rather have as your MP?’. The answer options were randomized: trial_neither_offered = 1 when a ‘neither’ option was shown (opt-out; 1,603 of 3,960 chose neither, both profiles 0), 0 for a forced choice. Level text from the article’s vignette template mapped to the authors’ value labels; candidate names not stored. Pairs and profile order inferred from row order. N = 7,903 as in the article. | 2026-10-09 | |
martineztomas_2026_pseudowords |
Rating scales differ by item: valence 1-9, arousal 1-9, wordlikeness 1-7. These are the raw ratings: raters the authors screened out (>= 95% identical responses, or low agreement with the group) are included and can be identified and dropped from the rater column. One questionnaire’s list number differs between the source’s participants and responses files; the responses file’s value is used. | #1443 | 2026-10-01 |
martini_2021_good_politician |
Conjoint table (irw_conjoint). choice = which of the two candidates the respondent would have preferred to win a seat in the European Parliament (forced choice, no opt-out; article’s English). Levels are the authors’ English labels (respondents saw Italian); all 3,096 respondents are kept, including 420 who saw the rare ‘artisan’ job level that the article’s main analysis (n = 2,676) excludes; the 7-point rating is not built because only a dichotomized version was deposited. Attribute order was randomized but not recorded; cov_attention_pass flags the authors’ attention subsample. | 2026-10-07 | |
mathisen_2026_satisfaction |
Conjoint table (irw_conjoint). Three single-profile text vignettes (Country A, B, C) per respondent, each with three randomized components (influence equality/inequality, economic and moral policy position); rating = ‘As an ordinary citizen of Country A, how satisfied would you be with the government?’ 0 = extremely dissatisfied to 10 = extremely satisfied. Level text from SPSS labels, with truncated labels completed from the article; influence level text names the country. N = 2,448 as in the article. | 2026-10-09 | |
matsaid_2022_abcdm_intention |
Code 0 (‘non-applicable’) dropped as a non-response (49 cells); the paper’s totals score it 0. IC23 and IC26 are stored reverse-scored, as in the scoring guide. In the deposit IC25/IC26 correlate 0.88 with each other and about 0.1 with IC21-IC24, which does not reproduce the paper’s single-factor loadings (0.72-0.97 for all six items). | 2026-10-05 | |
matsaid_2022_abcdm_knowledge |
Only the scored item is deposited (1 = correct; 0 = incorrect or don’t know, so ‘don’t know’ cannot be separated from a wrong answer). Item 8 is the one keyed False. | 2026-10-05 | |
matsaid_2022_abcdm_perceived_benefits |
Code 0 (‘non-applicable’) dropped as a non-response (26 cells); the paper’s totals score it 0. | 2026-10-05 | |
matsaid_2022_abcdm_perceived_risk |
Code 0 (‘non-applicable’) dropped as a non-response (171 cells); the paper’s totals score it 0. PR15 (‘I am not worried…’) is stored reverse-scored, as in the scoring guide. The paper’s Table 2 loadings for items 15/16 (0.263/0.539) are not reproduced from the deposit (PR15 0.51, PR16 0.14 in a three-factor CFA), so the two columns may be transposed relative to the questionnaire numbering. | 2026-10-05 | |
matsuo_2018_uk_policy_packages |
Conjoint table (irw_conjoint). choice: “Which of these two candidates would you be most likely to vote for?” between candidates of two parties with the displayed policy packages; forced choice, no opt-out. Levels drawn uniformly and independently; attribute order randomized per respondent (not recorded); profile position inferred from row order. Unanswered tasks and one respondent ID holding two complete sessions dropped: 1,403 respondents against 1,394 in the paper. Platform IDs re-keyed; party ID, education and nation kept as undocumented codes. | 2026-10-08 | |
mattis_2025_ai_disclosure |
Conjoint table (irw_conjoint_2). Single-profile ratings of an article’s trustworthiness, 1 (not at all) to 7 (extremely trustworthy), no choice. 7 binary attributes (human or AI for idea generation, image generation, background research, article writing, proofreading, fact checking; human in the loop yes/no) from a fixed 8-row fractional design repeated in three topic blocks (fixed blocked design; trial_topic, trial_article_image). Level text is the authors’ code, not the displayed Dutch wording; the display order within a topic is not recorded. N = 683 respondents (the authors’ filtered sample), 24 profiles each. | 2026-10-11 | |
maxey_2025_intervention_success |
Conjoint table (irw_conjoint). choice: “Which intervention was the MOST successful?” (forced, no opt-out; 43 pairs unanswered); rating: “How successful were each of the interventions?” 1 = Very Unsuccessful .. 4 = Very Successful. Task and profile inferred from row order. Level text is the deposit’s value labels (may abbreviate the displayed wording); restrictions undocumented. cov_attention_pass flags the pre-treatment attention check used in the authors’ robustness check. | 2026-10-07 | |
mayne_2026_electoral_reform |
Conjoint table (irw_conjoint). Four forced choices between two electoral-reform scenarios per task (Scenario 1/2, no opt-out): choice = “If you had to choose one of these scenarios, which would you pick?”; choice_voice, choice_governability (1 = scenario judged to make legislation more difficult to pass) and choice_responsiveness. Attribute order randomized by respondent (attrpos_*); the ideology attribute was a graph, stored as its lead sentence plus the authors’ label. Qualtrics IDs and free text dropped; 2,177 respondents and 12,842 rows as in the Supplementary Material. | 2026-10-07 | |
mazepus_2021_judicial_reform_ua |
Conjoint table (irw_conjoint_2). Between-subject 3x2 scenario, one per respondent: the winning party is unnamed, supported or opposed, and incomes under its government stayed the same or rose, followed by a judicial reform. Five 1-7 ratings (support, trust, reform justified, reform acceptable, vote likelihood; higher = more); no opt-out. cov_survey_weight is the authors’ weight. Level text is the English original of the Russian/Ukrainian vignettes. Ukraine sample only. | 2026-10-14 | |
mbozi_2024_tally_sheets |
Conjoint table (irw_conjoint). Image conjoint of vote tally sheets; attributes are the author’s coding of six manipulated features. choice: which of two sheets has “more reliable information about the vote outcomes” (forced, no opt-out); rating_misconduct 1 = very unlikely .. 5 = very likely misconduct occurred, and four 1-5 agree items (tally changed, party interests unprotected, party influence, officer pressure); all ratings: higher = more suspicion, kept as asked. Clean and corrected images came from separate pools (non-uniform combinations); 21 tasks with identical feature sets are kept (the author drops them, 2,298 rows). Vote numbers on the images not stored. | 2026-10-07 | |
mccammon_2024_phage_therapy |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical infection treatments (incl. phage therapy) shown to UK health professionals; choice = which treatment the respondent would prefer to recommend (forced, no opt-out); rating = likelihood of recommending each treatment, 0-10 (survey text says 1 to 10, slider and data run 0-10), stored raw. Attribute order is randomized per task (attrpos_). Two pilot respondents with an earlier 5-attribute design are dropped; 190 respondents, article N not checked. | 2026-10-12 | |
mccrea_2026_therapist_choice |
Conjoint table (irw_conjoint). choice: which of two therapists the respondent picks (forced, no opt-out); rating: 1-7 rating of each therapist, higher = more favourable, anchors not deposited (wording paraphrased). 3 scenarios x 2 therapists; each respondent saw one of three versions hiding two attributes (stored as ‘(not shown)’); the authors analyse only two versions. Specialty 1 and 2 never share a level. Prolific IDs and free text dropped. | 2026-10-09 | |
mcdonnell_2022_gp_referral |
Conjoint table (irw_conjoint). Single-profile text vignette, choice = “Given the information provided above, would you refer the patient to the ED or not refer?” (Refer = 1), not referring is the outside option (opt_out = yes); fixed blocked design, 6 vignettes per block (blocks 1 and 3), levels coded from the instrument’s scenario texts. Task order follows the instrument (not recorded). 150 GPs answered all six scenarios; the record reports n = 142. | 2026-10-09 | |
mcdowell_2025_immigrant_partisanship |
Conjoint table (irw_conjoint_2). One question per respondent; rating = agreement that the United States should permit higher levels of immigration from [Vietnam/Venezuela], 0 = strongly disagree to 10 = strongly agree. Full 2 x 2 (origin x partisanship sentence; the sentence is ‘(not shown)’ in control); panel ids dropped; 3,002 respondents vs 3,000 reported. | 2026-10-09 | |
mclaughlin_2025_spending_projects |
Conjoint table (irw_conjoint_2). Ten forced-choice pairs of federal spending projects (project type, amount, administrator, time), US panel; choice = the project picked, rating = 1-7 rating of each project (wording and anchors not in the deposit). The block is not analysed in the paper; versions a/b (trial_version) are not explained; withdrawn respondents dropped. | 2026-10-11 | |
mclean_2017_sanctions |
Conjoint table (irw_conjoint_2). One vignette per respondent (2x2x2 between-subject: sanction type, expected effectiveness, success/failure frame). choice = ‘Should the U.S. impose the sanctions recommended by the President? (Yes / No)’, a single-profile accept/reject (opt-out); ratings: likelihood of success 0-100, pain for the Erqatan public and leaders 1-10. Attribute text is the authors’ condition names, not the full paragraph; id is the row number; the paper’s N was not checked. | 2026-10-11 | |
mclean_2022_mps |
Items OOP2, OOP5, OOP6, OOP14, OOP15 carry reversed value labels in the source .sav (1 = strongly agree .. 7 = strongly disagree): they are stored reverse-scored. Shipped as stored, not recoded. The same applies to PSP3, NDPI6, NDCI9 in mclean_2022_psps and to the LSE items in mclean_2022_rses. | 2026-10-07 | |
meco_l1_comprehension |
Pooled from MECO L1’s two collection rounds, which are different people at mostly different sites: cov_sample 1 = wave 1 (13 sites), 2 = wave 2 (16 sites). It is not a panel, so the round is cov_sample rather than wave. cov_language is the language and cov_site the release’s site code. The 42 wave-1 Norwegian readers answered 47 questions (text 6 question 4 was not administered). | #2207 | 2026-10-01 |
meco_l1_comprehension |
Item codes are shared only where the question is the same. Only texts 1, 3, 7, 11 and 12 are translations of a common original (the authors’ is.matched flag); their questions are text{t}_q{q} across sites. The other 7 texts were written per language, so their items carry a language or text-version prefix (e.g. du_text2_q1) and are answered only by that site’s readers: the table is sparse by design, with a median of 45 readers per item. A version is shared across sites or rounds only where the text is identical (no and tr across rounds, en with en_uk, the two Hindi sites except text 5, German wave 2 as ge_v2, sp with sp_ch on texts 2, 4, 5 and 9). Eight wave-1 matched questions that are a different question in that language get site-specific codes (ee_text3_q1, ee_text3_q2, sp_text3_q3, he_text1_q3, he_text12_q4, ru_text12_q4, gr_text12_q2, tr_text12_q3). | #2207 | 2026-10-01 |
meco_l1_comprehension |
Wave 2 published no comprehension-question wording, so for wave-2 sites the equivalence of the matched questions (text{t}_q{q}) is assumed from the authors’ design and is unverified. Two wave-2 sites contradict it in their responses and are coded site-specific instead: ru_mo for all 48 items (its per-question accuracy does not match any other site, nor wave-1 Russian on identical texts) and ba for text 11. | #2207 | 2026-10-01 |
meisels_2024_donor_candidates |
Conjoint table (irw_conjoint_2). Factorial vignettes (US donors): each donor rated 5 House-candidate vignettes (primary or general context, trial_election) varying district lean, opponent extremity/incumbency, ideology relative to donor and four viability cues; rating = very likely/almost certain to contribute (1) vs not (0), the authors’ dichotomization. Level text paraphrased from the ReadMe and figure labels (not the displayed wording); viability cue 0 coded ‘(not shown)’ by inference. 451 vignettes with no outcome dropped; inverse-propensity weight kept (NA for some donors). | 2026-10-14 | |
mercea_2025_movement_parties |
Conjoint table (irw_conjoint). choice = “If you had to choose between them, which candidate do you prefer?” (forced, no opt-out); rating = approval of each candidate, 1 = strongly disapprove to 7 = strongly approve. Six countries pooled (cov_country); attribute text is the English instrument with [COUNTRY] in the migration levels, since the national-language labels in the deposit are truncated. Levels randomized without restrictions; attribute order randomized once per respondent (not recorded). N = 10,347 as in the article. | 2026-10-08 | |
mey_2024_regional_electricity |
Conjoint table (irw_conjoint). choice: which of two regional electricity-system designs (left/right) the respondent prefers, 8 tasks, forced choice, no opt-out (exact wording not in the deposit). Six attributes stored as the exporter’s English labels. DNK, DEU, POL, PRT pooled (cov_country); pre-test respondents excluded following the authors’ workflow (4,103 of 4,526). IP, city, postcode, coordinates, panel codes and free text dropped. | 2026-10-09 | |
meyer_2022_bfi |
478 respondents who took the survey more than once are excluded: the deposit stores their answers averaged across attempts. Big Five items are coded 0-6 (seven points), although the paper says every measure was answered on a five-point 1-5 scale. Codes are shipped as stored. | #2877 | 2026-10-06 |
meyer_2022_covid_beliefs |
478 respondents who took the survey more than once are excluded: the deposit stores their answers averaged across attempts. The paper dropped three information-seeking items for low loadings; all ten are shipped. Codes are 0-4, where the paper describes a 1-5 scale. | #2877 | 2026-10-06 |
meyer_2022_covid_selfeff |
478 respondents who took the survey more than once are excluded: the deposit stores their answers averaged across attempts. Codes are 0-4, where the paper describes a 1-5 scale. | #2877 | 2026-10-06 |
meyer_2022_pss10 |
478 respondents who took the survey more than once are excluded: the deposit stores their answers averaged across attempts. Codes are 0-4, where the paper describes a 1-5 scale. | #2877 | 2026-10-06 |
mgkt_options |
Each MGKT question showed 10 options, 5 of them correct, and asked respondents to tick only those they knew. Each row is one option: resp = 1 if ticked, 0 if not; a question with nothing ticked is all zeros. Key: for every question, opt0-opt4 are correct and opt5-opt9 incorrect (the codebook’s answer codes A0-A9; the key was checked against the stored scores of all 19,218 respondents). Option texts are in the codebook (MGKT_data.zip). The order of clicks is not kept. The hits-minus-false-alarms score is in mgkt, with the same ids. | #2662 | 2026-10-01 |
mikkelborg_2025_primary_lucid1 |
Conjoint table (irw_conjoint). choice: “Which candidate for Congress would you support in this Democratic Primary election?” (Candidate A/B, forced, no opt-out); one task per respondent. Restricted randomization: one candidate always Black, one White; no shared endorsement. Deposit contains white respondents only. Ratings stored 0-1 as deposited. | 2026-10-07 | |
mikkelborg_2025_primary_lucid2 |
Conjoint table (irw_conjoint). choice: “Which candidate for Congress would you support in this Democratic Primary election?” (Candidate A/B, forced, no opt-out); one task per respondent. Restricted randomization: one candidate always Black, one White; no shared endorsement. Deposit contains white respondents only (all parties; the paper analyses Democrats: cov_party_id7 Lean, Mostly or Strong Democrat). Ratings stored 0-1 as deposited. Profile order inferred from row order. One respondent with no answers dropped. | 2026-10-07 | |
mikkelborg_2025_primary_lucid3 |
Conjoint table (irw_conjoint). choice: “Which candidate for Congress would you support in this Democratic Primary election?” (Candidate A/B, forced, no opt-out); one task per respondent. Restricted randomization: one candidate always Black, one White; no shared endorsement. Deposit contains white respondents only (all parties; the paper analyses Democrats: cov_party_id7 Lean, Mostly or Strong Democrat). Profile order inferred from row order. One respondent with no choice dropped. | 2026-10-07 | |
mikkelborg_2025_primary_lucid4 |
Conjoint table (irw_conjoint). choice: “Which candidate for Congress would you support in this Democratic Primary election?” (Candidate A/B, forced, no opt-out); one task per respondent. Restricted randomization: one candidate always Black, one White; no shared endorsement. Deposit contains white respondents only (all parties; the paper analyses Democrats: cov_party_id7 Lean, Mostly or Strong Democrat). Profile order inferred from row order. | 2026-10-07 | |
miller_2022_blame_crisis |
Conjoint table (irw_conjoint_2). Factorial vignettes, three per respondent (US MTurk): executive response (control / blame claim / blame deflect) x crisis severity (low / moderate / high) x executive party relative to respondent; outcomes rating_handling (approve of handling, 1-4, higher = more approval) and rating_vote (likelihood of voting for the executive, 1-4, higher = more likely), authors’ codings. Attribute levels are the authors’ dummy names (vignette text not in the deposit); task = scenario index, tasks lacking a party-match code or outcome dropped (872 to 845 respondents). | 2026-10-14 | |
miller_2022_lobbyist_access |
Conjoint table (irw_conjoint). choice = member chosen as the more attractive access target (codebook paraphrase), forced pick of 2; rating = five-point interest in meeting the member, 1-5 as stored (anchors undocumented; higher = more interest by its relation to choice). Restrictions from design.dat (party-specific subcommittee posts, leadership only on committee, cosponsor implies Support); legislative stage randomized per task (trial_stage). 657 tasks with no answer dropped; 670 respondents (all with covariates in resp_data.csv). | 2026-10-09 | |
miller_2024_backsliding_travel |
Conjoint table (irw_conjoint). choice = ‘At which of these destinations are you most interested in vacationing?’ (Destination 1/2/3 or ‘None of these destinations’; OPT-OUT, 16% of tasks have choice 0 on all three); rating = ‘How interested are you in vacationing at each destination?’ 1 Not at all interested to 5 Extremely interested (some blanks). 3 profiles per task; attribute order randomized and recorded (attrpos_). Unanswered tasks omitted; 2,076 respondents vs ‘nearly 2,100’ in the article. Two news levels lack the apostrophe in ‘voters’ (removed by the authors). | 2026-10-07 | |
miller_2024_news_selection |
Conjoint table (irw_conjoint). choice: “Would you prefer to read News Selection A or News Selection B?”; OPT-OUT in the abstention format only (“Neither”: choice = 0 on both profiles, about 25% of abstention-format tasks); forced format has no opt-out. trial_arm / trial_format / trial_repeat_of identify the arm and the repeated 13th task. Two attributes (headline, source), row order randomized and recorded; restrictions undocumented. All respondents with an answered task kept (no screening applied, as in the authors’ code). | 2026-10-07 | |
miller_2024_scandal_candidates |
Conjoint table (irw_conjoint). choice: “If this election were being held in your district, would you vote for Candidate A or Candidate B?” (forced format) or “… Candidate A, Candidate B, or neither candidate?” (abstention format, OPT-OUT: choice = 0 on both, about 26% of tasks); blank when only ratings were given. rating_shares_views / rating_good_morals / rating_cares: agreement that the candidate shares my views / has good morals / cares about people like me, 1 strongly disagree - 5 strongly agree. Characteristics hidden by the information environment are “(not shown)”. Gender and race levels unequal (observed). No screening applied. | 2026-10-07 | |
miller_2026_backsliding_jobs |
Conjoint table (irw_conjoint_2). Three pairs of hypothetical state-government jobs per respondent, six attributes; choice = preferred job (no opt-out, one per pair), rating = 1-7 rating of each job (direction undocumented, presumably higher = more attractive). Task and profile are inferred from the state_gov index (verified: one preferred per pair); wording is a paraphrase since no codebook is deposited; 3 pairs with missing choice dropped. | 2026-10-11 | |
miwa_2021_ideological_labels |
Conjoint table (irw_conjoint). Between-subject label arm (trial_label): choice_more_right = which candidate is more to the right (right arm), choice_more_left = which is more to the left (left arm), both forced choice with no opt-out, NA on the other arm; rating = placement of each candidate on a 1-10 left-right scale (anchors not in the deposit; higher = more right in the data). Respondents over 69 dropped as by the authors (1); the authors’ satisficer and fast-task exclusions (1,330 attentive respondents) are left to the user via cov_attention_pass_1/_2 and trial_page_submit_sec. Wording paraphrased from the authors’ code (no questionnaire in the deposit). | 2026-10-08 | |
miwa_2023_dynastic_candidates |
Conjoint table (irw_conjoint). rating: favourability of a single candidate profile, 1 (not favorable at all) to 8 (very favorable), as in the source. Each respondent rated 10 House of Representatives and 10 House of Councillors candidates (trial_hoc); task numbers follow the deposit’s row order, which is not the display order (election-block order was randomized but the deposit is sorted HoR first). Nonuniform level weights matched to real-world distributions. Typo ‘Buniness employee’ corrected. 1,126 respondents, matching the paper. | 2026-10-08 | |
miwa_2023_policy_positions |
Conjoint table (irw_conjoint_2). Paired candidate profiles (hypothetical candidates’ stances on ten policy issues, 2 levels each, Japan 2022 survey), 15 tasks x 2 profiles; attribute order randomized per respondent and recorded. rating = likelihood of voting for each candidate on a 6-point scale from ‘definitely would not vote’ to ‘high probability of voting’; the numeric direction (higher = more likely) is inferred, not labelled in the deposit. No choice column; attribute text is Japanese as stored. | 2026-10-14 | |
miwa_2024_populist_pilot1 |
Conjoint table (irw_conjoint). Pilot 1 for Study 2: ratings of perceived anti-elitism and people-centrism of each candidate, 1-6. No choice question. Manipulated-mediator arm adds a ‘special note’ per candidate (8 notes, each used once per respondent; note set differs between pilots); natural-mediator arm ‘(not shown)’. Wording from the deposit’s English codebook (Japanese survey). | 2026-10-08 | |
miwa_2024_populist_pilot2 |
Conjoint table (irw_conjoint). Pilot 2 for Study 2: ratings of perceived anti-elitism and people-centrism of each candidate, 1-6. No choice question. Manipulated-mediator arm adds a ‘special note’ per candidate (8 notes, each used once per respondent; note set differs between pilots); natural-mediator arm ‘(not shown)’. Wording from the deposit’s English codebook (Japanese survey). | 2026-10-08 | |
miwa_2024_populist_pilot3 |
Conjoint table (irw_conjoint). Pilot 3 for Study 2: ratings of perceived anti-elitism and people-centrism of each candidate, 1-6. No choice question. Manipulated-mediator arm adds a ‘special note’ per candidate (8 notes, each used once per respondent; note set differs between pilots); natural-mediator arm ‘(not shown)’. Wording from the deposit’s English codebook (Japanese survey). | 2026-10-08 | |
miwa_2024_populist_study1 |
Conjoint table (irw_conjoint). Ratings of perceived anti-elitism and people-centrism of each candidate, 1-6 (codebook anchors). No choice question. Arms: party affiliation shown or not (‘(not shown)’). Level frequencies follow the 2017 HoR candidate distribution (nonuniform). Wording from the deposit’s English codebook (Japanese survey). | 2026-10-08 | |
miwa_2024_populist_study2 |
Conjoint table (irw_conjoint). Rating of each candidate’s desirability as a member of the House of Representatives, 1-6. No choice question. Manipulated-mediator arm adds a ‘special note’ per candidate (4 notes, each used once per respondent); in the natural-mediator arm the note is ‘(not shown)’. Wording from the deposit’s English codebook (Japanese survey). | 2026-10-08 | |
moehlecke_2025_fdi |
Conjoint table (irw_conjoint). choice_state: which firm would be better for the respondent’s state (paraphrase of the authors’ ‘sociotropic’ outcome); choice_family: which firm would be better for the respondent and their family (paraphrase, ‘egotropic’); forced choice, no opt-out. Profiles come from 20 fixed design versions (trial_version), so only 223 of 960 profiles occur. Panel IDs, municipality identifiers and undocumented items dropped; N (2,008) not checked against the paywalled article. | 2026-10-07 | |
mohamed_2021_covid_attitude |
Five rows identical on every column (double submissions) were dropped, so N is 982 rather than the paper’s 987. | 2026-10-04 | |
mohamed_2021_covid_practice |
The three own-practice items are the authors’ binary collapse (yes frequently vs. sometimes or no), while the three family items keep the raw coding 1 = yes frequently to 3 = no, so the two halves run in opposite directions. Five duplicate rows were dropped (N 982 vs the paper’s 987). | 2026-10-04 | |
mohr_2024_museum_visits |
Conjoint table (irw_conjoint). choice = which of two museums (Museum A / Museum B) the respondent chose; forced choice, verbatim wording not deposited. Attribute levels are the authors’ short Stata value labels, not necessarily the displayed text; cost is the stored dollar amount (the .dta cost labels are mis-attached). Randomization restrictions not documented, level shares uniform. One unanswered task dropped (2,315 tasks). | 2026-10-07 | |
moise_2025_eu_defence |
Conjoint table (irw_conjoint_2). Factorial vignettes (7 EU countries, 2022): each respondent evaluated 2 scenarios varying Russian actions (4), the US response (2) and the EU defence-capacity response (4); rating = evaluation of the scenario, 0-10 (11-point scale; question wording and anchors not verbatim, higher read as more approval). Attribute level text taken from the article (Table 1); the deposit stores short tags. Task order inferred from file order; 4,314 unrated rows dropped; countries pooled (cov_country). | 2026-10-14 | |
mon_2024_covid_democracy |
Conjoint table (irw_conjoint). choice = which scenario was selected as preferred (wording not deposited), forced choice, no opt-out. Attribute levels are associated and unequal in the data (pre-generated choice design), so estimate with all attributes in the model. Level text rebuilt from dummies with codebook wording (English; display language undocumented); the codebook’s attribute names for A3/A4 are swapped relative to the Stata labels, which were followed. | 2026-10-07 | |
montal_2024_journal_choice |
Conjoint table (irw_conjoint). Forced choice of the journal the respondent would submit to, between two hypothetical journals (3 tasks); exact wording not deposited, no opt-out. Attribute levels are the deposit’s short English labels; covariates are Spanish answer text named by question number. 4 unanswered tasks dropped, leaving 445 of the paper’s 446 respondents; Qualtrics IDs and one free-text field dropped. | 2026-10-08 | |
montana_blasco_2026_digital_skills |
Codes 6 (no sabe) and 7 (no contesta) dropped (2,262 responses); native coding kept, 1 = can do and could teach others. | 2026-10-10 | |
montana_blasco_2026_mh_statements |
Stored 1 = verdadera, 2 = falsa; no scoring key applied. | 2026-10-10 | |
montana_blasco_2026_tiktok_wellbeing |
TikTok users only (B1107 = 1), as the codebook directs. | 2026-10-10 | |
mortenson_2025_democracy_tradeoffs |
Conjoint table (irw_conjoint). choice = ‘Please select which country you would MOST prefer to live in’ (two countries with different features to their democracy), forced choice, no opt-out, 5 tasks. 15 respondents screened out at an attention check before the conjoint are excluded (623 = article N); 6 who failed the post-conjoint check are kept with cov_attention_pass_2 = 0. Demographics are unlabelled Qualtrics codes. | 2026-10-08 | |
much_tte_2025_concentrationtask |
wave numbers three measurement points in one session: before matrix block 1, between the blocks, and after block 2. No concentration item was answered under a speed instruction; treat = 1 marks participants randomised to take the speeded matrix block first. |
#2513 | 2026-09-28 |
much_tte_2025_currentmotivation |
wave numbers three measurement points in one session: before matrix block 1, between the blocks, and after block 2. treat = 1 marks participants randomised to take the speeded matrix block first, so for them wave 2 follows the speeded block. |
#2513 | 2026-09-28 |
much_tte_2025_effort |
wave 1 and 2 are retrospective effort ratings given after matrix block 1 and block 2 of the same session. treat = 1 marks participants randomised to take the speeded block first, so for them wave 1 rates the speeded block. |
#2513 | 2026-09-28 |
much_tte_2025_matrixreasoning |
Items MR01 to MR10 form block 1 (wave 1) and MR11 to MR20 block 2 (wave 2), in one session. treat = 1 marks items answered under the speeded instruction; participants were randomised to meet it in block 1 or block 2, so item, block and condition are only partly crossed. |
#2513 | 2026-09-28 |
mueller_2024_covid_governance |
Conjoint table (irw_conjoint). choice = which of two scenarios for taking COVID-19 response decisions the respondent preferred (paraphrase; forced choice, no opt-out), 4 tasks. Task/profile inferred from row order; attribute order fixed; level text = article Table 1 English with country-specific government terms. Six countries pooled (cov_country); N = 7,599 as in the article; cov_attention_pass from the authors’ Covid knowledge check. | 2026-10-09 | |
mullinix_2015_political_knowledge |
resp is the deposit’s own scoring (CVeto, CMajHouse, CConstitu, CSecState); each closed item’s correct option is the single code the deposit scores 1 (Veto 4, MajHouse 2, Constitu 3). SecState is open-ended and has no resp_raw. PartyKnow is not shipped (no documented key). | 2026-10-09 | |
mummolo_2016_news_choice |
Conjoint table (irw_conjoint). choice = which of the two news items the respondent would choose to read (paraphrase; wording not in the deposit), forced choice, no opt-out, 12 tasks per respondent. Partisans only (cov_dem / cov_rep). Two attributes: headline text and outlet; the two items share a topic in about 9% of tasks and an outlet in about 34%, with no restriction documented. Qualtrics response ids re-keyed; the author’s derived topic and friendly/unfriendly-source dummies dropped. | 2026-10-08 | |
mummolo_2016_news_choice_mturk |
Conjoint table (irw_conjoint). choice = which of the two news items the respondent would choose to read (paraphrase; wording not in the deposit), forced choice, no opt-out, 10 tasks per respondent. Partisans only (cov_dem / cov_rep). Two attributes: headline text and outlet; the two items share a topic in about 9% of tasks and an outlet in about 34%, with no restriction documented. Qualtrics response ids re-keyed; the author’s derived topic and friendly/unfriendly-source dummies dropped. | 2026-10-08 | |
mummolo_2017_community_choice |
Conjoint table (irw_conjoint_2). Forced choice between two randomly generated communities (Community A/B), 5 tasks per respondent, 7 attributes fully randomized; choice = preferred community (paraphrase of the paper), no opt-out. 16 task-pairs with missing choice dropped. Housing cost levels are 15/30/40 percent in the data, 15/25/40 in the paper’s text; the deposit’s second experiment (single trait labels, no profiles) is not included. | 2026-10-14 | |
mummolo_2021_partisan_loyalty |
choice = which of “Candidate 1” or “Candidate 2” the respondent chose in each of 7 tasks; forced, no opt-out, no rating. Attribute row order was randomized per respondent and is recorded in attrpos_*. Repeat survey takers reduced to their first response (3,073 -> 3,015 respondents), undocumented items dropped; the paper’s N could not be checked and Study 2 is not included because its outcome file is not deposited. | #2887 | 2026-10-07 |
munzert_2022_vaccine_priority |
Conjoint table (irw_conjoint). choice: ‘Which person would you prioritize for access to the vaccine?’ Person A / Person B, forced choice. 4 tasks x 2 persons, 7 attributes, completely randomized; job ‘Teacher’ only in the German survey. Five countries in one table (cov_country); levels in English, citizenship as authors’ labels (‘Country’ = citizen). Repeated rows of 237 respondents removed; 120 tasks without a choice dropped (4,354 respondents vs 4,366 surveyed). | 2026-10-09 | |
myers_2024_welfare_stereotypes |
Conjoint table (irw_conjoint). One profile per task: rating = how typical the person described is of welfare recipients, 0-10 (higher = more typical; end labels not in the deposit); no choice question. Respondents saw one of two instruction versions (trial_instruction_code 1/2; wording not deposited). All 1,893 respondents with ratings are kept; the article analyses the 1,293 non-Hispanic whites. 94 respondents rated fewer than 30 profiles. | 2026-10-08 | |
myrick_2019_covert_exp1 |
Conjoint table (irw_conjoint_2). Factorial vignettes, two per respondent (dictator and civil-war scenarios; trial_scenario): region x military force/money-and-weapons x secret/informed x success/failure. rating is approval before the outcome (1 disapprove strongly .. 7 approve strongly), rating_after approval after it; attr_outcome had not been shown when rating was given. Secrecy is a within-respondent cross-over; task order (random) is not recorded, so task 1 = dictator is by numbering only. | 2026-10-11 | |
myrick_2019_covert_exp2 |
Conjoint table (irw_conjoint_2). Factorial vignettes, two per respondent (dictator and civil-war scenarios): region x public-opinion statement (polls show >80% opposition, or not shown) x secret/informed x success/failure; the action is fixed (small military force). rating is approval before the outcome (1 disapprove strongly .. 7 approve strongly), rating_after approval after it; task order is recorded; secrecy is a within-respondent cross-over. | 2026-10-11 | |
myrick_2019_covert_exp3 |
Conjoint table (irw_conjoint_2). Factorial vignettes, two per respondent (dictator and civil-war scenarios): region x force/money-and-weapons x informed vs secrecy kept despite experts agreeing it gave no advantage x success/failure. rating is approval before the outcome (1 disapprove strongly .. 7 approve strongly), rating_after approval after it. Secrecy is a cross-over; task order (random) not recorded; one age ‘99 +’ set missing. | 2026-10-11 | |
myrick_2019_covert_exp4 |
Conjoint table (irw_conjoint_2). Factorial vignettes, two per respondent (dictator and civil-war scenarios): region x force/money-and-weapons x secret/informed x success/failure, all shown before a single approval rating (1 disapprove strongly .. 7 approve strongly). Secrecy is a cross-over; task order (random) is not recorded, so task 1 = dictator is by numbering only. | 2026-10-11 | |
nadar_2024_art_pedodontists |
Conjoint table (irw_conjoint_2). Single-profile factorial vignettes: each pedodontist rated 11 fixed child-patient scenarios (age 4/6/12, cooperation, vulnerability); rating = how likely to use ART, 1 = very likely to 5 = very unlikely, stored raw (lower = more likely). Fixed design shared by all respondents; task is the scenario number, display order not documented. 432 respondents (matches the article); 4,748 rows because a few scenario ratings are missing. | 2026-10-12 | |
nagawa_2026_aid_departments |
Conjoint table (irw_conjoint). Three paired forced-choice tasks (3 attributes, displayed sentences); choice = preferred department (paraphrase), no opt-out. Follow-up hours-of-effort items (asked only for the chosen profile) are stored as unlabelled codes and not kept. N = 559 after dropping 3 incomplete surveys, as in the paper. | 2026-10-09 | |
nagawa_2026_aid_projects |
Conjoint table (irw_conjoint). Three paired forced-choice tasks (4 attributes, displayed sentences); choice = preferred project (paraphrase), no opt-out. Follow-up hours-of-effort items (asked only for the chosen profile) are stored as unlabelled codes and not kept. N = 559 after dropping 3 incomplete surveys, as in the paper. | 2026-10-09 | |
nair_2022_covax_agreement_lucid |
Conjoint table (irw_conjoint). Four pairs of international vaccine agreements; choice = ‘Which of these two agreements do you prefer?’ (forced, no opt-out; 36 tasks without a choice keep their ratings), rating = likelihood of voting for the agreement, definitely against (1) to definitely in favor (7). Levels drawn uniformly and independently; N = 1,749 (paper 1,751). | 2026-10-09 | |
nair_2022_covax_agreement_norc |
Conjoint table (irw_conjoint). Two pairs of agreements; choice = ‘If you had to choose, which agreement do you favor?’ (forced, no opt-out; 277 tasks without a choice keep ratings), rating = likelihood of voting for the agreement, asked 1-5 (definitely against to definitely in favor) but deposited as the authors’ 0-1 rescale (0, .25, .5, .75, 1). Duration levels 1-9 years in the data vs 1-5 in the SM; derived implied-burden column dropped; N = 4,204 (paper 4,214). | 2026-10-09 | |
nair_2022_vaccine_recipients |
Conjoint table (irw_conjoint). Five pairs of people waiting for a COVID-19 vaccine; choice = ‘If you had to choose between them, which of these two people should be given priority to receive the vaccine?’ (forced, no opt-out), rating = 1-7 (1 should definitely not, 7 should definitely receive the vaccine). Levels drawn uniformly and independently; N = 1,747 in the file vs 1,751 in the paper (25 respondents with fewer than 5 tasks). | 2026-10-09 | |
neisser_2026_mp_outside_income |
Conjoint table (irw_conjoint). Single-profile ratings of hypothetical Bundestag members, 1 = stimme gar nicht zu .. 7 = stimme voll und ganz zu: rating_voter_interest (primarily represents voters’ interests), rating_competent, rating_own_interest (primarily represents own or third-party interests; higher = less favourable, kept as asked), rating_hardworking; wording paraphrased from truncated labels. Outside activity and income shown as one line (13 levels; ‘none’ 25%, others about 6% each); party shares follow 2021 vote shares. 3 respondents failing the attention check kept with cov_attention_pass = 0; cov_survey_weight = the authors’ weight. | 2026-10-07 | |
nelson_2020_covid_job_offers |
Conjoint table (irw_conjoint). choice = which of the two job offers was most attractive / more likely to be accepted (paraphrase), forced choice, no opt-out; rating = attractiveness of each offer, 1 = Not at all attractive to 5 = Very attractive (the authors’ recode of the raw Qualtrics codes). Attributes fully randomized (article); level text from the article’s example trial and notes, with the two Trump political-climate levels and two location levels not seen displayed. Demographics other than birth year, 7-point party ID and COVID concern dropped (unlabelled); N = 984 (article: about 1,000). | 2026-10-07 | |
nelson_2022_backsliding_mturk |
Conjoint table (irw_conjoint). choice is a forced choice between two job offers and rating a 5-point rating of each offer (the authors’ recode of the raw Qualtrics codes; higher = more favourable inferred; wording not deposited). Attribute levels are the authors’ labels, and the backsliding levels are their code names for statements respondents read. IP addresses, locations and Qualtrics IDs dropped; unanswered tasks omitted. | 2026-10-07 | |
nelson_2022_backsliding_students |
Conjoint table (irw_conjoint). choice is a forced choice between two job offers and rating a 5-point rating of each offer (the authors’ recode of the raw Qualtrics codes; higher = more favourable inferred; wording not deposited). Attribute levels are the authors’ labels, and the backsliding levels are their code names for statements respondents read. IP addresses, locations and Qualtrics IDs dropped; unanswered tasks omitted. | 2026-10-07 | |
nelson_2025_abortion_jobs |
Conjoint table (irw_conjoint). Ten paired job-offer tasks with nine displayed-text attributes, order randomized per respondent; choice = which job the respondent would take (paraphrase; forced, no opt-out), rating = each job on a 4-point scale stored raw, 1 = favourable end (direction inferred from agreement with the choice; anchors not deposited). YouGov weight in cov_survey_weight; N = 1,200. | 2026-10-09 | |
nettle_2024_welfare_schemes |
Conjoint table (irw_conjoint_2). Forced choice of the preferred welfare policy of two (no opt-out; no rating collected). Payment size (child/adult/pensioner) and income tax (three rates) are fixed bundles, so those two attributes have restricted combinations; level text is the authors’ short labels. Prolific ID and free text dropped; one respondent with no choices removed. N = 800 respondents, as in the paper. | 2026-10-11 | |
neundorf_2026_electoral_autocracy |
Conjoint table (irw_conjoint). Forced choice of the country the respondent would prefer to live in (no opt-out); rating_democracy (‘How would you rate the level of democracy in each country?’, 0 very low - 10 very high, stored as the authors’ 0-1 rescale) and rating_quality_of_life (0-1 rescale). Attribute levels are the authors’ short binary labels, not the displayed sentences. 41 tasks with a missing attribute dropped; partial completers kept; 35,292 respondents with a choice vs 35,281 reported. | 2026-10-09 | |
neuner_2022_populist_amended_2018 |
Conjoint table (irw_conjoint). choice: forced choice of the candidate the respondent would rather vote for (article paraphrase), no opt-out. First and second priority are drawn from one list and never identical (restriction). Task and pair order inferred from row order. cov_survey_weight is the panel weight; thin-populism item wording/direction not deposited. Amended arm of the 2018 sample; the other half is neuner_2022_populist_candidates_2018. | 2026-10-07 | |
neuner_2022_populist_candidates_2017 |
Conjoint table (irw_conjoint). choice: forced choice of the candidate the respondent would rather vote for (article paraphrase), no opt-out. First and second priority are drawn from one list and never identical (restriction). Task and pair order inferred from row order. cov_survey_weight is the panel weight; thin-populism item wording/direction not deposited. | 2026-10-07 | |
neuner_2022_populist_candidates_2018 |
Conjoint table (irw_conjoint). choice: forced choice of the candidate the respondent would rather vote for (article paraphrase), no opt-out. First and second priority are drawn from one list and never identical (restriction). Task and pair order inferred from row order. cov_survey_weight is the panel weight; thin-populism item wording/direction not deposited. Replication arm of the 2018 sample; the other half is neuner_2022_populist_amended_2018. | 2026-10-07 | |
neuner_2025_populism_uk |
Conjoint table (irw_conjoint). Text vignette: statements of two hypothetical politicians; choice_listen/_crooked/_immigration/_conservative = which politician is more likely to say that politicians should always listen to the people / many in the political class are crooked / immigration should be reduced / to have conservative ideology (forced, no opt-out); rating_listen/_crooked/_immigration 1 = not at all likely to 5 = extremely likely; rating_ideology 1-5 left-right (UK wording; end labels unavailable). Control levels (no people-centrism or anti-elitism statement) and the undisplayed party are ‘(not shown)’. Finished respondents with any answer: 690. | 2026-10-08 | |
neuner_2025_populism_us1 |
Conjoint table (irw_conjoint). Text vignette: statements of two hypothetical politicians; choice_listen/_crooked/_immigration/_conservative = which politician is more likely to say that politicians should always listen to the people / many in the political class are crooked / immigration should be reduced / to have conservative ideology (forced, no opt-out); rating_listen/_crooked/_immigration 1 = not at all likely to 5 = extremely likely; rating_ideology 1-5 1 = very liberal to 5 = very conservative. Control levels (no people-centrism or anti-elitism statement) and the undisplayed party are ‘(not shown)’. Finished respondents with any answer: 699. | 2026-10-08 | |
neuner_2025_populism_us2 |
Conjoint table (irw_conjoint). Text vignette: statements of two hypothetical politicians; choice_listen/_crooked/_immigration/_conservative = which politician is more likely to say that politicians should always listen to the people / many in the political class are crooked / immigration should be reduced / to have conservative ideology (forced, no opt-out); rating_listen/_crooked/_immigration 1 = not at all likely to 5 = extremely likely; rating_ideology 1-5 1 = very liberal to 5 = very conservative. Control levels (no people-centrism or anti-elitism statement) and the undisplayed party are ‘(not shown)’. Finished respondents with any answer: 704. | 2026-10-08 | |
nguyen_2021_farm_subsidies_ch |
Conjoint table (irw_conjoint). choice = “Which proposal do you prefer?” (forced, no opt-out); rating = each proposal rated on a seven-point scale (1 = not like .. 7). Goal row order randomized but not recorded; levels from the article’s English text (0/1/2 = None/Moderate/Large); task/profile inferred from the source row index and verified. Deposit N 2,843 vs the article’s 2,310 valid answers; nothing dropped. | 2026-10-09 | |
nguyen_2021_farm_subsidies_us |
Conjoint table (irw_conjoint). choice = “Which proposal do you prefer?” (forced, no opt-out); rating = each proposal rated on a seven-point scale (1 = not like .. 7). Goal row order randomized but not recorded; levels from the article’s English text (0/1/2 = None/Moderate/Large); task/profile inferred from the source row index and verified. Deposit N 2,854 vs the article’s 2,605 valid answers; nothing dropped. | 2026-10-09 | |
nguyen_2024_cpd_benefits |
The deposit has no codebook, questionnaire or paper: this table is the deposit’s B1-B8 block (expected benefits of CPD), read from the English header abbreviations. The 1-6 response anchors are undocumented, and the four covariates are the source’s unlabelled codes. | 2026-10-09 | |
nguyen_2024_cpd_conditions |
The deposit has no codebook, questionnaire or paper: this table is the deposit’s C1-C12 block (enabling conditions; C1-C6 inside the school and C7-C12 the same six outside it), read from the English header abbreviations. The 1-6 response anchors are undocumented, and the four covariates are the source’s unlabelled codes. | 2026-10-09 | |
nguyen_2024_cpd_motivation |
The deposit has no codebook, questionnaire or paper: this table is the deposit’s A1-A9 block (reasons for joining CPD), read from the English header abbreviations. The 1-6 response anchors are undocumented, and the four covariates are the source’s unlabelled codes. | 2026-10-09 | |
nguyen_2024_cpd_willingness |
The deposit has no codebook, questionnaire or paper: this table is the deposit’s D1-D5 block (D1-D2 support received and given, D3-D5 willingness to chair, operate and present; kept together as the deposit groups them), read from the English header abbreviations. The 1-6 response anchors are undocumented, and the four covariates are the source’s unlabelled codes. | 2026-10-09 | |
nicholson_2023_paediatric_care |
Conjoint table (irw_conjoint). Discrete choice experiment: choice = preferred health service (Service A or B) for a child who is unwell, forced choice with no opt-out; fixed Bayesian-efficient design of 24 choice sets in 2 blocks of 12. Appointment-level text from the article (fielded grid not deposited); 458 respondents kept, the article analyses 450. | 2026-10-09 | |
nicoli_2025_defense_union |
Conjoint table (irw_conjoint). choice = which of the two security packages the respondent preferred (forced choice; wording paraphrased in the article); rating = each package on 1 strongly against - 5 strongly in favor. Attribute text is the English master (Table 1), not the national-language text shown; attribute order randomized per respondent but not recorded. Five countries pooled (cov_country, cov_survey_weight). Energy-framing arm of the same survey not included. 3,839 respondents vs 3,750 in the article; 2,893 pass the attention check. | 2026-10-07 | |
noble_2024_qanon_candidates |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent preferred (no opt-out; 4 tasks without a recorded choice keep their ratings); rating_favor: 7-point favorability (higher = more favorable, anchors not documented); rating_ideology: perceived ideology 1 extremely liberal to 7 extremely conservative. Each respondent answered one of the two ratings (346 favorability, 344 ideology). Exact question wording, sample source and fielding date are not documented in the deposit. | 2026-10-09 | |
nukui_2025_female_successors |
Conjoint table (irw_conjoint_2). One vignette per respondent (2 x 2: successor gender Man/Woman x age 32/65 years old, block-randomized); rating = willingness to nominate the person as successor, rating_support = perceived acceptance by the respondent’s supporters, rating_electability = perceived ability to win, each 1-4 with 4 most favourable (codebook wording, translated). No respondent id (id = row number; rows shuffled by the authors); level text from the codebook’s English labels, full Japanese vignette wording not deposited; covariates are the authors’ coarsened versions. N = 7,209 (paper: over 7,000). | 2026-10-09 | |
nyholt_2024_local_candidates |
Conjoint table (irw_conjoint). rating = “How likely is it that you would vote for Candidate A [B]?” reversed from the source to 1 = very unlikely .. 5 = very likely (Ved ikke set missing); choice = “Hvilken kandidat foretrækker du?” with a Ved ikke opt-out (1,801 of 5,105 tasks have no chosen profile). Behavioral and symbolic localism were omitted for a random subset of profiles (“(not shown)”); five long localism levels are truncated at 80 characters in the deposit’s labels. Gender, age, occupation and party follow the 2019 candidate distribution (non-uniform). All 1,021 respondents kept; the article analyses 779 after dropping speeders (cov_duration_sec < 180 s, 195 s with the prime, trial_prime). | 2026-10-07 | |
ocampo_2025_latinos_american |
Conjoint table (irw_conjoint). Five paired profile tasks; choice = ‘If you had to choose between them, which of them do you personally believe is more American?’ (forced, no opt-out); rating = how American each person is, 1-7 (7 = very much American); 21 skipped ratings (998) set NA. Restrictions (education x profession, background x language, faces x identifies) and unequal level probabilities for party and spouse per the authors’ cjoint design; survey weight kept. White and Black samples pooled (cov_race); N = 1,500. | 2026-10-09 | |
ofosu_2026_malawi_women_candidates |
Conjoint table (irw_conjoint). choice: “Which of these two candidates would you vote for as your MP?”; choice_others: “Which of these two candidates do you think others in your constituency would vote for as their MP?” Both forced choice, asked in random order (trial_self_first). Tasks with no recorded answer to one question have that outcome missing (25 and 56 tasks); tasks with neither are omitted. Attribute text is the authors’ English labels; party = constituency winner (Major) / runner-up (Minor) / Independent, the actual names shown not deposited. trial_video is the randomized pre-conjoint video. Pilot dropped; 2,233 of the paper’s 2,239 have answers. | 2026-10-07 | |
ogrady_2026_immigration_sanctions |
Conjoint table (irw_conjoint_2). One text vignette per respondent about a migrant to the UK (name code, reason: asylum/economic/family, sanction none or severity 1-3), YouGov June 2025. Four 1-4 ratings (support for detention on a 1-5 scale; migrant and self likelihood of migrating; fairness of the immigration system), 1 = strongest support/most likely/fairest; don’t know set to NA; no opt-out. Name and sanction levels are tags (the deposit has no text); reason labels come from the authors’ figure code. Weight kept. | 2026-10-14 | |
ogrady_2026_welfare_sanctions |
Conjoint table (irw_conjoint_2). One text vignette per respondent about a UK Universal Credit claimant (name, reason for redundancy, sanction none or severity 1-3), YouGov July 2025. Four 1-4 ratings (support for sanctions on a 1-5 scale; claimant and self likelihood of accepting a job; fairness of the JobCentre), 1 = strongest support/most likely/fairest; don’t know set to NA; no opt-out. Sanction levels are ordinal tags because the deposit has no text for them; an undocumented base-text variant is kept as trial_basetext_variant. Weight kept. | 2026-10-14 | |
ogura_2022_partisan_labels |
Conjoint table (irw_conjoint). choice: which of two persons is more of a Democrat (Democrat arm) or Republican (Republican arm), forced, no opt-out; rating: 7-point partisanship rating of each person, kept raw (anchors not in the deposit; the authors’ code implies 1 = Democrat end, 7 = Republican end in both arms). Wording paraphrased (Online Appendix A not deposited). Attribute order randomized per respondent (attrpos_). All 1,051 respondents kept with trial_page_seconds and two attention-check covariates; the authors analyse 864 attentive respondents. | 2026-10-08 | |
ohmura_2022_covid_vaccination_lucid |
Conjoint table (irw_conjoint). Forced choice of the more desirable of two situations (‘あなたにとって、以下の2つの状況のうちどちらが望ましいですか。’), 5 tasks, no opt-out; the ‘familiar’ attribute refers to the person the respondent picked beforehand (cov_familiar_entity). Level text is the author’s English labels (‘wait and see’ = ‘do not vaccinate now, vaccinate later’), not the Japanese shown. Unanswered tasks dropped; N 1,024 matches the article. Age (cov_age) is the free-entry age kept only when a whole number 18-110 (typed birth dates and text set to NA); gender and education codes are unlabelled. | 2026-10-08 | |
ohmura_2022_covid_vaccination_ycs |
Conjoint table (irw_conjoint). Forced choice of the more desirable of two situations (‘あなたにとって、以下の2つの状況のうちどちらが望ましいですか。’), 5 tasks, no opt-out; the ‘familiar’ attribute refers to the person the respondent picked beforehand (cov_familiar_entity). Level text is the author’s English labels (‘wait and see’ = ‘do not vaccinate now, vaccinate later’), not the Japanese shown. Unanswered tasks dropped; Article reports 2,975 participants; the file has 2,971, of whom 37 lack conjoint fields and 86 more answered no task: 2,848 kept. Age (cov_age) is the free-entry age kept only when a whole number 18-110 (typed birth dates and text set to NA); gender and education codes are unlabelled. | 2026-10-08 | |
okamura_2018_burnout |
cov_age is an age-band code, not age in years: 2 = “20代”; 3 = “30代”; 4 = “40代”; 5 = “50代”; 6 = “60代”; 7 = “70代”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
okamura_2018_eol_attitude |
cov_age is an age-band code, not age in years: 2 = “20代”; 3 = “30代”; 4 = “40代”; 5 = “50代”; 6 = “60代”; 7 = “70代”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
oluocharidi_2020_delivery_facility |
Conjoint table (irw_conjoint). Discrete choice experiment: choice = which delivery facility (A or B) the woman would prefer, with option C ‘None of the two health facilities - home delivery’ as an opt-out (794 of 7,456 tasks; both profiles 0). 16 choice sets x 2 facilities, 6 attributes (5 binary, cost Ksh3000/5000/8000), Ngene D-efficient design in 2 blocks; the article says 8 sets per respondent but the deposit has 16. Level text from the article’s sample card. Some covariate codes do not match the Readme (kept as _code). | 2026-10-09 | |
ono_2019_candidate_sex |
Conjoint table (irw_conjoint). choice = which candidate the respondent would vote for (Congress wording: ‘If you had to choose between them, which of these candidates would you vote to be a member of the U.S. House of Representatives?’; presidential wording not deposited), forced choice, no opt-out. Attributes independently randomized (level probabilities not stated); attribute order randomized per respondent, not recorded. Tasks 1-5 are President and 6-10 Congress for all, though the block order was randomized, so task order is not display order. | 2026-10-07 | |
ono_2020_gender_stereotypes |
Conjoint table (irw_conjoint). choice = which of two same-party candidates the respondent would vote for (‘Which of the two candidates would you like to vote for? Even if you are not entirely sure, please indicate which of the two you would prefer if you had to choose either one of them.’, translated from Japanese), forced choice, no opt-out. All 1,152 profiles possible (no restrictions); attribute order randomized per respondent, not recorded; attribute text is the authors’ English translation. N = 2,686 as in the paper. | 2026-10-08 | |
ono_2025_bicameral_gender |
Conjoint table (irw_conjoint). rating: ‘To what extent is this candidate favorable as a [HoR/HoC] member?’ 8-point scale, 1 = not favorable at all, 8 = very favorable (authors’ translation of the Japanese item); single profile per task, 10 tasks per chamber, chamber block order randomized (trial_chamber; task = order shown, rebuilt from HOR_first as in the authors’ code); half of respondents read a priming paragraph on the lower house’s supremacy (trial_priming). Level probabilities follow real-world distributions (nonuniform); attribute order randomized per respondent (attrpos_). Hometown text keeps the template X (respondent’s prefecture was inserted). N 2,267 matches the paper; inattentive respondents were removed by the authors before deposit. | 2026-10-09 | |
ono_2026_multiple_vote |
Conjoint table (irw_conjoint_2). Six-candidate municipal-election tasks (5 per respondent), ballot rule randomized: choice_sntv = ‘vote for one’ (SNTV arm), choice_bv_first/second/third = the 1st/2nd/3rd candidates named under a three-vote ballot (BV arm), wording in Japanese from the Qualtrics export; no opt-out, no repeated picks. Hometown levels keep the piping structure with a placeholder for the respondent’s municipality. Attribute order randomized per respondent; level shares unequal (not documented in the deposit). N = 5,400 (1,456 SNTV, 3,944 BV), matching the authors’ preprocessing. | 2026-10-09 | |
opladen2025_edeq |
Items EDEQ_13 to EDEQ_18 are the EDE-Q’s behavioural frequency questions (number of episodes or days over the past 28 days), so resp is a self-reported count (here 0 to 112), not a rating like the other items (stored 1-7 here). The EDE-Q’s subscale and global scores use only the 22 rating items. | #1700 | 2026-10-01 |
ornstein_2024_hometown_advantage |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent chose in a hypothetical nonpartisan local election (wording not deposited); forced, no opt-out. Task order inferred from row order (five stacked blocks of the same respondents); profile recorded. The article says levels were uniform with no restrictions, but White-coded names appear about three times as often as each Black- or Latino-coded name. CES covariates kept as codes, except cov_gender and cov_party_id (text); cov_survey_weight blank for 308 respondents. | 2026-10-07 | |
oviedo_2026_waterbirds |
Conjoint table (irw_conjoint). Discrete choice experiment: “Please, indicate the program that you would choose” PROGRAM A / PROGRAM B / NO PROGRAM; the status quo NO PROGRAM shows its fixed outcomes and is profile 3, so exactly one of three is chosen. Species-change text rebuilt from the coded levels with the questionnaire template (English version); tax amount as coded (displayed currency format unknown). Fixed blocked design. Spain and Portugal pooled (cov_country). | 2026-10-09 | |
pan_2023_hiring_managers |
Conjoint table (irw_conjoint). rating = reported callback for each resume (1 = would call back, 0 = no), each resume judged on its own; no choice question. Study-group attribute blocked within respondent (each saw comic, socialism and Western groups once); other attributes binary. Level text is paraphrase from the ReadMe; task order inferred from row order; income missing for 36 respondents. | 2026-10-07 | |
pan_2023_resume_pretest |
Conjoint table (irw_conjoint). Three ratings of each applicant, 1-10, higher = more: rating_merit (academic merit), rating_connections (strength of government connections), rating_conformity (political conformity to the CCP regime); no choice question. Three binary attributes; randomization rules not documented. Wording is the ReadMe’s paraphrase; task order = profile letter, display order undocumented. | 2026-10-07 | |
pan_2025_crossstrait_agreement |
Conjoint table (irw_conjoint). Five single agreements per respondent; choice = support (1) or oppose (0) the cross-Strait agreement (paraphrase; rejecting is the outside option, opt_out = yes). Each of 8 concessions shown with probability ~1/2, ‘(not shown)’ otherwise, never all hidden; respondent-level US-support/China-threat framing in trial_us_china_frame; task order inferred from row order; N = 3,446 vs 2,905 in published summaries. | 2026-10-09 | |
pan_2025_diplomatic_visits |
Conjoint table (irw_conjoint). choice: which of two politicians the respondent prefers as a national security defender (paraphrase; questionnaire not deposited); an opt-out is inferred from the data (both profiles never chosen; 5,105 of 11,225 tasks with neither chosen, choice = 0 on both). Attribute levels are the deposit’s English value labels; respondents saw Traditional Chinese. Covariates are the authors’ coarsened versions (born before 1978, college, residence in the 6 capitals, perceived relations). | 2026-10-08 | |
panel_2023_technocrats_de |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. | 2026-10-08 | |
panel_2023_technocrats_es |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. | 2026-10-08 | |
panel_2023_technocrats_fr |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. | 2026-10-08 | |
panel_2023_technocrats_it |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. Three tasks with blank attribute text (2 respondents) dropped. | 2026-10-08 | |
panel_2023_technocrats_pl |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. | 2026-10-08 | |
panel_2023_technocrats_uk |
Conjoint table (irw_conjoint). Forced choice of the preferred of two hypothetical ministers (codebook: ‘chosen as the respondent’s preferred profile’; wording not deposited), 5 tasks, one per post (trial_post); no opt-out. Level text is the deposit’s English; occupation and policy-position levels are specific to the post, and education is ‘Phd’ on every profile. Respondent ids re-keyed from Qualtrics ResponseId. | 2026-10-08 | |
panth_2026_bipq |
Responses are as deposited; nothing is reversed. In the B-IPQ’s own scoring items 3, 4 and 7 (personal control, treatment control, understanding) run opposite to the other five. Item 9 (open-ended causes) is not deposited. | #2940 | 2026-10-10 |
panth_2026_cdrisc10 |
One cell held 9 on a 0-4 scale (a single entry error) and was dropped. The deposit’s serial number is duplicated once (two different respondents both numbered 153), so id is the row order of the deposit, not its serial number; the same ids are used in panth_2026_bipq. | #2940 | 2026-10-10 |
papp_2023_covid_vaccine_hungary |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical COVID-19 vaccines in Hungary (Sept 2022 online panel, quotas on vaccination status and party preference); choice = the vaccine selected in the task (forced, no opt-out). Level text comes from Stata value labels and is a short form of the codebook wording (e.g. severe side effects ‘Rare’). Randomization restrictions and attribute-order randomization are not documented; the question wording is a paraphrase. 324 respondents, 3,888 profiles, matching the paper. | 2026-10-12 | |
papp_2024_democracy_satisfaction |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical political systems fielded identically in Italy, Hungary and France (pooled, cov_country). choice = in which of the two countries the respondent would be more satisfied with the way democracy works (forced, no opt-out); rating = 0-10 satisfaction with democracy in each country (higher = more satisfied), stored raw. Randomization restrictions and attribute-order randomization are not documented; the deposit has no questionnaire, so the rating wording is a paraphrase. | 2026-10-12 | |
pardosprado_2026_nhs_tax |
Conjoint table (irw_conjoint). choice = “Generally speaking, if you had to choose, which one of these two options would you prefer?” (Option A / Option B / Don’t know; Don’t know is an opt-out, 3,805 tasks with no choice); rating = “how do you feel about it?” for each option, 1 like it a lot to 4 don’t like it at all (Don’t know -> NA). Tax rates across the five income brackets were restricted to non-decreasing schedules. | 2026-10-09 | |
parkins_2025_forest_disturbance |
Conjoint table (irw_conjoint). Factorial vignette: ‘Please indicate how acceptable the following management scenario are to you … completely acceptable (+5) to completely unacceptable (-5)’, stored as codes 1-11 (1 = -5 completely unacceptable, 11 = +5 completely acceptable). 6 of 54 fixed scenarios per respondent; display order not recorded, so task follows scenario number. Panel (2,569) and non-panel (37) samples pooled with cov_sample. | 2026-10-09 | |
pedersen_2019_candidate_background |
Conjoint table (irw_conjoint_2). One text vignette per respondent; rating = likelihood of voting for the candidate 0 (very unlikely) - 10 (very likely); rating_left_right = candidate placement 0-10 (0 = left); rating_ |
2026-10-09 | |
pedersen_2023_pay_for_performance |
Conjoint table (irw_conjoint). choice: “Which of the two jobs would you personally prefer?” Job A / Job B, forced, no opt-out. rating: “To what extent do jobs A and B reflect your ideal job arrangement?” 1 (Definitely not ideal) to 7 (Definitely ideal), each job. 3 pairs per respondent, 8 attributes drawn uniformly and independently, fixed order. White and Non-White Prolific samples pooled (cov_sample); 7 failed the attention check (cov_attention_pass), all kept. | 2026-10-08 | |
pedersen_2025_mp_tasks_uk |
“Which of the two candidates would you be most inclined to vote for…” (Candidate 1/2, forced, no opt-out, 4 tasks); young candidates (25-34) always have 1-3 years’ experience; attribute order not recorded. Levels are the deposit’s English master labels; respondents who did not consent (no conjoint) are dropped, so 2,231 of the paper’s 2,849. | #2887 | 2026-10-07 |
pedersen_2025_mp_tasks_us |
“Which of the two candidates would you be most inclined to vote for…” (Candidate 1/2, forced, no opt-out, 4 tasks); young candidates (25-34) always have 1-3 years’ experience; attribute order not recorded. Levels are the English master labels (the US screens probably said House rather than Parliament); non-consenters dropped, so 2,185 of the paper’s 2,554. | #2887 | 2026-10-07 |
pelzer_2019_facebook_use |
Conjoint table (irw_conjoint). choice = which of two Facebook-use situations the respondent picks (question wording not deposited), forced, 7 tasks from one of 4 fixed blocks; 2 attributes, levels are the author’s English variable names, not the displayed German text. One respondent with inconsistent block data dropped (N = 529 of 530). | 2026-10-09 | |
pepinsky_2023_refugee_gender |
Conjoint table (irw_conjoint_2). Single-question 2x2 factorial vignette: rating = agreement that the government should allow the entry of male / female-and-children refugees from Afghanistan / Ukraine, 1 = strongly disagree to 5 = strongly agree; ‘does not know’ and refusals dropped (984 of 1,000). Levels are the value labels of the randomization variables (Hungarian wording not in the deposit); TARKI weight kept. | 2026-10-11 | |
perez_hernandez_2021_cesd7 |
Unlabelled code 4 dropped (7-18 responses per item). | 2026-10-10 | |
perez_hernandez_2021_cesd7 |
June 2020 wave added 2026-10-10 (ids 3000000+); June and October are different respondents. A re-upload must delete-then-recreate the table. | 2026-10-10 | |
perez_hernandez_2021_elcsa |
June 2020 wave added 2026-10-10 (ids 3000000+). A re-upload must delete-then-recreate the table. | 2026-10-10 | |
perez_hernandez_2021_gad7 |
October 2020 respondents answered the GAD-2 only (gad1, gad2); July respondents the full GAD-7. Unlabelled code 4 (October) and 9 NS/NR dropped. | 2026-10-10 | |
perez_hernandez_2021_gad7 |
June 2020 wave added 2026-10-10 (GAD-2 only, ids 3000000+). A re-upload must delete-then-recreate the table. | 2026-10-10 | |
perino_2023_meat_tax |
Conjoint table (irw_conjoint). Single-proposal referendum vote ‘Stimmen Sie diesem Vorschlag zu?’ (Ja / Nein / Ich möchte nicht abstimmen): choice = Ja, choice_against = Nein, abstention 0 on both (opt-out); 6 proposals per respondent in a fixed order of rising levy, justification and differentiation randomized between respondents, levy rows depend on the justification. The codebook’s vote coding (0 = yes) contradicts the authors’ code and the data; 1 = yes is used. All 3,169 kept; the article drops the fastest and slowest 5% by survey time. | 2026-10-08 | |
perlstein_2025_risk_conversations |
Conjoint table (irw_conjoint). choice: “In welk gesprek zou u liever willen deelnemen?” (which conversation would you rather take part in), Gesprek 1 / Gesprek 2, forced, no opt-out. 9 tasks: the intro example pair (answered, used by the authors) plus Taak 1-8; blank tasks dropped. Observed restrictions: a friend is never ‘niet hecht’; ‘niet bezorgd’ never with the ‘meer maatregelen’ or ‘minder ongerust’ motivations; ‘minder maatregelen’ never with ‘redelijk/zeer bezorgd’; level shares unequal. Attribute order randomized per respondent. | 2026-10-08 | |
perna_2025_healthcare_be |
Conjoint table (irw_conjoint). choice: which of two patients the respondent would prioritise to be visited today, forced, no opt-out; rating: 0 (not at all) - 10 (totally) how probable it is they would prioritise the chosen patient, asked of the chosen patient only. Codebook English labels; profile inferred from row order. One corrupt respondent dropped. | 2026-10-09 | |
perna_2025_healthcare_es |
Conjoint table (irw_conjoint). choice: which of two patients the respondent would prioritise to be visited today, forced, no opt-out; rating: 0 (not at all) - 10 (totally) how probable it is they would prioritise the chosen patient, asked of the chosen patient only. Codebook English labels; profile inferred from row order. | 2026-10-09 | |
pertiwi_2025_educational_fashion |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent scored 16 fixed children’s-clothing cards (theme, size, colour, purchase channel) on a 1-5 scale stored raw; the deposit gives no question wording or scale labels. Fixed 16-card subset of the 48 combinations (M appears on 8 cards, L and XL on 4 each); task = card number in column order (display order not documented). Covariates are two Indonesian-text purchase-habit answers. | 2026-10-12 | |
petersen_2025_online_abuse |
Conjoint table (irw_conjoint_2). Each respondent saw four single comments (task 1 criticism, 2 insult, 3 threat, 4 sexism; comment within type random) about a fictitious politician; ratings 1-7 (abusive; politicians should tolerate comments like this; would report this message), 99 = don’t know set to NA, no choice or opt-out. Gender is carried by the politician’s name only; comments 19-20 keep the template ‘[woman/man]’; comment-rows with all three outcomes missing are omitted; age is truncated at 81. Paper reports about 2,000 respondents; the file has 2,050 rows, 1,985 retained. | 2026-10-11 | |
peterson_2017_information_environment |
Conjoint table (irw_conjoint). choice = chosen candidate, forced choice; question wording not in the deposit. Party always shown, plus a random 1-9 of nine other attributes per task (‘(not shown)’ = not shown; trial_information_condition); party level rebuilt from the authors’ copartisan flag. Level shares are unequal (e.g. Male 67%, Lawyer 30%), not documented. 69 tasks with no recorded choice dropped (1,059 -> 1,037 respondents); N vs the paper not checked (paywalled). | 2026-10-07 | |
peterson_2024_economic_statecraft |
Conjoint table (irw_conjoint). Single-profile vignette; six statements rated on a 0-100 slider (anchors not in the deposit): ‘I support the proposed policy’ (rating), advances US interests, too costly to the US (higher = more costly), effective in changing the target’s behaviour, demonstrates US strength, sends a strong message about US commitment. Three experiments pooled (trial_experiment); level text names the target country, so background, rationale and inaction texts depend on behaviour and policy. 44 respondents without attributes or ratings dropped; demographic codes unlabelled. | 2026-10-09 | |
petley_2025_flanker_study1_childwoman |
Rows are expanded from per-condition counts (number correct out of trials); trial order is not observed, so the rows within a person and condition are exchangeable. Do not model order, practice or sequential effects. | #2401 | 2026-09-29 |
petley_2025_flanker_study1_girlman |
Rows are expanded from per-condition counts (number correct out of trials); trial order is not observed, so the rows within a person and condition are exchangeable. Do not model order, practice or sequential effects. | #2401 | 2026-09-29 |
petley_2025_flanker_study1_manwoman |
Rows are expanded from per-condition counts (number correct out of trials); trial order is not observed, so the rows within a person and condition are exchangeable. Do not model order, practice or sequential effects. | #2401 | 2026-09-29 |
petley_2025_flanker_study2_auditory |
Rows are expanded from per-condition counts (number correct out of trials); trial order is not observed, so the rows within a person, condition and session (wave) are exchangeable. Do not model order, practice or sequential effects. |
#2401 | 2026-09-29 |
petley_2025_flanker_study2_visual |
Rows are expanded from per-condition counts (number correct out of trials); trial order is not observed, so the rows within a person, condition and session (wave) are exchangeable. Do not model order, practice or sequential effects. |
#2401 | 2026-09-29 |
petrovic_2026_ai_hiring_score |
Conjoint table (irw_conjoint). Choice-based conjoint, 8 tasks of 4 candidates. choice: forced pick among the four (no opt-out); choice_commit: 1 on the picked candidate if the participant then said they would actually advance that candidate, otherwise no candidate (392 of 3,456 tasks); wording not deposited. Level text is the design’s Serbian level names, including ‘no information’ levels for gender, education, age and hobby. Covariates in Serbian answer text; 1-5 attitude items raw, anchors not deposited. | 2026-10-09 | |
petrovici_2025_culture_appreciation |
The deposit codebook documents 1 Yes / 0 No, but the data file holds 1 and 2 (plus 99 don’t-know, dropped); shipped as deposited with 2 read as No. | 2026-10-11 | |
peyton_2022_police_recruit_officers |
Conjoint table (irw_conjoint). choice: which of two police applicants the officer would prefer to see recruited into the Yonkers PD (forced, no opt-out); rating: 1 (definitely not be recruited) to 7 (definitely be recruited) per applicant. 5 pairs; teachers/social workers always hold a bachelor’s or graduate degree (restricted randomization). Level text is the authors’ shortened labels. 250 officers, 4 with no answers dropped; no covariates in the deposit. | 2026-10-09 | |
peyton_2022_police_recruit_residents |
Conjoint table (irw_conjoint). choice: which of two police applicants the respondent would prefer to see recruited into the Yonkers PD (forced, no opt-out); rating: 1 (definitely not be recruited) to 7 (definitely be recruited) per applicant. 5 pairs; teachers/social workers always hold a bachelor’s or graduate degree (restricted randomization). Level text is the authors’ shortened labels. Yonkers residents; 1,412 in deposit vs 1,413 in paper, 3 with no answers dropped; raking weight kept but unused by the authors. | 2026-10-09 | |
pham_2026_rural_jobs |
Conjoint table (irw_conjoint). choice: ‘Between jobs A and B, which job do you prefer?’ forced A/B; choice_final: ‘Which job would you choose in practice?’ A, B or current job (opt-out: 588 of 2,835 tasks chose the current job, both profiles 0). Fixed blocked design of 20 cards in 2 blocks; the repeated reliability card was removed by the depositors, so 9 tasks per respondent with original positions kept; attribute text from the bilingual cards (English line); 315 respondents as in the README. | 2026-10-09 | |
philipps_2022_self_mediatization |
Conjoint table (irw_conjoint). Factorial vignette survey: each MP read one German-language negotiation vignette (2x2: decision pressure x election-campaign proximity); attribute text is the codebook’s English description, not the vignette prose. 23 rating_ outcomes on 1-5 scales (likelihood of publicizing via 10 channel/content combinations, of agreeing on confidentiality, of compromising; vignette realism; two manipulation checks whose anchors differ, see conj_outcomes). No choice, no opt-out; one row per respondent, id is the row number. | 2026-10-08 | |
phillips_2021_emental_health |
Conjoint table (irw_conjoint). DCE, choice: ‘Bitte wählen Sie eine der beiden Varianten eines Online-Therapieprogramms, die Sie bevorzugen würden:’ (OPTION A / OPTION B), no opt-out, 16 fixed choice sets (Bayesian D-efficient design, one block). German displayed level text from the codebook. Respondents were randomized to one of two context scenarios (trial_c_0001, code meaning undocumented). 20 unanswered tasks dropped; 2,004 respondents vs 1,984 analysed in the article. | 2026-10-09 | |
phillips_2022_blended_care |
Conjoint table (irw_conjoint). DCE. choice: ‘Bitte wählen Sie eines der beiden Anwendungsszenarien für Blended Care, das Sie bevorzugen würden:’ OPTION A / B, forced (no opt-out by design). Fixed blocked D-efficient design (trial_block; block 1 is the pre-change design seen by 37 respondents); German level text as displayed; free-text answers dropped; N = 200 as in the article. | 2026-10-09 | |
phillips_2025_political_violence |
Conjoint table (irw_conjoint). Five single-profile tasks; rating = ‘What punishment, if any, do you think they should receive?’ 1 = a warning … 13 = more than 30 years in jail (answer position, higher = harsher); rating_good_cause = ‘How likely do you think it is that this individual was protesting for a good cause?’ 1 = extremely unlikely to 5 = extremely likely. Task order inferred from row order; residence stored as the authors’ group label (state piped into displayed text not saved); perpetrator race nonuniform by design. The double-profile experiment in the deposit is not built (pairs not recoverable). N = 3,402 as in the article. | 2026-10-09 | |
pickering_2025_peacekeeping_japan |
Conjoint table (irw_conjoint). choice between missions A and B, rating 1 = not at all effective to 5 = extremely effective; exact wording not deposited. The source’s non-substantive code 8 is set missing. Note the deposit’s file names are swapped: replication_data_jp.csv holds the UK sample. Respondents saw Japanese; labels are the authors’ English ones. 2 respondents with no answers dropped. | 2026-10-07 | |
pickering_2025_peacekeeping_uk |
Conjoint table (irw_conjoint). choice between missions A and B, rating 1 = not at all effective to 5 = extremely effective; exact wording not deposited. The source’s non-substantive code 8 is set missing. Note the deposit’s file names are swapped: replication_data_jp.csv holds the UK sample. | 2026-10-07 | |
pickering_2026_ebike_policy |
Conjoint table (irw_conjoint). Four forced choices per pair of e-bike policy packages, no opt-out: choice = preferred standard, choice_increase_use = more likely to increase e-bike use, choice_more_restrictive = more restrictive for the individual, choice_more_intervention = requires more government intervention (the authors’ labels; question wording not in the deposit). Levels in English; display language not documented. cov_survey_weight kept. | 2026-10-08 | |
pickett_2024_officer_conjoint |
Conjoint table (irw_conjoint). choice: which officer in each pair would make the respondent the most afraid (paraphrase of the preprint; Officer A / Officer B), forced choice, 5 tasks. Levels randomized independently; attribute order randomized per respondent and held across tables (attrpos_). General-population and Black oversamples pooled with the authors’ combined weight; 14 unanswered tasks dropped. | 2026-10-09 | |
pickett_2024_officer_photos |
Conjoint table (irw_conjoint). Pictured two-officer team, race (White/Latino/Black) and sex of each officer and stop location (empty/busy street) randomized; one picture per respondent. rating_hurt/force/weapon/arrest: ‘Afraid these officers will – physically hurt you / use excessive force against you / point a weapon at you / wrongfully arrest you’, raw 1 = Very Afraid to 5 = Very Unafraid. Two samples pooled with the authors’ combined weight. | 2026-10-09 | |
pinedo_2026_genai_utaut2 |
Nine UTAUT2 construct blocks (PE, EE, SI, HM, PV, FC, HT, BU, UT) of one questionnaire shipped as one table; the item prefix is the construct. | 2026-10-11 | |
pinho_2026_ai_literacy |
The items AWA2_R, USA2_R and ETH2_R are stored already reverse-scored (each correlates positively with the other items of its dimension), so higher = more AI literacy on every item. The deposit has no codebook; construct names for all pinho_2026 tables come from the record’s description. | 2026-10-07 | |
pklmart_pickleball_2021_2025 |
agent_a is always the winning team (the source records only winner and loser, not side or serve), so winner is always agent_a; do not read an agent_a advantage into it. Agents are doubles teams keyed by their two pklmart player ids. Data are player-entered, mostly amateur games, and sparse: most teams appear in one game. | #2714 | 2026-10-01 |
polgahecimovich_2021_admissions |
Conjoint table (irw_conjoint). choice: forced choice of which of two applicants should be given priority in admission to the Naval Academy (wording paraphrased; not in the deposit), no opt-out. Level text as recorded by the survey; the appendix’s race level ‘Other’ never occurs; gender Non-binary appears on about 10% of profiles. One task with no recorded choice and two free-text items dropped. 582 respondents, as in the paper. | 2026-10-08 | |
polgahecimovich_2021_faculty |
Conjoint table (irw_conjoint). choice: “Which candidate do you think should be given priority in faculty recruitment?”, forced choice between two candidates, no opt-out. Level text as recorded; department has 9 levels in the data although the appendix lists 4, race ‘Other’ never occurs, gender Non-binary appears on about 10% of profiles. Two free-text items dropped. 572 respondents, as in the paper. | 2026-10-08 | |
portmann_2021_candidate_stereotypes |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent (Italy, 2019): candidate name (Italian or Algerian) x positive information (none = ‘(not shown)’, character traits, civic citizenship, both). Ratings: likelihood of voting for the candidate and support for nominating him (0-10), and overall evaluation (1 negative, 2 neutral, 3 positive, coded from the answer text). Level text is paraphrased from the article (vignette appendix closed); the deposit stores only short condition codes; 1,594 of the article’s 1,597 respondents have an outcome. | 2026-10-14 | |
portmann_2022_candidate_cues |
Conjoint table (irw_conjoint). choice = preferred candidate of two (codebook ‘Dummy whether respondent prefers candidate’; wording not deposited), one task, no opt-out; trial_competence_arm = 1 adds a competence/political-office sentence (attr_competence), ‘(not shown)’ in arm 0. Attribute text is the German Qualtrics field text (encoding repaired); the authors’ recoded policy factor mislabels candidate 2 in 745 of 1,000 tasks and the raw fields are used instead. Survey Experiment 1 (fixed-pair 2 x 2 vignette, wording not deposited) is not built. | 2026-10-09 | |
prasetyo_2024_massage_parlor |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent rated 27 fixed stimulus cards (6 attributes, Indonesian text) for a massage parlor service; rating = preference 1-7, 1 = strongly not preferred, 7 = strongly preferred, stored raw. Fixed orthogonal design shared by all respondents; task = card number in column order. Therapist gender has a third level ‘Terapis wanita/pria’. 212 respondents (matches the paper). | 2026-10-12 | |
prather_2020_foreign_aid_latino |
Conjoint table (irw_conjoint_2). One news-article vignette per respondent (id = row number); choice = should NOT cut the program (1) vs cut (0), a single-profile accept/reject (opt-out yes); rating_obligation (moral obligation) and rating_closeness are 1-4, higher = more agreement/closer. Levels from the article’s Figure 2; the deposit’s US and UK samples are not built (treatment codes undocumented). N = 1,000 matches the article. | 2026-10-11 | |
prilyasinta_2026_connectedness |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
prilyasinta_2026_identity |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
prilyasinta_2026_intensity |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
prilyasinta_2026_multilingualism |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
prilyasinta_2026_pilot_items |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
prilyasinta_2026_responsibility |
The pilot (n=126) and main (n=259) samples answered the same items; retained items are pooled with cov_study and pilot_/main_ id prefixes. Subscale membership is the deposit’s questionnaire PDF (main-study sections); the response points are not labelled in the deposit. | 2026-10-09 | |
project_kids_wj_ak_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
project_kids_wj_lwid_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
project_kids_wj_pc_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
project_kids_wj_pv_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
project_kids_wj_spell_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
project_kids_wj_wa_grade |
Woodcock-Johnson forms A and B are pooled under shared item names; the source records form per grade, but very few children (7 or fewer per subtest) took form B. | #2513 | 2026-09-28 |
prokop_2021_digital_services |
Conjoint table (irw_conjoint). rating_procedure / rating_result: “Wie zufrieden wären Sie… mit dem Ablauf? / mit dem Ergebnis?” on a 9-point scale stored as codes 1-9 (1 = -4 gar nicht zufrieden, 9 = 4 sehr zufrieden). Single-profile text vignettes, three per respondent, factors independently and evenly randomized per vignette; the third vignette pairs the self-service-terminal text with the desk picture (attr_image). 1,255 respondents with any rating (authors’ complete-case sample 1,234; cov_progress); IP, ResponseId and panel token dropped. | 2026-10-08 | |
PROMISPME_Forrest_2021_LS_Proxy |
56 of the 54,040 rows repeat an (id, item) pair already in the table, and some of these repeats give a different resp. They cannot be resolved: the source deposit (OSF f7rp3) no longer holds the file this table was built from. Every other row is one response per person-item. If your analysis needs unique (id, item) pairs, drop the repeated ones. | #1856 | 2026-09-30 |
PROMISPME_Forrest_2021_MP_Proxy |
55 of the 51,040 rows repeat an (id, item) pair already in the table, and some of these repeats give a different resp. They cannot be resolved: the source deposit (OSF f7rp3) no longer holds the file this table was built from. Every other row is one response per person-item. If your analysis needs unique (id, item) pairs, drop the repeated ones. | #1856 | 2026-09-30 |
PROMISPME_Forrest_2021_Physical_Proxy |
79 of the 121,418 rows repeat an (id, item) pair already in the table, and some of these repeats give a different resp. They cannot be resolved: the source deposit (OSF f7rp3) no longer holds the file this table was built from. Every other row is one response per person-item. If your analysis needs unique (id, item) pairs, drop the repeated ones. | #1856 | 2026-09-30 |
PROMISPME_Forrest_2021_Strength_Proxy |
25 of the 22,975 rows repeat an (id, item) pair already in the table, and some of these repeats give a different resp. They cannot be resolved: the source deposit (OSF f7rp3) no longer holds the file this table was built from. Every other row is one response per person-item. If your analysis needs unique (id, item) pairs, drop the repeated ones. | #1856 | 2026-09-30 |
przybylski_2016_igd |
The deposit’s US cohort in Study 2 contained 1,258 Canadian respondents duplicated from the Canadian cohort; each was kept once under Canada, so N and cohort sizes differ from the paper (which analysed 10,009 rows). Respondents who do not play games never saw the checklist and are excluded, and the distress item exists only in Study 2. | 2026-10-04 | |
purohit_2026_bureaucrat_help |
Outcome: which of two sarpanches the officer would help “get work done” for (exact wording in the article, not the deposit), forced choice, no opt-out. Attribute levels are the author’s short labels (e.g. UR/BC/SC reservation). 221 of the 253 surveyed officials appear in the conjoint file; only respondent gender is kept to limit re-identification. | #2887 | 2026-10-07 |
qi_2024_welfare_programs |
Conjoint table (irw_conjoint_2). choice = forced choice between two programs (‘Chose this program’ / ‘Chose other program’; wording not deposited, no opt-out); rating = program support 1-7 per program (wording and anchors not deposited). Attribute levels are the authors’ value labels (summaries for the description and problem texts). Restriction: problem statements match the program family of the description (EITC or TANF) or are ‘None’. Task/profile inferred from the profile identifier. 2,177 respondents (liberal subgroup 1,025 as in the authors’ log); IP addresses dropped. | 2026-10-09 | |
qi_2025_health_redistribution |
Conjoint table (irw_conjoint). choice: forced choice between two health-insurance programs (wording not deposited), no opt-out. Levels are the author’s short value labels; cost amounts are uncertain ($2/3/4/6 trillion in the value labels vs $2.4/3.2/4.3/6.2 trillion in the author’s figures). Task/profile inferred from the profile identifier. The author’s main analyses keep respondents passing 2+ of 4 screeners (cov_screener_count >= 2). IP addresses dropped. | 2026-10-09 | |
qi_2026_welfare_taxes |
Conjoint table (irw_conjoint). Forced choice between two welfare programs (wording not deposited; no opt-out). Randomization restricted: a property-tax source appears only with state and local government. The paper’s main estimates drop respondents passing fewer than 2 of 4 screeners (cov_attention_pass_1..4); all 1,996 respondents kept. IP addresses dropped. | 2026-10-09 | |
rains_2023_ward_candidate |
Conjoint table (irw_conjoint). choice = respondent prefers this hypothetical ward-leader candidate over the other (codebook paraphrase; original wording and language not deposited), no opt-out; 3 pairs per respondent, task inferred from row order and verified (one choice per pair), position within pair not recorded. Levels are the codebook’s English descriptions. 1,704 of 7,878 respondents whose candidate traits were not saved are dropped. | 2026-10-09 | |
ramamurthi_2022_india_coal_policy |
Conjoint table (irw_conjoint_2). Paired choice between two policy packages (6 attributes); forced choice, no opt-out. Question wording, codebook and the authors’ article are not in the deposit; task is the source question index 1-5 (order shown undocumented). Covariates kept as undocumented codes; survey items and timing dropped. 1,467 respondents; no paper count available. | 2026-10-12 | |
raman_2022_covid_booster |
Conjoint table (irw_conjoint). Single-profile tasks (5 per respondent). choice: “If you had to choose, would you choose to get this vaccine [booster]?” (1 = would choose to get it; opt-out = would not); rating: “How likely or unlikely would you be to get the vaccine [booster] …” 1 Extremely unlikely .. 7 Extremely likely. Randomized Omicron prime in trial_omicron_info. Includes all vaccination groups (article analyses fully vaccinated unboosted, n = 548, and unvaccinated, n = 619, separately); 3 respondents with missing attributes dropped. | 2026-10-08 | |
rapallini_2023_raven |
resp is 1 correct / 0 incorrect against a key derived from the deposit: the option per item that reproduces the deposited Raven total for every respondent (key D1 3, D2 4, D3 3, D4 7, D5 8, D6 6, D7 5, D8 4, D9 1, D10 2, D11 5, D12 6). The option chosen is not shipped. The deposit’s person code is reused across sessions, so id is the row index. | 2026-10-11 | |
rasmussen_2023_interest_groups |
Conjoint table (irw_conjoint). Factorial vignette (up to two per respondent: hybrid-car tax reduction and sugar restrictions, order randomized; 3 factors). rating_fair/right/actors/citizens/democratic/affected: agreement with six legitimacy statements, 0 = strongly disagree .. 4 = strongly agree as deposited. No choice. Levels are codebook labels, not vignette text; UK/US/DE pooled. Failed-check respondents kept (cov_attention_pass, cov_failed_manipulation); rows with no outcome omitted; authors’ derived scales dropped. | 2026-10-09 | |
rathbun_2024_nuclear_strike_prolific |
Conjoint table (irw_conjoint_2). One Tomz-Weeks style vignette per respondent (2 x 2: regime (democracy / not a democracy) x country race bullet (predominantly non-white / no race bullet); the white arm of the first survey was dropped; Fixed bullets: developing nuclear weapons (first bomb within six months), not a military ally, low trade, military half as strong as US forces in the region, predominantly Christian.) rating = support for using the US military to attack the country’s nuclear sites, stored 1-7 (higher = favor) although the article describes a five-point scale; rating_threat and rating_immoral are 1-7 agreement items (7 = strongly agree). No opt-out; no respondent id (id = row number). Authors’ factor/additive scales dropped; N = 2,659 as in the paper. | 2026-10-09 | |
rathbun_2024_nuclear_strike_qualtrics |
Conjoint table (irw_conjoint_2). One Tomz-Weeks style vignette per respondent (2 x 3: regime (democracy / not a democracy) x country race bullet (predominantly white / predominantly non-white / no race bullet); Fixed bullets: developing nuclear weapons (first bomb within six months), not a military ally, low trade, military half as strong as US forces in the region, predominantly Christian.) rating = support for using the US military to attack the country’s nuclear sites, stored 1-7 (higher = favor) although the article describes a five-point scale. No opt-out; no respondent id (id = row number). Authors’ factor/additive scales dropped; N = 1,626 as in the paper. | 2026-10-09 | |
reeves_2026_councillor_candidates |
Conjoint table (irw_conjoint). choice: “Which of the following two persons do you think is the most desirable as a proportional representation list member of the House of Councillors?” (Japanese instrument), forced choice, no opt-out. 10 tasks per respondent, 7 attributes, Japanese level text with the respondent’s prefecture substituted into hometown and assembly levels; attribute order randomized per respondent (attrpos_*). trial_treatment_group = main-experiment arm (1-4). 2,018 respondents; Qualtrics IDs re-keyed; free-text answers not kept. | 2026-10-08 | |
regalado_2023_tourism_value |
cov_age is an age-band code, not age in years: 2 = “18 a 30 años”; 3 = “31 a 40 años”; 4 = “41 a 50 años”; 5 = “51 a 60 años”; 6 = “Más de 60 años”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
reher_2025_disabled_candidates |
Conjoint table (irw_conjoint). choice: “Which candidate would you be more likely to vote for?” (A or B, no opt-out; 2,632 tasks unanswered, left blank); rating_ideology: candidate placement 0 = Left .. 10 = Right; rating_represent: “Candidate A/B represents people who are under-represented in politics” 0 = strongly disagree .. 4 = strongly agree. Each respondent saw a no-party and a party pair in random order (trial_party_shown; attr_party (not shown) when not shown). Gender and minority status were signalled by names not in the deposit (codebook codings kept); disability and minority names had non-uniform weights. 6,011 ids in the deposit vs N=3,000 per country reported; 42 with no answers dropped. | 2026-10-07 | |
rehmert_2022_party_delegates |
Conjoint table (irw_conjoint). Outcome: forced choice between two candidates (wording not in the deposit; paraphrase), no opt-out. Level shares are unequal and dependent: legal and farming occupations appear almost only with Abitur, and a 6-year membership dominates for candidates with legislative experience; no pair is fully absent, the scheme is undocumented, and AMCEs should condition on these pairs as the authors do. Attribute text is German (the .dta value labels); 50 respondents with no conjoint answers dropped. | 2026-10-08 | |
rehmert_2026_geographic_responsiveness |
Conjoint table (irw_conjoint_2). Paired conjoint of German party delegates choosing which of two hypothetical candidates they prefer in a nomination (exact question wording not in the deposit); forced choice, no opt-out. Eight attributes; some levels are relative to the respondent (e.g. ‘in Respondent’s PR District’). 349 tasks with no recorded choice were dropped (648 delegates in the deposit, 591 kept). Respondent IDs are re-keyed. | 2026-10-14 | |
reidinger_2025_gender_representation |
Conjoint table (irw_conjoint_2). Single-profile attribute-table descriptions of a decision-making body, rated 0-10: rating_cond and rating_recog (fairness items) on tasks 1-7, rating_rep and rating_stat (output expectations) on tasks 8-14, the other columns NA (trial_block). Attribute level text for gender and ideology is the authors’ recode; age, class and ethnicity are the deposit’s short tags because the displayed text is not in the deposit. Outcome wording and anchors are paraphrased from the README; the order of the two blocks is unknown. | 2026-10-14 | |
reidinger_2026_committee_decisions |
Conjoint table (irw_conjoint_2). One respondent evaluates three committee decisions (policy order randomized, approval randomized per policy) made by one committee whose composition (women, age, parents, degree) was randomized; ratings 1-10: fairness of the process (1 very unfair to 10 very fair), agreement with the decision (1-10) and support for the policy statement after seeing the decision (1-10). No opt-out; text of the composition levels is the per-respondent reminder string; 10,066 of 10,101 respondents have at least one rating. | 2026-10-11 | |
reijnders_2025_cocreation |
Conjoint table (irw_conjoint). Two paired tasks, 4 attributes (Dutch text as displayed); choice = ‘Aan welk participatie initiatief zou je het liefst willen deelnemen?’ (forced, Initiatief 1/2); choice_results, choice_process, choice_trust = which initiative would give the best results / the best process / most trust in the government, each with a ‘Geen verschil’ (no difference) opt-out. French-language respondents (457) dropped because their displayed text is not saved; N = 683. | 2026-10-09 | |
reinsberg_2026_imf_messages |
Conjoint table (irw_conjoint). Factorial vignette (one per respondent): six statements each shown with probability one-half after a fixed base text. rating = approval of the IMF program (1 disapprove a lot - 4 approve a lot); rating_gov_responsible and rating_imf_responsible = responsibility of the government / the IMF for the program (1 not - 3 much responsible); rating_imf_confidence = confidence in the IMF (0-10). Statement order in the vignette undocumented. cov_survey_weight = poststratification weight. N 1,691 (articles: 1,600; analysis sample 1,514). | 2026-10-09 | |
ren_2019_scpv |
SCPV1 does not behave like the other eleven SCPV items. Its responses stay within 1-4 where the rest use 1-5, its median correlation with the other SCPV items is 0.15 (theirs are 0.32-0.54), and it tracks CPTI2 (r 0.77) and SDQ1 (r 0.70) closely. It may not be the item the deposit labels it as; consider dropping it from SCPV scores. | #2425 | 2026-09-30 |
retrosheet_mlb_1871_2025 |
Forfeited games: winner follows Retrosheet’s forfeit code, not the score (forfeit V = agent_b, H = agent_a; T, ruled a no-decision, = draw). homefield is always agent_a, the team Retrosheet lists as home; neutral-site games (Tokyo, London, Mexico City series, etc.) are not marked, and park identifies them. league NAssn is Retrosheet’s NA (National Association, 1871-75), renamed so it is not read as missing. | #2702 | 2026-10-01 |
ridge_2024_regime_egypt |
Conjoint table (irw_conjoint). choice: “Which of these two political systems of government would you prefer the most?” (forced choice, OPTION 1 / OPTION 2). Unconstrained randomization except that the two profiles were never identical. Expats kept with cov_nationality (the author drops them; Egyptians = 1,000 matches the article); 3 tasks with no choice and 3 with two chosen profiles dropped. Level text English; most respondents saw Arabic. | 2026-10-07 | |
ridge_2024_regime_morocco |
Conjoint table (irw_conjoint). choice: “Which of these two political systems of government would you prefer the most?” (forced choice, OPTION 1 / OPTION 2). Unconstrained randomization except that the two profiles were never identical. Expats kept with cov_nationality (the author drops them; Moroccans = 991 matches the article); 1 task with no choice and 1 with two chosen profiles dropped. Level text English; most respondents saw Arabic. | 2026-10-07 | |
rinscheid_2019_coal_phaseout_germany |
Conjoint table (irw_conjoint). Forced choice between two coal phase-out scenarios (no opt-out) and a 1-7 rating of each (1 = very poor, 7 = very good); exact wording not deposited. Fixed blocked design (500 versions); attribute text is the article’s English Table 1. The paper excludes attention-check failures (N = 1984); all 2161 are kept with cov_attention_pass. Cost framing (household only vs household plus economy-wide) is in trial_conjoint_type with undocumented codes; postcode and place of residence dropped. | 2026-10-08 | |
rinscheid_2019_coal_phaseout_lusatia |
Conjoint table (irw_conjoint). Forced choice between two coal phase-out scenarios (no opt-out) and a 1-7 rating of each (1 = very poor, 7 = very good); exact wording not deposited. Fixed blocked design (500 versions); attribute text is the article’s English Table 1. The paper excludes attention-check failures (N = 473); all 501 are kept with cov_attention_pass. Cost framing (household only vs household plus economy-wide) is in trial_conjoint_type with undocumented codes; postcode and place of residence dropped. | 2026-10-08 | |
rinscheid_2019_coal_phaseout_rhineland |
Conjoint table (irw_conjoint). Forced choice between two coal phase-out scenarios (no opt-out) and a 1-7 rating of each (1 = very poor, 7 = very good); exact wording not deposited. Fixed blocked design (500 versions); attribute text is the article’s English Table 1. The paper excludes attention-check failures (N = 491); all 533 are kept with cov_attention_pass. Cost framing (household only vs household plus economy-wide) is in trial_conjoint_type with undocumented codes; postcode and place of residence dropped. | 2026-10-08 | |
ripamonti_2026_transparency_trust |
Conjoint table (irw_conjoint_2). Single-profile text vignettes of a public organization (five binary attributes), eight per respondent, each rated on three 0-100 sliders (what percent of the time citizens can trust it to act in a principled way, in the city’s interest, efficiently); no opt-out. Attribute text for the disclosure motive taken from the appendix (the deposit stores short tags). Only 16 of the 32 combinations occur (half-fraction); 20 blocks of 8 (trial_deck). Respondent demographics are restricted and not included. | 2026-10-14 | |
robinson_2022_cf_candidates |
Conjoint table (irw_conjoint). choice: ‘If you had to choose, which candidate would you vote for?’ (Candidate A or B, no opt-out). Subjects were randomly assigned to see party, ideology and previous office in addition to the 5 finance attributes (trial_arm; ‘(not shown)’ otherwise). 437 rounds missing one profile in the deposit (implausible donation combinations apparently removed after fielding) are dropped; the candidate approval rating is not in the deposit. | 2026-10-09 | |
robinson_2022_cf_initiatives |
Conjoint table (irw_conjoint). choice: ‘If you had to choose, would you vote for or against this initiative?’ (For = Support campaign profile chosen, Against = Opposition; no opt-out); rating: approval of each side of the campaign, 1 strongly disapprove to 7 strongly approve. Issue display order not saved (task follows a fixed issue order). 103 rounds missing one campaign in the deposit are dropped. | 2026-10-09 | |
rodon_2020_fair_redistribution |
Conjoint table (irw_conjoint). rating = “On a scale from 0 to 10, how FAIR do you think the impact of these policies would be in:” each of two countries, stored as the export’s codes 1-11 (1 = 0 very unfair, 11 = 10 very fair); 5 tables per respondent, attribute order randomized per table. Table 4 used alternative row labels and mobility wording (trial_label_set). Duplicate Prolific submissions (2) dropped; 1,577 respondents vs 1,567 analysed in the article; speeders kept (cov_duration_sec). | 2026-10-09 | |
rogowski_2021_court_nominees |
Conjoint table (irw_conjoint). rating = ‘On a scale from strongly oppose to strongly support, where would you place your level of support for this potential nominee?’ 1-5 (strongly oppose to strongly support); rating_trust = trust in the nominee’s impartiality, 1-5 (wording not deposited). Single-profile ratings, task order inferred from row order; trial_rhetoric is a respondent-level arm (elite rhetoric shown or not). Attribute levels come from the authors’ code; the ‘Think-tank’ level is labelled ‘Chief counsel’ in their figures. The authors’ additive legitimacy index is not included; 7 rows with no outcome dropped. | 2026-10-08 | |
romero_2026_coyote_choice |
Conjoint table (irw_conjoint). choice: respondent’s pick between two hypothetical coyotes (question wording not deposited; codebook: 1 = this profile was chosen), forced choice, up to 3 tasks. Levels are the authors’ English short labels; the displayed text is not deposited. 549 unanswered tasks dropped; 1,415 of 1,534 respondent ids answered at least one task. | 2026-10-09 | |
rosendorf_2026_autonomous_weapons |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent, 2 x 2: target (soldiers / terrorists) x misidentification risk (equal for both drones / higher for the remote-controlled drone); four countries in one table (cov_country). rating is the forced-choice drone preference (1 = autonomous, 0 = remote-controlled); ethicality of each drone is rated 1-6 (HIGHER = more ethical). The attribute text is the authors’ description, not the verbatim vignette, and the exact higher risk figure shown cannot be recovered. | 2026-10-11 | |
rosyid_2025_academic_citizenship |
cov_age is an age-band code, not age in years: 1 = “17 - 20 Years”; 2 = “20 - 24 Years”; 3 = “> 24 Years”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
rosyid_2025_ethical_leadership |
cov_age is an age-band code, not age in years: 1 = “17 - 20 Years”; 2 = “20 - 24 Years”; 3 = “> 24 Years”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
rosyid_2025_prosocial_motivation |
cov_age is an age-band code, not age in years: 1 = “17 - 20 Years”; 2 = “20 - 24 Years”; 3 = “> 24 Years”. The labels are the source file’s own (see the covariate_labels table). The column will be renamed cov_age_band when the table is next rebuilt. | #1775 | 2026-10-01 |
rudolph_2022_supply_law_conjoint |
Conjoint table (irw_conjoint). Each pair of law proposals was rated 1 (total dagegen) to 7 (total dafür) and then chosen between for a hypothetical referendum (‘welchen Vorschlag würden Sie eher annehmen?’), no opt-out; unanswered choices are NA. Level text is the authors’ short English labels. trial_norms marks the random half shown a UN-norms text first (the article’s main conjoint estimates use trial_norms = 0). 38 tasks whose levels were not saved are dropped. | 2026-10-08 | |
rudolph_2022_supply_law_vignette |
Conjoint table (irw_conjoint). One 3 x 3 factorial text vignette per respondent (a law adopted by referendum; reciprocity and stringency sentences in German, verbatim), followed by 7 agreement ratings 1 (stimme überhaupt nicht zu) to 5 (stimme voll und ganz zu), raw; the instrument’s ‘Weiss nicht’ option does not appear in the data. trial_norms marks the random half shown a UN-norms text first. 47 respondents without any rating dropped. | 2026-10-08 | |
rudolph_2024_arms_exports |
Each of 6 pairs had a forced choice (“If you had to decide between the two arms deliveries, which one would you personally prefer and approve?”, no opt-out) and a 1-7 rating of each profile (1 = should definitely be rejected, 7 = should definitely be approved). Levels drawn uniformly with no restrictions; attribute order randomized per respondent but not in the data; English master labels, other-country level = France for German and Germany for French respondents. Vignette and party-position experiments, free text and timers dropped; N = 6,617 matches the paper, mean rating 3.20 vs reported 3.23. | #2887 | 2026-10-07 |
rudolph_2026_migration_vignette |
Conjoint table (irw_conjoint). One text vignette per respondent (6 fixed scenarios, 3 damage x 2 risk). choice: “If you were in this situation, would you stay here in your present village or would you prefer to move to another place?” 1 = move, 0 = stay; an unprompted ‘depends’ was followed by a forced choice (trial_depends = 1). Staying is the outside option. 1,561 heard a vignette; 27 refusals omitted (1,534). Text is the English master questionnaire; interviews were in Bengali. | 2026-10-08 | |
rudolph_2026_ukraine_support |
Conjoint table (irw_conjoint). choice: “If you had to choose between one of the two strategies, which would you personally prefer?” (translated from the German master questionnaire; forced, no opt-out). rating: each strategy on 1-7, 1 = definitely reject, 7 = definitely support (endpoint order randomized; reversed half recoded as in the author’s code), asked in 2 of 4 tasks only. Attributes uniformly randomized without restrictions per the article; ties kept. All 2,334 respondents kept, matching the article. | 2026-10-07 | |
rudolph_2026_ukraine_western |
Conjoint table (irw_conjoint). choice: ‘If you had to choose between one of the two strategies, which would you personally prefer?’ (German master wording; forced, no opt-out). Four of the article’s five countries (US omitted: already in IRW as rudolph_2026_ukraine_support from the same survey); the 1-7 rating asked in 2 of 4 tasks is not included because its endpoint alignment is undocumented here; survey weight kept; postcode dropped. | 2026-10-09 | |
ruedin_2020_hiring_population |
Conjoint table (irw_conjoint). One CV vignette per respondent (job x Swiss/migrant name, 2 x 2); nine hiring judgements (continuous slider stored 0-7 though documented as 1:7, anchors undocumented, rating_risk and rating_often_ill negative) plus three manipulation checks (1 = yes, 2 = no), all raw. Level text is the French version; German-language respondents saw German job titles and the Swiss name Martin Baumgartner, and the language used is not recorded. | 2026-10-08 | |
ruedin_2020_hiring_students |
Conjoint table (irw_conjoint). One CV vignette per respondent (job x Swiss/migrant name, 2 x 2); nine hiring judgements (categorical 1-7, anchors undocumented, rating_risk and rating_often_ill negative) plus three manipulation checks (1 = yes, 2 = no), all raw. Level text is the French version; German-language respondents saw German job titles and the Swiss name Martin Baumgartner, and the language used is not recorded. Two respondents without any answer dropped. | 2026-10-08 | |
rugby_union_tier1_1871_2026 |
Wikipedia prints each score without saying whose comes first, so score_a/score_b give the higher score to the listed winner. Matches before about 1890 were decided on goals: 58 have no points score (score_a/score_b blank, printed score in score_raw), and the early points values are as the pages give them. Home side is the venue’s country (Belfast counts as Ireland); World Cup and other third-country matches have homefield blank. Only matches both sides capped are included: the 1920-28 Australia (New South Wales) v New Zealand and South Africa matches and uncapped touring XVs are excluded. | #2708 | 2026-10-01 |
rzeszutek_2023_danieli |
Each respondent rated one ancestor generation, apparently assigned by age band (parents n=658, grandparents n=736, great-grandparents n=204); 114 respondents gave one constant answer to all 60 items (100 of them all 3s; straight-lining), kept as observed. | 2026-10-04 | |
rzeszutek_2023_ees |
96 of 1,598 respondents (6%) answered 3 to all 34 items, and a few more gave key-consistent alternating patterns (e.g. 4/2 by item keying); online-panel straight-lining, kept as observed. | 2026-10-04 | |
rzeszutek_2023_pds5 |
Administered only to the 964 respondents who endorsed at least one traumatic event on the checklist; the 634 with no event have no PDS-5 symptom rows. | 2026-10-04 | |
rzeszutek_2023_swls |
Present in the data file but not mentioned in the paper; identified by a 5-item 1-7 block whose sum equals the file’s SWLS column on every row. | 2026-10-04 | |
sadish_2021_covid_beliefs |
resp is endorsement (1 = Yes), not correctness: only the asymptomatic-spread statement is true. ‘Don’t know’ (1,872 answers over two waves) is a non-response and is dropped. | 2026-10-09 | |
sagebiel_2026_tofu |
Conjoint table (irw_conjoint). DCE: choice = which of two tofu variants the respondent would buy (‘Vilken produkt skulle du föredra?’), with opt-out ‘I would not buy any of the variants’ (both 0). Fixed 27-row design drawn at random per choice set; the two variants never share a level. trial_info_treatment = randomized information page (Environment/Health/Recipe/Control). Includes 82 timed-out respondents (cov_status) and an attention item (cov_attention_pass, ~50% pass); the article’s ~1,400 sample is not reproduced. | 2026-10-09 | |
saha_2023_meat_candidates |
Conjoint table (irw_conjoint). Five pairs of hypothetical same-party presidential primary candidates; choice = which candidate the respondent would be most likely to vote for in the primary (forced choice, no opt-out). Attributes drawn independently at random (article). N = 857 as in the article; the paper’s AMCEs (strong animal-rights supporter +17.3, vegan -6, rescued farm animals ~+9 points) are approximately reproduced (+18.7, -5.2, +7.9 by OLS; the authors ran cregg on na.omit of the full file). | 2026-10-09 | |
saikkonen_2023_elite_transgressions |
Conjoint table (irw_conjoint). choice = the profile respondents would prefer as prime minister of Finland (paraphrase; forced choice, no opt-out); choice_vote = follow-up whether they would also vote for the selected leader (1 on the chosen profile when yes; opt-out). Randomization without restrictions (article); level text is the article’s English translation of the Finnish profiles. N = 1,030 as in the article; authors’ recodes and municipality dropped. | 2026-10-09 | |
sangwin_2021_proof_rigour_insight |
excluded = 1 marks the decisions of judges whose median time per judgement was under 5 seconds (484 of 3,969). Sangwin & Kinnear dropped these as nonserious judges, and their analyses and the Kinnear et al. CJ meta-analysis use the remaining 3,485. | #2692 | 2026-10-01 |
sanz_martos_2024_stai_state |
Baseline (pre-massage) responses only. The deposit’s post-massage item columns are near-copies of the baseline columns (96.5% of cells identical) while its post totals disagree with them, so they were not shipped. | 2026-10-09 | |
sanz_martos_2024_stai_trait |
Baseline (pre-massage) responses only. The deposit’s post-massage item columns are copies of the baseline columns (99.5% of cells identical) while its post totals disagree with them, so they were not shipped. | 2026-10-09 | |
sasmaz_2026_candidate_selection |
Conjoint table (irw_conjoint). choice = ‘Which person would you contact first to offer a place in your list?’ (English questionnaire; asked in Arabic), forced choice between two aspirants, 3 rounds; 103 unanswered rounds omitted and 11 candidates with no answered round dropped (1,896 of 1,907). Levels stored in Arabic as shown; the profession level ‘وظيفة عمومية’ (public-sector job) is labelled ‘lawyer’ in the author’s code and questionnaire. Attribute order randomized per respondent (attrpos_). Restrictions undocumented. cov_ = candidate’s list type and party membership. | 2026-10-07 | |
schaap_2022_ai_agency_dating |
Conjoint table (irw_conjoint). Factorial vignette, one scenario per respondent (Study 3, 16 cells). rating: ‘Reviewing the options above, which would you choose to assist you?’ 1 = Definitely the expert to 7 = Definitely the algorithm; rating_agent: ‘If you merely had to choose between either the expert or the algorithm, which would you choose?’ 1 = The expert, 2 = The algorithm. No opt-out; full factorial, no restrictions. Free-text reasons, nationality and the authors’ derived variables dropped; gender ‘Other/don’t want to tell’ set to NA. N matches the paper. | 2026-10-09 | |
schaap_2022_ai_agency_stocks |
Conjoint table (irw_conjoint). Factorial vignette, one scenario per respondent (Study 1, 16 cells). rating: ‘Reviewing the options above, which would you choose to assist you?’ 1 = Definitely the expert to 7 = Definitely the algorithm; rating_agent: ‘If you merely had to choose between either the expert or the algorithm, which would you choose?’ 1 = The expert, 2 = The algorithm. No opt-out; full factorial, no restrictions. Free-text reasons, nationality and the authors’ derived variables dropped; gender ‘Other/don’t want to tell’ set to NA. N matches the paper. | 2026-10-09 | |
schaap_2023_ai_agency_incentivized |
Conjoint table (irw_conjoint). Factorial vignette, one incentivized scenario per respondent (16 cells, real GBP 1 investment). rating_agent: ‘Please choose your option now’ 1 = Option A: Expert, 2 = Option B: Algorithm; rating: ‘Reviewing the options above, please tell us how strongly you prefer one option over the other for this investment.’ 1 = I definitely prefer the expert to 7 = I definitely prefer the algorithm; rating_invest (only if the picked agent offered control): 1 = BUY!, 8 = ABORT!. No opt-out; full factorial. Prolific IDs, free text and derived variables dropped; gender ‘Other/don’t want to tell’ set to NA. N = 210, no paper. | 2026-10-09 | |
scheerhagen_2019_reproq |
Two questionnaires linked by survey token: wave 1 = antenatal (pregnancy phase, items *_P), wave 2 = postnatal (birth *_A and postnatal *_B). 94 exact duplicate submissions (same token, every column identical) dropped; 92 tokens that recur with different answers are treated as different women and not linked; 1,590 postnatal rows with a blank token are not linked. Not-applicable answers (no emergency, do not remember, no referral, etc.) are dropped, as are Pnb code 4 ‘did not want to be involved’ and Svd ‘don’t recall’. Covariates are taken from the antenatal record when present. | 2026-10-05 | |
schick_2026_hazard_adaptation |
Conjoint table (irw_conjoint). choice: ‘Which option do you prefer?’ Option 1/Option 2, forced (no opt-out); rating_acceptance: ‘Irrespective of your choice, how acceptable do you find the two options?’ 1 (totally unacceptable) to 6 (totally acceptable). Attribute text is as displayed in four survey languages; task 6 repeats task 1 with options swapped; observed restriction: company CO2-based costs never shown with an exemption; free-text fields dropped; 923 respondents in the file vs n = 891 in the README. | 2026-10-09 | |
schillemans_2025_accountability |
Conjoint table (irw_conjoint). Up to two paired tasks (trial_forum principal, then stakeholder) with five binary attributes described by the codebook’s English level text; choice = which demand to prioritize (paraphrase; wording not deposited), forced, no opt-out. Task inferred from row order; 9 respondents have one pair only. N = 385 respondents, 761 decisions as in the article; no rows dropped. | 2026-10-09 | |
scholten_2025_military_action |
Conjoint table (irw_conjoint). choice = against which of the two countries military action is more justified (paraphrase; forced choice); rating = justification of military action against each country, 0 = Completely unjustified to 10 = Completely justified. Mass (Lucid) and State Department expert samples pooled with cov_sample. Task is inferred from row order; 132 tasks with no recorded choice are blank (the authors count them as not chosen); 57 rows with no outcome dropped, leaving 1,131 respondents vs 1,132 reported. | 2026-10-07 | |
scholten_2025_return_migrants |
choice = “¿Cuál de estas opciones prefieres?” (forced, no opt-out); rating = “Indique qué tan favorables considera ambas opciones” 0-10 (higher = more favourable; anchors not in deposit). All respondents with any outcome kept (1,642, incl. partial completes); the authors analyse the 1,572 answering all 6 choices. IP, location, Qualtrics ID, Netquest ticket and consent item dropped; article N not checked. | #2887 | 2026-10-07 |
schroeder_2025_vp_nominee |
Conjoint table (irw_conjoint). choice: which of two hypothetical candidates the respondent prefers as vice-presidential nominee (paraphrase; no questionnaire deposited), forced, no opt-out. Task 2 is the same fixed pair for every respondent (trial_fixed_pair = 1), which makes pooled level shares unequal; tasks 1, 3-5 are fully randomized. 16 of 599 respondents lack 1-3 tasks. Covariates keep the source codes. Prolific IDs and per-task follow-up items with no recorded wording dropped. | 2026-10-08 | |
schuler_2024_gender_clientelism |
Conjoint table (irw_conjoint). One task: which of two National Assembly candidates the respondent would vote for (Candidate 1/2; DK/refusal respondents omitted, 540), no opt-out; prime arm (control / honesty / clientelism) as trial_prime. Levels are the survey’s Vietnamese level names; kinship coding in the authors’ do-file conflicts with these names. 2 respondents with a missing level dropped. | 2026-10-09 | |
schulz_2020_housework_sharing |
Conjoint table (irw_conjoint_2). Single-profile text vignettes of a couple (six randomized factors), four per respondent, each answered three times: who should do the housework in total, routine and non-routine housework, on a 7-point scale stored 1-7 (1 = woman alone, 4 = both equally, 7 = man alone; higher = more to the man). No opt-out. Restricted design: implausible employment-income combinations removed from the full factorial and 120 vignettes drawn d-efficiently in 30 decks (trial_deck). | 2026-10-14 | |
schwartz_2020_audience_costs_mturk |
Conjoint table (irw_conjoint_2). rating = ‘What are your views about how the US President handled the situation?’ (1 = strongly approve … 7 = strongly disapprove, as coded; higher = more disapproval); rating_regime = how democratic the attacking country is (-10 to 10); rating_nonwhite = guessed president race non-White (authors’ 0/1). One of 24 equally likely vignettes per respondent (2 president genders via names x 2 opponent genders x 2 parties x 3 actions). Open-ended answers, derived dummies and IDs dropped. Exploratory MTurk pilot (party quota); no respondent id in the deposit, id = row number. | 2026-10-09 | |
schwartz_2020_audience_costs_tess |
Conjoint table (irw_conjoint_2). rating = ‘To what extent do you disapprove or approve of how the US President handled the situation?’ (1 = strongly approve … 7 = strongly disapprove, as coded; higher = more disapproval); rating_regime = how democratic the attacking country is (-10 to 10); rating_nonwhite = guessed president race non-White (authors’ 0/1). One of 24 equally likely vignettes per respondent (2 president genders via names x 2 opponent genders x 2 parties x 3 actions). Open-ended answers, derived dummies and IDs dropped. TESS randomization blocked on respondent party ID; 1 respondent with no outcome omitted; N 2,342 in the deposit. | 2026-10-09 | |
scottbuechler_2024_dac_siting |
Conjoint table (irw_conjoint). choice: ‘Please choose which plan you prefer from each pair’ (plan 1/2), forced, no opt-out, 13 tasks; taken from the recorded answers because the deposit’s long Choice column only marks the first plan. Fixed CBC design, 210 versions (trial_version). Authors’ cleaned sample only (1,195 of 2,198). cov_survey_weight = authors’ raking weight. Date of birth, ZIP codes and panel ids dropped. | 2026-10-09 | |
secen_2026_refugee_cities_de |
Conjoint table (irw_conjoint). choice = which of two cities the respondent would prefer to move to (Arabic question text from the export), forced, one per task; no opt-out. Six binary attributes, Arabic level text; attribute row order randomized once per respondent (attrpos_). Respondents who stopped part-way keep their answered tasks. N 735; article count not checked (paywalled). | 2026-10-09 | |
secen_2026_refugee_cities_tr |
Conjoint table (irw_conjoint). choice = which of two cities the respondent would prefer to move to (Arabic question text from the export), forced, one per task; no opt-out. Six binary attributes, Arabic level text; attribute row order randomized once per respondent (attrpos_). Respondents who stopped part-way keep their answered tasks. N 4,322; article count not checked (paywalled). | 2026-10-09 | |
segovia_2020_candidate_emotions |
Conjoint table (irw_conjoint). Outcome: forced choice of which of two candidates the respondent would prefer as president of Chile; 2,245 of 9,000 tasks have no recorded choice and are omitted, so 1,542 of 1,800 respondents remain. Level text is the deposit’s Spanish value labels, four of which Stata had truncated at 60 characters and are completed here; the article’s attribute table words several levels differently, so the exact displayed wording is uncertain. cov_survey_weight is the survey weight the author uses in every model. | 2026-10-08 | |
seitz_2024_artist_vaccine_job |
Conjoint table (irw_conjoint_2). One hypothetical-choice vignette per respondent (Germany, arts and culture, 2021; German text): vaccine immunity (50/70/90%) x pay of an alternative outside job (same / 20% lower); rating_decision = raw 1 (resume artistic work with vaccination) or 2 (take the outside job), a decision about the vignette rather than a pick among displayed profiles. 654 of 788 respondents kept (100 not assigned, 34 unanswered). | 2026-10-14 | |
sejna_2025_academic_self_concept |
73 item cells (and 2 sex cells) were mean-imputed in the deposit (each fractional value equals its column mean) and are dropped. The deposit’s ID is not unique (534 twice), so id is the row index. | 2026-10-09 | |
sejna_2025_saqc |
73 item cells across the file were mean-imputed in the deposit and are dropped. The record describes 5-point items, but no SAQ-C item uses 5 in 740 pupils except SA02 twice; those two cells are dropped as entry errors, so the table is on 1-4. id is the row index (the deposit’s ID repeats 534). | 2026-10-09 | |
sejna_2025_social_integration |
73 item cells (and 2 sex cells) were mean-imputed in the deposit (each fractional value equals its column mean) and are dropped. The deposit’s ID is not unique (534 twice), so id is the row index. | 2026-10-09 | |
selfcompassionscale_shortform_fuochi_2025 |
3,024 of the 24,816 rows repeat an (id, item) pair. The repeats sit inside a single sample (cov_sample) and disagree on resp, and the source file cannot say which copy is right. Every other row is one response per person-item. If your analysis needs unique (id, item) pairs, drop the repeated ones. | #1856 | 2026-09-30 |
semeval2013_scientsbank_nom |
text is not an option the student chose: it is the graders’ classification of the free-text answer (in answer_text) into one of five accuracy categories, converted from finer facet-level annotations and hand-revised by the task organisers. resp collapses it to SemEval’s 2-way scheme (1 = correct, 0 = any other label). id is module_student: student numbers restart in each module, and each module was given to a different group of children. | #2651 | 2026-10-01 |
senninger_2023_mp_eu_oversight |
Conjoint table (irw_conjoint). choice = which of the two MPs the respondent would vote for (forced choice, no opt-out; wording paraphrased from the preprint). Restriction: AfD candidates always have two years in parliament, so party and experience effects must be estimated within allowed combinations. Attribute order randomized (attrpos_*). Half the respondents answered the survey questions before the conjoint (trial_block_order). Level and covariate text is German as displayed. | 2026-10-07 | |
sergeeva_2026_xr_operators |
Conjoint table (irw_conjoint_2). Factorial vignette: machinery operators (Prolific) rate up to 3 XR-system vignettes varying data storage, switch-off option, operator reminders and effect on communication (binary conditions, a half-fraction design) on how acceptable the technology is (rating, 1 = not at all acceptable to 20 = totally acceptable). No opt-out. Attribute text is paraphrased from the codebook (the deposit stores 1/2 codes); derived scale sums are dropped; 53 vignette rows with failed attention checks and 2 consent-revoked respondents are removed. | 2026-10-14 | |
sevi_2026_election_outcomes |
Conjoint table (irw_conjoint). Six paired tasks of two hypothetical election outcomes. choice: “If you had to choose one of these two scenarios, which would you pick as better for Canada?” (forced, no opt-out); rating: each outcome 0 very bad .. 10 very good. Level text as displayed in English or French (cov_language), attribute order randomized per respondent (attrpos_). All 2,487 who completed the conjoint are kept; the authors’ analysis keeps 2,015 (drops non-citizens, under 240 s, over 1,200 s). | 2026-10-08 | |
shaffer_2020_local_refugees |
Conjoint table (irw_conjoint). rating: receptive to this refugee group settling in the community (1 = answered ‘Either group’ or ‘Group |
2026-10-08 | |
shafranek_2021_roommate_choice |
Conjoint table (irw_conjoint). choice: which of two roommates was chosen (wording unknown), no opt-out; rating: each roommate 1-7, higher = more favourable (chosen profiles rate higher), anchors unknown. Task/profile inferred from row order and verified (one choice per pair). Sexual orientation LGBT drawn with probability 1/5 (authors’ design); race shares also unequal. 203 of 205 respondents kept (2 with no saved attributes); article N not checked. | 2026-10-07 | |
shandler_2023_cyberterror |
Conjoint table (irw_conjoint). Single-profile design: rating = 1 if the respondent classified the incident as cyberterrorism, 0 if not; exact wording not deposited. US, UK and Israeli samples in one table (cov_country), as the authors analyse them; attribute text is the English version, mapped through the level codes for Israeli respondents, who saw Hebrew. Online databases never paired with an explosion outcome. The deposit has 21,252 classifications vs 21,238 in the paper. Replaces the three country tables of irw_conjoint v1.1. | #2887 | 2026-10-07 |
shi_2025_restaurant_hiring |
Conjoint table (irw_conjoint). choice is the preferred restaurant in each of 5 pairs (exact wording not deposited); forced choice, no opt-out. Task and profile are inferred from row order. Level text is the .dta value labels; the deposit’s Data Dictionary gives the staff-composition codes the other way round and labels the worst review ‘Terrible’, so those two attributes carry some label uncertainty. cov_survey_weight is the YouGov weight. | 2026-10-07 | |
shockley_2024_qatar_groups |
Forced choice between two baskets of non-citizen groups to prioritise for citizenship, up to 3 contests; each group coded Mentioned / Not mentioned, with basket position in attrpos_. Labels are the authors’ English labels (Arabic shown). N = 488 matches the authors’ log. | #2887 | 2026-10-07 |
shockley_2024_qatar_residency |
Choice between two permanent-residency candidates with the option to reject both (35 tasks have no chosen profile). Religion was cued by a first name that is not deposited, so the level is the cued identity. 657 of 733 completes deposited (the paper’s model uses 554); the profile ratings the article mentions are not deposited. | #2887 | 2026-10-07 |
short_2025_candidate_pilot |
Conjoint table (irw_conjoint_2). Paired candidate choice (7 attributes: party, five issue positions, experience) from pilot 3 of the study, a June 2023 convenience sample: which candidate would you vote for (paraphrase), forced choice. 20 tasks with no recorded choice dropped; weights kept as cov_survey_weight. | 2026-10-11 | |
short_2025_merger_challenge |
Conjoint table (irw_conjoint_2). Single-profile ratings of five hypothetical mergers (7 attributes) by US adults and antitrust lawyers: should the government allow or challenge the merger, four points (1 = definitely allow, 4 = definitely challenge); no opt-out. Survey weight kept as cov_survey_weight; attribute order randomized per the article but not recorded; the article says four mergers per respondent, the data hold five. | 2026-10-11 | |
shum_2026_hk_diaspora_policy_au |
Conjoint table (irw_conjoint). Outcome: forced choice between two parties/candidates described by six policy proposals (wording not deposited; paraphrased from the authors’ code), exactly one chosen per task, no opt-out. One of four country tables from the same study; the domestic-policy text is specific to the host country. Respondents saw English or Traditional Chinese; only English level text is deposited. Qualtrics IDs re-keyed; the authors’ derived bins dropped. | 2026-10-08 | |
shum_2026_hk_diaspora_policy_ca |
Conjoint table (irw_conjoint). Outcome: forced choice between two parties/candidates described by six policy proposals (wording not deposited; paraphrased from the authors’ code), exactly one chosen per task, no opt-out. One of four country tables from the same study; the domestic-policy text is specific to the host country. Respondents saw English or Traditional Chinese; only English level text is deposited. Qualtrics IDs re-keyed; the authors’ derived bins dropped. | 2026-10-08 | |
shum_2026_hk_diaspora_policy_uk |
Conjoint table (irw_conjoint). Outcome: forced choice between two parties/candidates described by six policy proposals (wording not deposited; paraphrased from the authors’ code), exactly one chosen per task, no opt-out. One of four country tables from the same study; the domestic-policy text is specific to the host country. Respondents saw English or Traditional Chinese; only English level text is deposited. Qualtrics IDs re-keyed; the authors’ derived bins dropped. | 2026-10-08 | |
shum_2026_hk_diaspora_policy_us |
Conjoint table (irw_conjoint). Outcome: forced choice between two parties/candidates described by six policy proposals (wording not deposited; paraphrased from the authors’ code), exactly one chosen per task, no opt-out. One of four country tables from the same study; the domestic-policy text is specific to the host country. Respondents saw English or Traditional Chinese; only English level text is deposited. Qualtrics IDs re-keyed; the authors’ derived bins dropped. | 2026-10-08 | |
sievert_2024_burden_perceptions |
Conjoint table (irw_conjoint_2). One vignette set per respondent, 2x2x2 between subjects: three German Arbeitsagentur information pages on ALG II (learning demands, compliance demands, sanctions), each shown or withheld ((not shown)); text is the German page text. Ratings: approval of the policy (1-7), spending preference (1-3 less/same/more), five 0-100 estimates (stored 1-101) of recipient prejudice items, nine 1-7 legitimacy items; no opt-out. The deposit stores only a group number 1-8; the mapping to the factors is inferred (binary coding) and confirmed against the paper’s Table 2 coefficients. Authors’ attention-check exclusions applied. | 2026-10-14 | |
simsalrbim_human |
Three studies of the same paired-choice task, pooled; cov_sample names the study. LargeValence_2017 and LargeValence_2018 share 7 pictures (Cat, Crow, Doctor, Fire, Frustrated, Lake, War); LowValence_2017 uses 7 others (Flowers, Food, Grass, House, Monkey, Sunset, Timber). No one saw both sets, so the two halves share no items. Each trial is two rows: resp 1 for the picture chosen, 0 for the one passed over; trial numbers restart within each study. | #1700 | 2026-10-01 |
singh_2021_chile_democracy |
Conjoint table (irw_conjoint). Up to five pairs of hypothetical democratic systems; choice = the system chosen (source Chosen_country; question wording not deposited), forced choice, no opt-out. 2,242 unanswered contests dropped; all respondents kept with cov_attention_pass (the authors keep passers only). Levels in Spanish as stored; restrictions and attribute order undocumented. | 2026-10-09 | |
singh_2022_compulsory_voting |
Conjoint table (irw_conjoint). choice = which of two candidates the respondent would vote for (wording not in the deposit; one chosen per task, no opt-out observed). Attribute text is the deposit’s English labels for a Spanish-language survey; ‘No Opinion’/‘No Information’/‘No Favorite…’ are stored as levels, though whether they were displayed as text is undocumented. The codebook’s 0 = voted for is contradicted by the data and analysis (1 = chosen used). The deposited day counts past ages 18/70 (which fix the date of birth, the RD running variable) are dropped; cov_age keeps completed years. N = 2,042. | 2026-10-08 | |
skigin_2023_victim_solidarity |
Conjoint table (irw_conjoint). Factorial vignette experiment, 4 single-victim vignettes per respondent. choice = donated for this victim (0/1; not donating is the outside option); rating_donation (amount, 0-150, unit not stated), rating_thermometer (0-100), rating_social_distance, rating_compassion, rating_norms (0-3). No wording or anchors in the deposit; meanings from the author’s variable names. Level text is the author’s English labels (survey in Spanish). Duplicated panel ids deduplicated; 1 id with two sessions dropped. | 2026-10-09 | |
sleiman_2025_online_dating |
Conjoint table (irw_conjoint). Outcome: forced choice of which of two dating profiles to select (wording not in the deposit), no opt-out. A fixed D-optimal design of 16 pairs shown to everyone in random order (order not recorded; task is the set number); race and attractiveness were conveyed by profile photos. Profile gender is set by respondent orientation, not randomized; IP, location, Prolific IDs and free text dropped. | 2026-10-08 | |
smeele_2025_health_insurance |
Conjoint table (irw_conjoint). Discrete choice experiment, 14 tasks of two unlabelled policy alternatives; choice = forced choice between alternative 1 and 2 (wording not deposited), rating_certainty = certainty about the choice, 1 = very uncertain to 5 = very certain (asked per task, repeated on both rows). Levels are the codebook values (patient deaths x 1,000, signed; premium change in euros per month), not displayed text. Fixed blocked design; task = design scenario number. N = 708 as in README. | 2026-10-09 | |
smeele_2025_organ_transplant |
Conjoint table (irw_conjoint). Dual-response discrete choice experiment, 10 tasks; choice = stage-2 choice among alternative 1, 2 and a constant third alternative (current situation: 1,200 deaths, no premium change, good condition; stored as profile 3), choice_stage1 = forced choice between alternatives 1 and 2 (profile 3 not offered, 0). Wording not deposited; levels are codebook values (deaths x 100, premium change in euros per month, quality good/moderate). Survey arms 1-3 undocumented. N = 691 as in README. | 2026-10-09 | |
smith_2023_candidates_brazil |
Conjoint table (irw_conjoint). choice = “Em qual dos candidatos você votaria?” between Candidato A and B, forced (no opt-out); 3 tasks. Nine binary attributes independently randomized; the abortion attribute offers only keep current laws vs complete ban. The three policy attributes were shown first or last (trial_order); order within blocks randomized but not recorded. 1,817 respondents = the article’s valid N after its filter (IP in country, consent, citizen, resident, 18+, finished). Questionnaire codes for covariates differ by country. | 2026-10-07 | |
smith_2023_candidates_chile |
Conjoint table (irw_conjoint). choice = “¿Por cuál de los candidatos votaría usted?” between Candidato A and B, forced (no opt-out); 3 tasks. Nine binary attributes independently randomized; the abortion attribute offers only keep current laws vs complete ban. The three policy attributes were shown first or last (trial_order); order within blocks randomized but not recorded. 3,732 respondents = the article’s valid N after its filter (IP in country, consent, citizen, resident, 18+, finished). Questionnaire codes for covariates differ by country. | 2026-10-07 | |
smith_2023_candidates_peru |
Conjoint table (irw_conjoint). choice = “¿Por cuál de los candidatos votaría usted?” between Candidato A and B, forced (no opt-out); 3 tasks. Nine binary attributes independently randomized; the abortion attribute offers only keep current laws vs complete ban. The three policy attributes were shown first or last (trial_order); order within blocks randomized but not recorded. 3,698 respondents = the article’s valid N after its filter (IP in country, consent, citizen, resident, 18+, finished). Questionnaire codes for covariates differ by country. | 2026-10-07 | |
smith_2024_supply_chain |
Conjoint table (irw_conjoint_2). rating = whether the respondent would be in favour of (1) or opposed to (0) each proposal (‘If you were presented Proposal A [B] today, would you be opposed or in favour of it?’); both can be favoured; the forced choice between proposals is not in the deposit. 5 pairs, task/profile inferred from row order; text is the English India-version instrument. Indonesia is excluded: its support variable contradicts the opposed-to-both flag in a third of pairs. | 2026-10-11 | |
smith_2025_second_look |
Conjoint table (irw_conjoint_2). Each respondent judged three hypothetical second-look sentencing applications (text vignettes, eight randomized attributes); rating is the decision, 1 = release now, 2 = reduce sentence to 1-2 more years, 3 = deny but suggest reapplying in 3 years, 4 = deny permanently (higher = less favourable to release). Global-support item and its message arm, racial-resentment items and most demographics dropped; id is the row number; N = 800 as in the paper. | 2026-10-11 | |
smyk_2025_working_time |
Conjoint table (irw_conjoint_2). Three vignettes per respondent (fixed occupation order), one profile each; rating = ‘Jak Twoim zdaniem powinno zmienić się wynagrodzenie |
2026-10-09 | |
snagovsky_2020_candidates_australia |
Conjoint table (irw_conjoint). choice: which of two candidates (Q2: Candidate 1 / Candidate 2; question wording not deposited); ‘Don’t know’ was an answer option (1,077 tasks, both profiles 0); 176 refused tasks dropped (18 respondents lost). Ideology stored as two source label columns (political_id: Centre/Hard left/right; left_right), combined display not deposited. Observed restrictions: other language tied to cultural background, left_right tied to party. 5 pairs per respondent; W7 survey weight kept (missing for 78 respondents). | 2026-10-08 | |
sobczak_2024_bmast |
Deposited as MAST-weighted scores (0/2 for items 1, 2, 4-7, 10; 0/5 for items 3, 8, 9); shipped as 0/1 endorsement. | 2026-10-11 | |
song_2023_rses |
In this Chinese version of the Rosenberg Self-Esteem Scale, item 8 is scored in the positive direction, unlike the English original; the data reproduce the published scoring. | #2513 | 2026-09-28 |
song_2026_migrants_japan |
Outcome: choose the one of two immigrants whose admission you would support, even if unsure (forced, no opt-out). Attribute text is the authors’ English (shown in Japanese); the between-respondent framing arm is trial_framing and the post-stratification weight is cov_survey_weight. Respondent gender (cov_gender) is labelled from the appendix’s sex-by-age census tables; the other covariates keep the deposit’s unlabelled codes (education as cov_education_code) with missing codes set empty. Qualtrics IDs re-keyed. | #2887 | 2026-10-07 |
song_2026_migrants_korea |
Outcome: choose the one of two immigrants whose admission you would support, even if unsure (forced, no opt-out). Attribute text is the authors’ English (shown in Korean); the between-respondent framing arm is trial_framing and the post-stratification weight is cov_survey_weight. Respondent gender (cov_gender) is labelled from the appendix’s sex-by-age census tables; the other covariates keep the deposit’s unlabelled codes (education as cov_education_code) with missing codes set empty. Qualtrics IDs re-keyed. | #2887 | 2026-10-07 |
sorelle_2024_student_debt |
Conjoint table (irw_conjoint). choice = which borrower most deserves to have a significant portion of their student loan debt forgiven (forced, no opt-out); rating = how deserving each borrower is, 1 Very Undeserving to 5 Very Deserving. Only tasks 1-4 are kept: the deposit records race only for those, whether race was shown in tasks 5-6 is undocumented, and the authors analyse tasks 1-4. The authors’ code gives profile B profile A’s repayment history; this table carries the correct value, so its AMCEs differ slightly from the published ones. | 2026-10-07 | |
spilker_2016_pta_partners_costa_rica |
Conjoint table (irw_conjoint). choice = which of two partner countries the respondent prefers Costa Rica to choose for a trade agreement (forced choice, no opt-out); rating = support for an agreement with each country, 1 not support at all to 7 strongly support. Full independent randomization; attribute order randomized per respondent but not recorded. Level text is the authors’ English labels, not the Spanish shown; 266 tasks dropped (political system missing in 311 rows; 13 tasks with no choice). | 2026-10-07 | |
spilker_2016_pta_partners_nicaragua |
Conjoint table (irw_conjoint). choice = which of two partner countries the respondent prefers Nicaragua to choose for a trade agreement (forced choice, no opt-out); rating = support for an agreement with each country, 1 not support at all to 7 strongly support. Full independent randomization; attribute order randomized per respondent but not recorded. Level text is the authors’ English labels, not the Spanish shown; 786 tasks dropped (economic size missing for task 5 profile 1 of 536 respondents; political system missing in 335 rows). | 2026-10-07 | |
spilker_2018_trade_costa_rica |
choice = “Which agreement would you prefer?” (forced, no opt-out); rating = 1 never support to 7 always support. Task 4 dropped because the deposit copies task 1’s first-profile outcomes into it; 821 respondents against the paper’s 820. Unlabelled covariates dropped. | #2887 | 2026-10-07 |
spilker_2018_trade_nicaragua |
choice = “Which agreement would you prefer?” (forced, no opt-out); rating = 1 never support to 7 always support. Task 4 dropped because the deposit copies task 1’s first-profile outcomes into it; N = 800 matches the paper. Unlabelled covariates dropped. | #2887 | 2026-10-07 |
spilker_2018_trade_vietnam |
choice = “Which agreement would you prefer?” (forced, no opt-out); rating = 1 never support to 7 always support. Task 5 dropped because one profile’s attribute codes are corrupt in the deposit; N = 700 matches the paper. Unlabelled gender/person columns dropped. In every task the chosen agreement is also rated strictly higher (2,800 of 2,800, no ties; Costa Rica 85% with 9% ties, Nicaragua 76% with 10%), so choice and rating were probably derived from or forced consistent with each other here: treat them as one answer, not two. | #2887 | 2026-10-07 |
sscait_ladder_2021_2024 |
No game ends in a draw. Under SSCAIT’s rules a crashing bot loses (crashed names it, 3,538 games), and a game that hits the time limit (90 in-game minutes, or 5 minutes with no unit dying) goes to the bot with the higher kills+razings score (timelimit = 1, 1,437 games; the API notes these as “draw;”). Filter on both if only games played to destruction are wanted. | #2681 | 2026-10-01 |
stamm_2023_vaccine_campaign |
Conjoint table (irw_conjoint). Outcomes: forced choice of the more appealing of two vaccination-call scenarios and a 0-10 rating of the likelihood of heeding each (0 ‘would definitely not get vaccinated’ .. 10 ‘would definitely get vaccinated’); no opt-out. Austria and Italy pooled (cov_country), level text is the English version of the German and Italian instruments. The testimonial was shown as a first name and age that were not saved; only its gender and same/different age group are stored. Levels drawn with equal probability from the full factorial. | 2026-10-08 | |
stamm_2023_vaccine_media |
Conjoint table (irw_conjoint). Outcomes: forced choice of the media report that makes vaccination more trustworthy and a 0-10 rating of the likelihood of getting vaccinated given each report (0 ‘would definitely not get vaccinated’ .. 10 ‘would definitely get vaccinated’); no opt-out. Austria and Italy pooled (cov_country); level text is the English version of the German and Italian instruments. Levels drawn with equal probability from the full factorial. | 2026-10-08 | |
stamuli_2020_fnp_outcomes_a |
Conjoint table (irw_conjoint). choice: “We would like you to choose which scenario you prefer out of a choice of two” (Scenario A or B; forced, no opt-out). Fixed blocked design A (15 sets); the two scenarios always differ on every attribute. task is the choice-set number, not display order. Respondents who did one design and those who did all four (trial_group) are pooled, as in the article; panel ID dropped. | 2026-10-08 | |
stamuli_2020_fnp_outcomes_b |
Conjoint table (irw_conjoint). choice: “We would like you to choose which scenario you prefer out of a choice of two” (Scenario A or B; forced, no opt-out). Fixed blocked design B (15 sets); the two scenarios always differ on every attribute. task is the choice-set number, not display order. Respondents who did one design and those who did all four (trial_group) are pooled, as in the article; panel ID dropped. | 2026-10-08 | |
stamuli_2020_fnp_outcomes_c |
Conjoint table (irw_conjoint). choice: “We would like you to choose which scenario you prefer out of a choice of two” (Scenario A or B; forced, no opt-out). Fixed blocked design C (15 sets); the two scenarios always differ on every attribute. task is the choice-set number, not display order. Respondents who did one design and those who did all four (trial_group) are pooled, as in the article; panel ID dropped. | 2026-10-08 | |
stamuli_2020_fnp_outcomes_d |
Conjoint table (irw_conjoint). choice: “We would like you to choose which scenario you prefer out of a choice of two” (Scenario A or B; forced, no opt-out). Fixed blocked design D (16 sets); the two scenarios always differ on every attribute. task is the choice-set number, not display order. Respondents who did one design and those who did all four (trial_group) are pooled, as in the article; panel ID dropped. | 2026-10-08 | |
stapleton_2021_lotr |
Items rescaled from the deposit’s 0-1 coding back to 1-5; the three pessimism items are stored in the pessimism direction (high = pessimistic), as deposited. Pooled 2008 CCAP and 2018 CCES samples (cov_study). | 2026-10-10 | |
stapleton_2021_tipi |
Items rescaled from 0-1 back to 1-7; the five negatively keyed adjectives (reserved, disorganized, conventional, critical, calm) were reversed by the authors before deposit, so every item is in the trait direction. | 2026-10-10 | |
stecula_2020_antivax |
wave = the survey’s own wave numbers 4 and 6; 2,771 and 1,924 respondents. | 2026-10-10 | |
steenwegen_2024_community_education |
Conjoint table (irw_conjoint_2). One flyer per respondent; rating_overall, rating_children and rating_society = whether it is good or bad (overall / for the community’s children / for Flemish society) that the community organizes such classes, 1 = Heel slecht to 7 = Heel goed. Full 3 x 2 factorial; English level descriptions stored for Dutch flyers; 2,653 respondents vs 2,650 in the article (3 without the second and third outcomes). | 2026-10-09 | |
stoetzer_2023_covid_triage |
Conjoint table (irw_conjoint). choice = which of two patients the respondent would prioritize for intensive care (forced choice, no opt-out; wording paraphrased from the preprint, not deposited); 2,053 unanswered pairs omitted. Arrival is randomized for patient A only and patient B’s arrival follows from it. Level text is the authors’ English labels; respondents saw English, German, Italian, Polish or Portuguese screens. Partisanship names a different left/right party per country (cov_country). The country-size weight and derived party-rating variables are dropped. | 2026-10-07 | |
stokes_2017_rps_support |
Conjoint table (irw_conjoint). Factorial vignette: a hypothetical state RPS bill (35% renewables by 2025) with five randomized statements, each with a control in which nothing was shown (‘(not shown)’); one vignette per respondent. rating: support for the bill on a 4-point scale from strongly oppose (1) to strongly support (4) (paraphrase; direction inferred from the authors’ binary recode); 43 non-respondents omitted. Level text is the authors’ short labels, not the displayed sentences. Deposit 2,564 rows / 2,521 answers vs paper’s ‘2,500’. | 2026-10-09 | |
stone_2025_judge_selection_2021 |
Conjoint table (irw_conjoint). One hypothetical trial-court judge profile per respondent (7 attributes, one value each randomly assigned, attribute order randomized but not recorded); rating = support for the judge, 1-5 strongly oppose to strongly support, rating_fairness = fairness, 1-5 very unfair to very fair; wording paraphrased from the README. Level text from article Table 1; partisanship level ‘No partisan information provided’ is Table 1’s label. 71 respondents with no treatment/outcome omitted (962 of 1033). | 2026-10-09 | |
stone_2025_judge_selection_2023 |
Conjoint table (irw_conjoint). One hypothetical trial-court judge profile per respondent (7 attributes, one value each randomly assigned, attribute order randomized but not recorded); rating = support for the judge, 1-5 strongly oppose to strongly support; wording paraphrased from the README. Level text from article Table 1; partisanship level ‘No partisan information provided’ is Table 1’s label. 52 respondents without the support item omitted (1172 of 1224). | 2026-10-09 | |
stoop_2025_host_communities |
Conjoint table (irw_conjoint). Forced choice of the host community the respondent would prefer to live in after imagining having fled violence (no opt-out); rating = how much they would like to live in each community, 0-4 raw (anchors undocumented; codebook says 1-4); five integration statements rated 0-4 (‘not at all’ .. ‘strongly agree’) in round 1 only. Level text is English; the survey was in Tshiluba with drawings. Attribute order randomized per respondent but not recorded. | 2026-10-08 | |
stress_deboeck2012 |
36 of the 185 people hold two or more complete response sets in the source (fsdatT/stressT, JSS v048c01 supplement), and the source does not say which set is which. So 792 rows repeat an (id, item, occasion) triple even after the occasion column. If your analysis needs unique (id, item, occasion), drop the repeated sets. | #1856 | 2026-09-30 |
strojny_2026_gdt |
19 respondents who appear twice in the deposit (15 rows identical on all 150 columns, 4 more with an identical 88-item GMI pattern) were dropped, keeping the first copy, so N is 911 rather than the paper’s 930. | 2026-10-04 | |
strojny_2026_gis |
Responses are self-reported minutes per day (one weekday and one weekend day for each of six gaming activities), not ratings; values run 0 to 1300 and are heavily skewed. 19 respondents who appear twice in the deposit (15 rows identical on all 150 columns, 4 more with an identical 88-item GMI pattern) were dropped, keeping the first copy, so N is 911 rather than the paper’s 930. | 2026-10-04 | |
strojny_2026_gmi |
19 respondents who appear twice in the deposit (15 rows identical on all 150 columns, 4 more with an identical 88-item GMI pattern) were dropped, keeping the first copy, so N is 911 rather than the paper’s 930. | 2026-10-04 | |
strojny_2026_igds9sf |
19 respondents who appear twice in the deposit (15 rows identical on all 150 columns, 4 more with an identical 88-item GMI pattern) were dropped, keeping the first copy, so N is 911 rather than the paper’s 930. | 2026-10-04 | |
su_2024_pss14 |
Items follow Cohen’s standard PSS-14 order (authors’ GitHub, njnklab/response-time-dataset). The positively worded items (4, 5, 6, 7, 9, 10, 13) are not reversed in the source or in IRW, and the source’s total score sums all 14 items without reversing any. | #2513 | 2026-09-28 |
suasapha_2023_tourism_education |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent scored 16 fixed stimulus cards (5 attributes, Indonesian text) describing a higher-education institution; rating = preference score 1-4 stored raw (larger = more preferred per the SPSS CONJOINT /SCORE syntax; the deposit gives no question wording or scale labels). Fixed orthogonal design, task = card number in column order. 238 complete respondents out of 259 form responses. | 2026-10-12 | |
sung_2023_chemical_compounds |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent rated 16 fixed stimulus cards (6 attributes) describing a chemical compound; rating = purchase probability 0-10 (higher = more likely to purchase), stored raw. Level text comes from the paper’s Table 1 and the mapping of the deposit’s dummies to levels follows its numbering; the descriptive covariates are aligned by row position as in the authors’ notebook. Task is the card order in the file. 130 respondents (the paper’s usable N). | 2026-10-12 | |
sung_2026_marital_quality |
The deposit stores five of the six items as label strings that disagree with the paper’s coding: read literally, no one of 616 spouses ever answers ‘often’. The shipped codes (Never=1, Hardly ever or rarely=2, Some of the time=3 for the four how-often items; Different/separate things=3, Some together=2, Together=1 for free time) are the only assignment that reproduces the deposit’s own positive and negative marital-quality composites exactly for every spouse, and they match the paper’s 1-3 coding; the strings appear to be shifted labels. | 2026-10-05 | |
suring_2025_forest_contracts |
Conjoint table (irw_conjoint). Labelled discrete choice experiment: 9 choices between Kontrakt A (management-based objective, profile 1) and Kontrakt B (results-based objective, profile 2) or ‘Ingen af dem’ (neither: OPT-OUT, choice 0 on both; 12,840 of 28,979 tasks). Objectives come in fixed pairs (deadwood-beetles, gaps-plants, age classes-birds). Compensation per ha for 20 years as shown: DKK in Denmark, EUR in Finland. Level text is the English codebook wording. Task inferred from row order. Danish pilots and unanswered tasks dropped; Danish N 1,716 vs 1,795 reported. | 2026-10-09 | |
syamsuar_2025_behavioral_intention |
The deposit has no codebook: this table is the BI1-5 block, read as behavioural intention to use antivirus software from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_effort_expectancy |
The deposit has no codebook: this table is the EE1-5 block, read as effort expectancy from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_facilitating_conditions |
The deposit has no codebook: this table is the FC1-5 block, read as facilitating conditions from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_perceived_severity |
The deposit has no codebook: this table is the PS1-5 block, read as perceived severity from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_perceived_vulnerability |
The deposit has no codebook: this table is the PV1-5 block, read as perceived vulnerability from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_performance_expectancy |
The deposit has no codebook: this table is the PE1-5 block, read as performance expectancy from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_protective_behavior |
The deposit has no codebook: this table is the PB1-5 block, read as protective cybersecurity behaviour from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_response_cost |
The deposit has no codebook: this table is the RC1-5 block, read as response cost from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_response_efficacy |
The deposit has no codebook: this table is the RE1-5 block, read as response efficacy from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_self_efficacy |
The deposit has no codebook: this table is the SE1-5 block, read as self-efficacy from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_social_influence |
The deposit has no codebook: this table is the SI1-5 block, read as social influence from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
syamsuar_2025_usage_behavior |
The deposit has no codebook: this table is the UB1-5 block, read as antivirus usage behaviour from the standard PMT/UTAUT abbreviation and the constructs named in the record’s abstract. | 2026-10-07 | |
takacs_2026_aces |
ACE items are kept as the source codes them: 1 = yes, 2 = no, so a higher value means the experience was NOT reported. | 2026-10-07 | |
takei_2024_public_threats |
Conjoint table (irw_conjoint). rating = ‘Given the information available, what is your best estimate about whether Country A will stand firm in this dispute, ranging form 0 percent to 100 percent?’ (0-100, higher = more likely to stand firm); choice = which of the two countries is more likely to stand firm (forced choice, no opt-out; the article quotes only the Kertzer et al. wording it adopted). No restrictions on combinations (so ‘the United States’ can be ‘a dictatorship’). Level text is the appendix sentence fragments; the wide popularity columns are coded opposite to the long label and the long label is followed. Authors’ inattention flags dropped. | 2026-10-07 | |
takei_2026_audience_costs |
Conjoint table (irw_conjoint_2). Ratings: threat credibility 1 very unlikely to 4 very likely, and situational and general approval at stage 1 and stage 2 on 1-7 composites (1 disapprove very strongly, 7 approve very strongly, from the authors’ do-file recode); question wordings are not in the deposit. Six two-level scenario fragments (regime, threat, military strength, motive, interest, region) are the attributes; authors’ main factors are regime and threat. The Korean pilot is not built (one factor). | 2026-10-11 | |
tavolacci_2017_audit |
Item codes are the deposit’s headers. AUDIT4 and AUDIT5 take only 0, 1, 2 and 4 (no 3), the pattern of the standard AUDIT items 9-10, while AUDIT9-10 take 0-4 and 0-3; the deposit’s numbering may therefore not follow the standard AUDIT order, and there is no codebook to settle it. AUDIT2-10 are blank for about 225 students (skip after AUDIT1). Four fractional or text cells dropped. | 2026-10-07 | |
tavolacci_2017_pss10 |
13 cells coded 5 or 6 (outside the PSS 0-4 range, scattered over six items) and three text missing codes dropped. No codebook; instrument identified from the ‘Cohen’ column prefix, 10 items and 0-4 range. | 2026-10-07 | |
tellez_2021_peace_packages |
Conjoint table (irw_conjoint_2). Forced choice between two hypothetical peace-agreement packages (A/B), 5 tasks, six 3-level attributes (land, elections, reintegration, justice, reparation, drug policy); choice = preferred package, no opt-out. The deposit stores short tags; attribute level text is the authors’ own plain-language labels from their analysis code (paraphrased), not the displayed wording. 9 respondents have 4 tasks. | 2026-10-14 | |
theodoridis_2023_discrimination |
Conjoint table (irw_conjoint). choice: which of two people (Person A / Person B) the respondent chose, in one of five between-subject contexts (Relative, Neighbor, Loan, Charity, Jury; trial_context); wording not deposited (paraphrase); forced choice, no opt-out. Each attribute was independently omitted from about 15% of profiles (the authors’ ‘Not Asked’ level, stored as (not shown)). Covariates keep YouGov codes; weight missing for 154 respondents. N = 2,654 matches appendix Table I. | 2026-10-08 | |
thontwa_2024_food_cassava |
Conjoint table (irw_conjoint). Discrete choice between two cassava options (which would you buy; paraphrase), no opt-out, exactly one chosen per card. Fixed blocked design of 8 versions x 8 cards; attribute text in French from the authors’ choice-card coding sheets, price in Congolese francs. Respondents come from three survey groups (trial_arm CG/IT/LT, meaning undocumented) and also answered the other four product experiments. | 2026-10-08 | |
thontwa_2024_food_maize |
Conjoint table (irw_conjoint). Discrete choice between two maize options (which would you buy; paraphrase), no opt-out, exactly one chosen per card. Fixed blocked design of 8 versions x 8 cards; attribute text in French from the authors’ choice-card coding sheets, price in Congolese francs. Respondents come from three survey groups (trial_arm CG/IT/LT, meaning undocumented) and also answered the other four product experiments. | 2026-10-08 | |
thontwa_2024_food_oil |
Conjoint table (irw_conjoint). Discrete choice between two cooking oil options (which would you buy; paraphrase), no opt-out, exactly one chosen per card. Fixed blocked design of 8 versions x 8 cards; attribute text in French from the authors’ choice-card coding sheets, price in Congolese francs. Respondents come from three survey groups (trial_arm CG/IT/LT, meaning undocumented) and also answered the other four product experiments. | 2026-10-08 | |
thontwa_2024_food_rice |
Conjoint table (irw_conjoint). Discrete choice between two rice options (which would you buy; paraphrase), no opt-out, exactly one chosen per card. Fixed blocked design of 8 versions x 8 cards; attribute text in French from the authors’ choice-card coding sheets, price in Congolese francs. Respondents come from three survey groups (trial_arm CG/IT/LT, meaning undocumented) and also answered the other four product experiments. | 2026-10-08 | |
thontwa_2024_food_sugar |
Conjoint table (irw_conjoint). Discrete choice between two sugar options (which would you buy; paraphrase), no opt-out, exactly one chosen per card. Fixed blocked design of 8 versions x 8 cards; attribute text in French from the authors’ choice-card coding sheets, price in Congolese francs. Respondents come from three survey groups (trial_arm CG/IT/LT, meaning undocumented) and also answered the other four product experiments. | 2026-10-08 | |
tian_2025_poms_anger |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tian_2025_poms_confusion |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tian_2025_poms_depression |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tian_2025_poms_fatigue |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tian_2025_poms_tension |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tian_2025_poms_vigor |
The record has no description or codebook. Items MD1-MD65 were identified as the POMS-65 in standard order because regressing each deposited subscale score on the item columns reproduces it exactly with the published POMS key; MD22_R and MD54_R are stored reverse-scored. MD66 (a 0-10 rating) and the seven unscored POMS adjectives are not shipped. | 2026-10-09 | |
tilley_2026_platform_ban |
Conjoint table (irw_conjoint_2). Paired politicians, 3 tasks x 2 profiles x 6 attributes. The answer to ‘Which of the two politicians do you think should be banned?’ (A only, B only, both, neither) is stored per profile as rating: 1 = should be banned, 0 = not (both = 1,1; neither = 0,0). The deposit’s variable labels for the office and years attributes are swapped; the value labels are used. Skipped answers dropped; attribute-order permutation not kept; unlabelled weight not kept; geography fields dropped; N not checked against the article. | 2026-10-11 | |
tilley_2026_protest_tolerance |
Conjoint table (irw_conjoint_2). Paired protests, 3 tasks x 2 profiles x 8 attributes. The answer to ‘Which of the two protests do you think should go ahead?’ (A only, B only, both, neither) is stored per profile as rating: 1 = this protest should go ahead, 0 = not (so both = 1,1 and neither = 0,0); the authors’ own newchoice reverses this to a restriction coding. Skipped answers dropped; attribute-order permutation not kept; unlabelled weight W8 not kept; postcode and constituency fields dropped; N not checked against the article. | 2026-10-11 | |
tilley_2026_school_speaker |
Conjoint table (irw_conjoint_2). Paired politicians invited to speak at a school, 3 tasks x 2 profiles x 6 attributes. The answer (A only, B only, both, neither) is stored per profile as rating: 1 = should be allowed to speak, 0 = not (both = 1,1; neither = 0,0); the authors’ newchoice reverses this to a restriction coding. Skipped answers dropped; attribute-order permutation not kept; unlabelled weight not kept; geography fields dropped; N not checked against the article. | 2026-10-11 | |
tingley_2021_paris_reply |
Conjoint table (irw_conjoint_2). Single-vignette factorial survey (Paris Agreement, wave 4): rating = overall approval of what the US government did, 1 = approve strongly to 5 = disapprove strongly (higher = less approval); emissions cut and energy price are ‘(not shown)’ when no laws were passed. All respondents kept (cov_attention_pass); no weight kept; N = 3,013. | 2026-10-11 | |
tingley_2021_paris_shamers |
Conjoint table (irw_conjoint_2). Single-vignette factorial survey (Paris Agreement, wave 6): rating = overall approval of what the US government did, 1 = approve strongly to 5 = disapprove strongly (higher = less approval); shamers attribute is ‘(not shown)’ when no criticism was shown. All respondents kept (cov_attention_pass); no weight kept; N = 3,525. | 2026-10-11 | |
tingley_2021_paris_shaming |
Conjoint table (irw_conjoint_2). Single-vignette factorial survey (Paris Agreement, waves 1, 2, 5): rating = overall approval of what the US government did, 1 = approve strongly to 5 = disapprove strongly (higher = less approval); emissions cut and energy price rise are ‘(not shown)’ when no laws were passed, and the criticism attribute is ‘(not shown)’ when none was shown. All respondents kept (the authors analyse only the attentive; cov_attention_pass); no weight kept; N = 3,012 + 635. | 2026-10-11 | |
tiradocastro_2026_gender_affinity |
Conjoint table (irw_conjoint). choice = which of the two candidates the respondent selected (paraphrase), forced choice, no opt-out; rating = 0-10 evaluation of each candidate (0 very negative, 10 very positive), present for chosen and non-chosen profiles although the codebook says the chosen one was rated. Task and profile are inferred from row order. Attribute levels are the codebook’s descriptions. | 2026-10-08 | |
titelman_2024_party_labels |
Conjoint table (irw_conjoint). choice = “If these were the only two candidates standing for Parliament in your constituency, which would you vote for?” (Candidate A / Candidate B / I am not sure / I would not vote); opt-out: 2,544 of 8,180 tasks have no chosen candidate (trial_no_choice says which). One Conservative and one Labour profile per task, whole profiles drawn from realistic candidate distributions (restricted), party label hidden for half of respondents (attr_party ‘(not shown)’); left-right stored as the 0-10 number without the displayed band word. N = 1,636 (808 labels / 828 no labels) as in the article. | 2026-10-09 | |
tjiong_2026_mobile_data |
Conjoint table (irw_conjoint). Discrete choice experiment, choice = chosen alternative among free campus Wi-Fi (status quo, profile 1, SP1/SP2) and 4G LTE data packages (wording paraphrased; no opt-out beyond the status quo profile). Three treatments (SP1 8 tasks, SP2 8, SP3 10) in fixed blocked designs; levels are the app’s display strings, Wi-Fi fee ‘(not shown)’ in SP3. 5 students with duplicate sessions dropped (N = 297 vs 302 in the paper). | 2026-10-09 | |
toenshoff_2025_climate_plans |
Conjoint table (irw_conjoint). choice: ‘Welchen Plan würden Sie bevorzugen?’ Forced choice, 5 tasks of 2 climate policy plans, German level text. Respondents were first randomized to one of three vignettes (trial_vignette: Control / Compensation Treatment / Carbon Tax Treatment; texts not deposited). Restrictions and attribute order undocumented. Sample = authors’ attention-check passers who finished (2,191); postcode, free text and platform ids dropped. | 2026-10-09 | |
toenshoff_2025_energy_candidates |
Conjoint table (irw_conjoint). choice: ‘Welche dieser beiden Kandidaten/Kandidatinnen würden Sie in der kommenden Bundestagswahl eher wählen? Bitte wählen Sie einen Kandidaten/eine Kandidatin aus, selbst wenn Ihnen keine der beiden Optionen gefällt.’ Forced choice, 5 tasks of 2 candidates, German level text. Energy policy levels unequally weighted (current speed 40%); restrictions and attribute order undocumented. Sample = authors’ attention-check passers who finished (2,191); postcode, free-text comments and Qualtrics/panel ids dropped; two full dates of birth set missing. | 2026-10-09 | |
tompa_2024_protein_sustainability |
Conjoint table (irw_conjoint). Choice-based conjoint: ‘Please choose the protein source that you think is more sustainable on the basis of the information at your disposal’ (English translation), two cards plus a ‘no choice’ option (opt-out; both cards 0). Fixed orthogonal design, the same 16 pairs for everyone; nutritional score, ecological score and price were each shown or not ((not shown)), with shown values fixed by product. Task order as displayed is not documented. All 639 respondents kept; the article excludes 98 (under 18 or diet due to disease) to reach 541. | 2026-10-08 | |
tomz_2019_israel_party |
Conjoint table (irw_conjoint_2). Israeli public, March 2016 (N=1,277), Hebrew display text: four pairs of hypothetical parties/candidates described by seven attributes (row order randomized per respondent); choice = preferred of the pair, derived from a 4-point preference question; also a 0-100 preference recode and a 0-10 rating of each; no opt-out. Question wording is not in the deposit. The authors drop respondents below 85% on attention checks; this table keeps them (cov_attention_pass flags). | 2026-10-11 | |
tomz_2019_us_party_apr |
Conjoint table (irw_conjoint_2). US public, April-May 2017 (N=2,051): four pairs of hypothetical parties/candidates described by seven attributes (row order randomized per respondent); choice = preferred of the pair, derived from a 4-point preference question; also a 0-100 preference recode and a 0-10 rating of each; no opt-out. Question wording is not in the deposit. The authors drop respondents below 85% on attention checks; this table keeps them (cov_attention_pass flags). | 2026-10-11 | |
tomz_2019_us_party_sep |
Conjoint table (irw_conjoint_2). US public, Sept-Oct 2017 (N=1,580): four pairs of hypothetical parties/candidates described by five randomized attributes (two more rows, economic policy and religion’s role, were held fixed and are dropped; row order randomized per respondent); choice = preferred of the pair, derived from a 4-point preference question; also a 0-100 preference recode and a 0-10 rating of each; no opt-out. Question wording is not in the deposit. The authors drop respondents below 85% on attention checks; this table keeps them (cov_attention_pass flags). | 2026-10-11 | |
tomz_2020_ballistic |
Conjoint table (irw_conjoint_2). One vignette per respondent (region (4), regime (2) and human-rights record (2)); rating = ‘Do you favor or oppose using the U.S. military to attack the country’s missile development sites?’ 1 = favor strongly to 5 = oppose strongly, stored as in the source (lower = more support for the strike). No opt-out (‘neither’ is a scale point). Ballistic study: region, regime and human-rights record are all randomized; level text is the exact randomization text. N = 1211 respondents. | 2026-10-11 | |
tomz_2020_hr_war_africa |
Conjoint table (irw_conjoint_2). One vignette per respondent (regime (democracy or not) and human-rights record (violates or not)); rating = ‘Do you favor or oppose using the U.S. military to attack the country’s nuclear development sites?’ 1 = favor strongly to 5 = oppose strongly, stored as in the source (lower = more support for the strike). No opt-out (‘neither’ is a scale point). Africa study: the region ‘in Africa’ is constant (dropped). Level text is the exact randomization text. N = 763 respondents. | 2026-10-11 | |
tomz_2020_hr_war_uk |
Conjoint table (irw_conjoint_2). One vignette per respondent (regime (democracy or not) and human-rights record (violates or not)); rating = ‘Do you favor or oppose using the British Armed Forces to attack the country’s nuclear development sites?’ 1 = favor strongly to 5 = oppose strongly, stored as in the source (lower = more support for the strike). No opt-out (‘neither’ is a scale point). UK study: level text is the short form decoded from randforce codes by the authors’ prepdata.do; the UK questionnaire shows the field names only. N = 1452 respondents. | 2026-10-11 | |
tomz_2020_hr_war_us |
Conjoint table (irw_conjoint_2). One vignette per respondent (regime (democracy or not) and human-rights record (violates or not)); rating = ‘Do you favor or oppose using the U.S. military to attack the country’s nuclear development sites?’ 1 = favor strongly to 5 = oppose strongly, stored as in the source (lower = more support for the strike). No opt-out (‘neither’ is a scale point). Main study: level text is the data’s condition labels, not the displayed sentences (questionnaire not in deposit). N = 1500 respondents. | 2026-10-11 | |
torcal_2026_epade_candidates |
Conjoint table (irw_conjoint). choice = “Which candidate do you prefer as Chancellor?” (Germany; Prime Minister in Portugal) (Candidate A / B, forced, no opt-out); rating = favourability of each candidate, 1 Very unfavorable to 7 Very favorable. 7 tasks, 9 two-level attributes randomized with randomized attribute order (order not recorded). Germany, Portugal and the Netherlands pooled (cov_country); level text is the master English wording (respondents saw German, Portuguese or Dutch). | 2026-10-09 | |
torcal_2026_epade_candidates_chile |
Conjoint table (irw_conjoint). choice = “Which candidate do you prefer as President of the Republic?” (Candidate A / B, forced, no opt-out); rating = favourability of each candidate, 1 Very unfavorable to 10 Very favorable. 7 tasks, 9 two-level attributes randomized with randomized attribute order (order not recorded). Level text is the English Chile questionnaire wording (respondents saw Spanish); the tax level differs from the EU version. | 2026-10-09 | |
torcal_2026_neighbours_argentina |
Conjoint table (irw_conjoint). choice: “Which profile would you prefer to have as your next-door neighbour?” (Neighbour A/B, forced, no opt-out; English master wording, shown translated). Levels are the panel file’s short English labels. Non-uniform level weights and party-ideology restrictions per the questionnaire; attribute order randomized per respondent but not recorded. cov_in_authors_sample = 0 for respondents without the right to vote, whom the authors drop (975 of 977 kept by them; counts match their files exactly). | 2026-10-07 | |
torcal_2026_neighbours_chile |
Conjoint table (irw_conjoint). choice: “Which profile would you prefer to have as your next-door neighbour?” (Neighbour A/B, forced, no opt-out; English master wording, shown translated). Levels are the panel file’s short English labels. Non-uniform level weights and party-ideology restrictions per the questionnaire; attribute order randomized per respondent but not recorded. cov_in_authors_sample = 0 for respondents without the right to vote, whom the authors drop (916 of 921 kept by them; counts match their files exactly). | 2026-10-07 | |
torcal_2026_neighbours_italy |
Conjoint table (irw_conjoint). choice: “Which profile would you prefer to have as your next-door neighbour?” (Neighbour A/B, forced, no opt-out; English master wording, shown translated). Levels are the panel file’s short English labels. Non-uniform level weights and party-ideology restrictions per the questionnaire; attribute order randomized per respondent but not recorded. cov_in_authors_sample = 0 for respondents without the right to vote, whom the authors drop (986 of 997 kept by them; counts match their files exactly). | 2026-10-07 | |
torcal_2026_neighbours_portugal |
Conjoint table (irw_conjoint). choice: “Which profile would you prefer to have as your next-door neighbour?” (Neighbour A/B, forced, no opt-out; English master wording, shown translated). Levels are the panel file’s short English labels. Non-uniform level weights and party-ideology restrictions per the questionnaire; attribute order randomized per respondent but not recorded. cov_in_authors_sample = 0 for respondents without the right to vote, whom the authors drop (821 of 823 kept by them; counts match their files exactly). | 2026-10-07 | |
torcal_2026_neighbours_spain |
Conjoint table (irw_conjoint). choice: “Which profile would you prefer to have as your next-door neighbour?” (Neighbour A/B, forced, no opt-out; English master wording, shown translated). Levels are the panel file’s short English labels. Non-uniform level weights and party-ideology restrictions per the questionnaire; attribute order randomized per respondent but not recorded. cov_in_authors_sample = 0 for respondents without the right to vote, whom the authors drop (1,071 of 1,080 kept by them; counts match their files exactly). | 2026-10-07 | |
tranvag_2021_precision_medicine |
Conjoint table (irw_conjoint_2). choice = ‘Given the information available, to whom would you allocate the new drug? [Choose A or B]’ (translated; forced choice, no opt-out, up to 6 tasks); the other patient receives standard treatment. 96 tasks with unsaved attributes and 40 without an answer were dropped, leaving 131 respondents (the article reports 115 completed responses and 690 observations, 21 partial respondents here); education wording from the article, not the deposit. | 2026-10-11 | |
traunmuller_2020_religion_vignette |
Conjoint table (irw_conjoint_2). One vignette per respondent (group origin x religiosity x religion x authorities propose to ban or permit); ratings: support for the ban or permit (1 strongly support to 5 strongly oppose, lower = more support, asked only in the matching arm), feelings thermometer 0-100, deservingness of an unemployed single mother from the group (0-10), and whether the group may vote or hold public office (1 agree strongly to 5 disagree strongly). No opt-out; the deposit has only factor labels, not the vignette wording; same data as the BJPolS ‘What is Islamophobia?’ deposit. | 2026-10-11 | |
treger_2022_policy_party_cues |
Conjoint table (irw_conjoint_2). Eight forced-choice pairs of low-salience policy proposals (five attributes: instrument, public support, effectiveness, promoter, timing) in four issue domains, US sample; choice = preferred proposal (no opt-out) and rating = support for each proposal, 1-7 (wording and anchors not in the deposit). Pair index is the task; display order of pairs not documented. The article analyses only the 924 Democrats and Republicans; all 1,370 are kept. | 2026-10-11 | |
treger_2025_policy_tools_il |
Conjoint table (irw_conjoint). rating = support for the proposed policy on 1-7 (higher = more support; wording not deposited, paraphrased from the author’s do-file); choice = forced choice between the two proposals, no opt-out. Within a task the two profiles never share a coercion level; the policy goal is fixed per task (trial_goal). Attribute levels are the author’s English value labels, the Hebrew display text is not deposited. 69 respondents flagged as low quality by the author (dropped in her analyses) are kept with cov_low_quality = 1. | 2026-10-08 | |
treger_2025_policy_tools_us |
Conjoint table (irw_conjoint). rating = support for the proposed policy on 1-7 (higher = more support; wording not deposited, paraphrased from the author’s do-file); choice = forced choice between the two proposals, no opt-out. Within a task the two profiles never share a coercion level; the policy goal is fixed per task (trial_goal). Attribute levels are the author’s English value labels. 200 respondents flagged as low quality by the author (dropped in her analyses) are kept with cov_low_quality = 1. | 2026-10-08 | |
treger_2026_party_or_policy |
Conjoint table (irw_conjoint). choice: one of two candidates chosen (wording not in deposit); rating: 1-7 rating of each candidate (anchors not in deposit; higher goes with being chosen). No opt-out recorded; 339 of 3,393 tasks have ratings but no choice (choice NA). Party label shown only in a randomized between-respondent arm (trial_arm), otherwise ‘(not shown)’. Task/profile inferred from row order; derived congruence factors and the Qualtrics ResponseId dropped. | 2026-10-09 | |
treglown_2016_burnout |
Item BOUT4REC is the source’s reverse-scored version of CBI work-burnout item 4; the response as administered was not deposited. Higher values on the other five items mean more burnout; on BOUT4REC the source has already aligned it to the same direction. | #2876 | 2026-10-06 |
trexler_2025_ideology_labels |
Conjoint table (irw_conjoint). Seven paired candidate choices on a 4-point scale (1 Strongly prefer A, 2 Slightly prefer A, 4 Slightly prefer B, 5 Strongly prefer B; codebook paraphrase), stored as choice (preferred candidate) plus the raw code in trial_preference; no opt-out. Each profile shows 5 of 27 issues (attr_ |
2026-10-09 | |
trinh_2026_family_wealth_transfers |
Conjoint table (irw_conjoint). Factorial vignette (4 per respondent, German text). rating: ‘Ist der Anteil den [Name] bekommen hat, Ihrer Meinung nach…’ (is the share [name] received too little, appropriate or too much), 11 points -5 too little .. 0 appropriate .. +5 too much, stored as answered. Fixed blocked D-efficient design from a 648-cell full factorial; gift vs inheritance between respondents. Representative (1,400) and affluent (1,592) samples pooled; levels are the authors’ English labels; authors’ recodes dropped. | 2026-10-09 | |
trondle_2023_phaseout_heating |
Conjoint table (irw_conjoint). choice = “Which package of measures [framing text] would you rather accept?” (forced, no opt-out); rating = “To what extent are you in support or opposition of the respective package of measures?” 1 Opposed .. 5 Support. German level text; trial_framing (phase-out vs renewables) and trial_first_sector record the between-subject framing and sector order. N = 1,870 complete responses vs 1,777 in the paper. | 2026-10-09 | |
trondle_2023_phaseout_transport |
Conjoint table (irw_conjoint). choice = “Which package of measures [framing text] would you rather accept?” (forced, no opt-out); rating = “To what extent are you in support or opposition of the respective package of measures?” 1 Opposed .. 5 Support. German level text; trial_framing (phase-out vs renewables) and trial_first_sector record the between-subject framing and sector order. N = 1,870 complete responses vs 1,777 in the paper. | 2026-10-09 | |
trump_2025_tax_rich_dk |
Conjoint table (irw_conjoint). Factorial vignette (Study 3, Denmark), three per respondent, one profile each; rating: income taxes for people like [him/her] should be 1 Reduced a lot … 5 Increased a lot (Danish: Reduceres meget … Øges meget). 8 slots randomized independently. Level text stored in English (codebook master text); respondents saw Danish. Survey weight kept. N 1,234 matches the paper. | 2026-10-09 | |
trump_2025_tax_rich_pilot |
Conjoint table (irw_conjoint). Factorial vignette (pilot), three per respondent, one profile each; rating: ‘What, in your opinion, would be fair when it comes to income taxes for people like [him/her]? Do you think their income taxes should be:’ 1 Reduced a lot … 5 Increased a lot. 8 slots randomized independently (income dollar and percentile drawn together); gender is the pronoun used throughout. Prolific convenience sample (US), all 599 respondents. | 2026-10-09 | |
trump_2025_tax_rich_us |
Conjoint table (irw_conjoint). Factorial vignette (Study 2), three per respondent, one profile each; rating: ‘What, in your opinion, would be fair when it comes to income taxes for people like [him/her]? Do you think their income taxes should be:’ 1 Reduced a lot … 5 Increased a lot. 8 slots randomized independently; loophole clause text includes the gendered pronouns. Includes 215 screener-passers outside the article’s sample (partials / missing Lucid income): cov_authors_sample = 1 gives the article’s 1,825. | 2026-10-09 | |
trump_endorsement_2022_mitchell |
Conjoint table (irw_conjoint_2). Factorial vignette (US, 2022): one vignette per respondent about candidate Terry Mitchell, 2 policy-stance packages x 3 Trump-endorsement arms (endorsed / will not endorse / sentence omitted = ‘(not shown)’); rating = likelihood of voting for him 1-5 (higher = more likely, from the authors’ rescaling) and rating_favorability 0-100 (wording not deposited). Only this experiment of an omnibus poll is included; covariates keep codes (no codebook); national weight kept. | 2026-10-14 | |
turnbulldugarte_2022_surrogacy |
Conjoint table (irw_conjoint_2). Factorial vignette (Britain): one vignette per respondent describing a couple seeking surrogacy, 2x2 of gay/straight couple x celebrity/ordinary couple; rating = attitude to surrogacy, 1-5 as recorded (higher = more supportive, inferred from the authors’ support = 4-5; wording and anchors not in the deposit). Level text is the authors’ figure labels, not the displayed wording (tags only). No-couple control arm and 25 respondents without outcome dropped. | 2026-10-14 | |
turnbulldugarte_2023_ivoting |
Conjoint table (irw_conjoint). choice: which of two i-voting reform proposals the respondent supports (paraphrase; forced, no opt-out); rating_support and rating_trust: 0-6 ratings of support for the proposal and of its trustworthiness relative to in-person voting (paraphrase; anchors not deposited; higher = more, inferred from the authors’ rescaling). Administered-by stored as one 6-level attribute (the authors analyse it as platform x private IT firm). Qualtrics ResponseId labels and unlabelled survey items dropped; 2 profiles lack a trust rating. | 2026-10-08 | |
turnbulldugarte_2025_dating_spain |
Conjoint table (irw_conjoint_2). Single-profile design: each respondent judged 10 dating-app profiles one at a time; choice = authors’ ‘selected’ (response code 1 = 0, codes 2 and 4 = 1; reject is the outside option, opt-out yes), raw code in trial_response_code. Question text and code labels are not in the deposit; flag, orientation, trans and photo-congruence levels are the authors’ labels; (not shown) is inferred from empty fields (party hashtag absent in the randomized no-party-label arm). Profile gender follows the respondent’s preference and is not randomized. Spain sample, 1,949 respondents, 19,187 profile rows (71 unfinished rows dropped). | 2026-10-11 | |
turnbulldugarte_2025_dating_uk |
Conjoint table (irw_conjoint_2). Single-profile design: each respondent judged 10 dating-app profiles one at a time; choice = authors’ ‘selected’ (response code 1 = 0, codes 2 and 4 = 1; reject is the outside option, opt-out yes), raw code in trial_response_code. Question text and code labels are not in the deposit; flag, orientation, trans and photo-congruence levels are the authors’ labels; (not shown) is inferred from empty fields (party hashtag absent in the randomized no-party-label arm). Profile gender follows the respondent’s preference and is not randomized. UK sample, 2,023 respondents, 20,062 profile rows (168 unfinished rows dropped); Prolific IDs dropped; file has no respondent age or education. | 2026-10-11 | |
tyllianakis_2025_agri_schemes_uk |
Conjoint table (irw_conjoint). Labelled DCE, 2 blocks x 6 cards, four scheme alternatives plus ‘No scheme’ (opt-out: 54 tasks with no chosen profile); choice = ‘Please choose your preferred scheme’; rating_land_share (chosen scheme only) = % of arable land or permanent grassland the respondent would enrol, 0-100. Payment levels are scheme-specific. Card contents transcribed from the questionnaire images. England sample only (N = 107, as in the article); German and Czech samples (<100 each) not included. | 2026-10-09 | |
umit_2022_wind_turbines |
Conjoint table (irw_conjoint). Factorial vignette (2x2x2, one vignette per respondent); rating = ‘Do you support or oppose these wind turbine proposals?’ 1-5 (5 = strongly support); rating_turnout, rating_vote_sp, rating_vote_svp = likelihood of voting in the 2019 federal election / for the SP / for the SVP, 1-5 (5 = very likely; eligible voters only). Don’t know is NA (authors’ recode); 145 respondents with no recorded treatment dropped; N = 3,984 of the 4,151 surveyed. | 2026-10-09 | |
useche_2022_cbq |
No respondent id in the source; id is row order. 229 rows exactly duplicate another row on all 55 source columns (mostly adjacent pairs, plus a few all-zero/all-four straight-liners). They are kept: the authors report removing duplicate MAC addresses before deposit and analyse N = 7001. Items are in questionnaire order (subscale means in the deposit confirm the F1/F2/F3 boundaries). | 2026-10-07 | |
usmani_2024_iwe |
Two questionnaires appear to have been entered twice (identical on all 39 items), so the second copy of each was dropped and N is 123, not the paper’s 125. | 2026-10-04 | |
usmani_2024_ks |
Two questionnaires appear to have been entered twice (identical on all 39 items), so the second copy of each was dropped and N is 123, not the paper’s 125. | 2026-10-04 | |
usmani_2024_pops |
Two questionnaires appear to have been entered twice (identical on all 39 items), so the second copy of each was dropped and N is 123, not the paper’s 125. The 12 items are three Kacmar-Ferris POPS subscales (general political behavior PPGP1-6, go along to get ahead PPGAL1-4, pay and promotion PPPP1-2); the paper scores them as one scale, and the 2-item pay-and-promotion subscale is short. | 2026-10-04 | |
uti_newlands_2023_wpai |
WPAI items 2-4 (hours missed because of health, hours missed for other reasons, hours actually worked, past 7 days) are self-reported time, so they are covariates rather than items: cov_wpai_hours_missed_health, cov_wpai_hours_missed_other and cov_wpai_hours_worked, on the rows of the wave they were reported in. Items are wpai_1 (employed), wpai_5 and wpai_6 (0-10 impairment ratings). | #1700 | 2026-10-01 |
valenzuela_2026_genai_news |
Conjoint table (irw_conjoint). Three tasks of Outlet A vs Outlet B; choice_credibility = ‘If you had to choose one, which do you think could be the most reliable to report the news in a complete, accurate, fair and unbiased manner?’, choice_selectivity = ‘Let’s say you have to pick one of these outlets to be up to date in current affairs … Which one would you choose?’ (English translation), both forced, no opt-out. Levels are the Spanish text shown; attribute order randomized per respondent (attrpos_); uniform randomization, identical profiles excluded. N = 2,145 as in the authors’ code. | 2026-10-09 | |
van_beveren_2019_cdi |
Non-integer cells (imputations) were dropped; only observed responses ship. | 2026-10-10 | |
van_beveren_2019_feelkj |
Non-integer cells (regression-style imputations on the 1-5 scale) were dropped; only observed responses ship. | 2026-10-10 | |
van_beveren_2019_panasc |
Non-integer cells (imputations) were dropped; only observed responses ship. | 2026-10-10 | |
vanderdoes_2021_pb_cost |
Conjoint table (irw_conjoint). rating = approval of each neighbourhood budget, 1 completely disapprove to 7 completely approve (article Figure 1, translated). The forced choice between the two budgets is not deposited. Only 5 of 6 randomized attributes: the winning project shown is not in the deposit (the authors’ derived favorability, the respondent’s own rating of that project > 4, is dropped). Levels are the authors’ English labels; attribute order not documented. Study 1 of this deposit is already in vanderdoes_2022_participatory_budgeting. | 2026-10-09 | |
vanderdoes_2021_pb_location |
Conjoint table (irw_conjoint). rating = approval of each neighbourhood budget, 1 completely disapprove to 7 completely approve (article Figure 1, translated). The forced choice between the two budgets is not deposited. The winning project shown in both budgets was the respondent’s own top-ranked project; attribute project_location varies how far from home it is built. Levels are the authors’ English labels; attribute order not documented. No survey weight for this study. | 2026-10-09 | |
vanderdoes_2022_participatory_budgeting |
Conjoint table (irw_conjoint). choice = preferred neighbourhood budget of two (forced choice; verbatim wording not published); rating = approval of each budget, 1 completely disapprove to 7 completely approve. Only 4 of 5 randomized attributes: the winning project shown is not in the deposit (the authors’ derived favorability, respondent rating of that project > 4, is dropped). Levels are the authors’ English labels. Attribute order randomized, not recorded. cov_ codes unlabelled except as noted in the script; cov_pref_* = respondent’s 1-7 preference for each project type. | 2026-10-07 | |
vandervoet_2024_performance_info |
Conjoint table (irw_conjoint). Discrete choice experiment, choice = the issue the respondent would choose to prioritize (article paraphrase), 2 issues per set, 4 sets, no opt-out; fixed blocked design, domain (roads or schools) per respondent in trial_service. Level text = the article’s English Table 2 wording mapped from Included/Not Included codes (displayed in Dutch/French). Age and tenure dropped (re-identification risk). | 2026-10-09 | |
vanlit_2025_democratic_defence |
Conjoint table (irw_conjoint_2). Factorial vignette (NL, DE, FR pooled): one news release per respondent about a proposed autocratizing organization, varying target (media/judiciary), justification (none ‘(not shown)’, anti-corruption, government control) and a Supreme Court defence (‘(not shown)’ or strongly negative opinion); rating = ‘how much more or less democratic would this action make Norway’, 1-7, higher = more democratic; rating_ambiguity and rating_defender_credibility (1-6, stored after the authors’ reversal; credibility only NL with a defence). Level text paraphrased from the authors’ dissertation chapter; the seven-item protest battery is not included. | 2026-10-14 | |
vanlit_2026_democratic_defence_status |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent, 2x2: the defending parliamentarian’s party is in government or opposition and is consistent or inconsistent. Outcome rating_credibility is a 1-7 rating of the credibility of the democratic defence (higher = more credible, assumed; wording and anchors not in the deposit); no opt-out. Level text is the authors’ condition labels because the vignette wording was unavailable. cov_survey_weight is the authors’ weight (missing for some). 4,785 respondents, matching the paper. | 2026-10-14 | |
vanloon_2026_older_adults_wellbeing |
Conjoint table (irw_conjoint_2). Factorial survey (text vignettes, one per task): rating = 0-10 satisfaction rating of how much well-being the described situation would bring (higher = more; wording and anchors not documented, paraphrase). 7 vignettes per respondent per wave over up to 5 waves (task = (wave-1)*7 + order); no choice or opt-out. Own-life items dropped; wave-1 n differs from the 2022 article (1,003 vs 800). | 2026-10-11 | |
vanoosten_2024_politicians_de |
Conjoint table (irw_conjoint). Text vignettes in 3 pairs (German). choice: “Which politician are you most likely to vote for?” (forced, no opt-out); rating_represents “Do you think this politician represents you?”, rating_trust “How much do you trust this politician?”, rating_capable “How capable do you think this politician is to perform well on the job?”, each stored as code 1-11 = displayed 0 (Überhaupt nicht) .. 10 (Sehr). Names drawn from background/gender lists, but 228 Turkish-background profiles carry German first names. Part 3a profiles dropped. N = 954. | 2026-10-08 | |
vanoosten_2024_politicians_fr |
Conjoint table (irw_conjoint). Text vignettes in 3 pairs (French). choice: “Which politician are you most likely to vote for?” (forced, no opt-out); rating_represents, rating_trust, rating_capable (0-10 scales) stored as code 1-11 = displayed 0 (Pas du tout) .. 10 (Oui, tout à fait). Sub-Saharan African background shown about twice as often as each other background; 182 Turkish-background profiles carry North-African first names. Religion level ‘pratique l’Islam’/‘pratique l’islam’ differs only in capitalisation by background, as stored. Part 3a profiles dropped. N = 1,199. | 2026-10-08 | |
vanoosten_2024_politicians_nl |
Conjoint table (irw_conjoint). Text vignettes in 3 pairs. choice: “Which politician are you most likely to vote for?” (forced, no opt-out); rating_represents “Do you think this politician represents you?”, rating_trust “How much do you trust this politician?”, rating_capable “How capable do you think this politician is to perform well on the job?”, each stored as code 1-11 = displayed 0 (Not at all) .. 10 (Very much). Names drawn from background/gender lists; Surinamese profiles about twice as frequent. Part 3a profiles dropped. N = 905. | 2026-10-08 | |
vargiu_2026_incivility_tweets |
Conjoint table (irw_conjoint). No choice question; five 0-100 ratings per profile: rating_impolite (polite-impolite), rating_unacceptable (acceptable-unacceptable), rating_notnormal (normal-not normal), rating_wrong (right-wrong), all kept as asked (higher = more uncivil), and rating = feeling thermometer toward the candidate (cold/unfavourable 0 - warm/favourable 100). Exact item wording not deposited (report paraphrase). Restrictions not documented. 70 respondents with no ratings or unsaved attributes omitted; cov_in_authors_sample marks the authors’ 1,884 (attention, straight-lining, speed), of whom 1,843 are in the table. | 2026-10-07 | |
vasquez_2013_reading_comprehension |
The nine literal-comprehension items (cl*) are answered correctly by all 120 pupils, as the deposit’s own literal total (9 for everyone) shows. Items c021 and c027 are renamed co21 and co27 (a zero typed for the letter o). | 2026-10-09 | |
vaughn_2025_police_park_r1 |
Conjoint table (irw_conjoint_2). Factorial vignette, one park police-encounter vignette per respondent (US, replication of Study 1, 2024): civilian race x officer race x demeanor of officer and civilian (4 pairs) x officer decision (escort or ticket). Ratings 0-4 of fairness to the civilian, of the civilian to the officer and of the decision, plus preferred punishment (0-5); paraphrased wording. 1,529 of 1,546 rows have an outcome (17 omitted); no respondent id. | 2026-10-14 | |
vaughn_2025_police_park_s1 |
Conjoint table (irw_conjoint_2). Factorial vignette, one park police-encounter vignette per respondent (US, 2021): officer race x demeanor of officer and civilian (5 pairs) x officer decision (escort or ticket). Five ratings: expected punishment (1-5), fairness before and after the decision (1-5), preferred punishment (1-6), overall fairness (1-5); question wording is paraphrased from the authors’ labels and the early items precede the decision attribute. One of three studies in the deposit; no respondent id (id = row number). | 2026-10-14 | |
vaughn_2025_police_weapon_s2 |
Conjoint table (irw_conjoint_2). Factorial vignette, one police-civilian encounter per respondent (US Lucid, 2021): officer race x demeanor of officer and civilian (4 pairs) x civilian weapon (unarmed or baseball bat) x punishment (ticketed or arrested). Nine 1-5 ratings (threat to victim, police and civilian; expected, preferred punishment; fairness before, of the punishment to civilian and victim, overall), paraphrased wording; the early items precede the punishment attribute. The vignette text is not deposited; undocumented W* columns not kept; no respondent id. | 2026-10-14 | |
vecchiato_2025_candidates_text |
Conjoint table (irw_conjoint). choice: “Which of these politicians would you rather have as your Representative in Congress?”, forced choice between two text profiles, no opt-out. Restricted randomization: some professions require higher education (Doctor, Lawyer, Teacher about half as frequent). cov_survey_weight is the YouGov weight; the same 500 respondents also answered the visual version (vecchiato_2025_candidates_visual). The article says 5 tasks per respondent; the data have 10. | 2026-10-07 | |
vecchiato_2025_candidates_visual |
Conjoint table (irw_conjoint). choice: “Which of these politicians would you rather have as your Representative in Congress?”, forced choice between two mock Twitter profiles, no opt-out; 8 skipped tasks omitted. Attribute levels are the authors’ labels decoded from the image file names (race shown only in the photo, feedback as follower/like counts, religion ‘No Religion’ = none in the bio); the two partisan tweet variants per profile are not stored. Restricted randomization (professions by education). The authors’ own processed file codes profession differently for 919 profiles; this table follows the images. | 2026-10-07 | |
vegetti_2021_intself |
Block name is the deposit’s code; the construct is undocumented in the deposit. | 2026-10-10 | |
ventura_2024_security_candidates |
Conjoint table (irw_conjoint). choice: which of two candidates (Candidato A / Candidato B) the respondent would vote for (paraphrase; wording not deposited), forced choice, no opt-out. Levels in Spanish as displayed; attribute order randomized per respondent (attrpos_). Tasks with no answer (-999; 48 and 57) dropped. Platform identifiers, IP and coordinates dropped. | 2026-10-09 | |
vera_2025_anticorruption |
Conjoint table (irw_conjoint). choice = vote (“which would you vote for”), with an opt-out: 3,424 of 10,290 tasks are ‘would not vote’ (0 on both profiles). choice_effective = “which will help reduce corruption the most” (forced). Task/profile inferred from row order. Six attributes carry the author’s English labels; ideology may have been shown relative to the respondent. 2,058 vs 2,060 in the paper (2 corrupted respondents dropped, as in the authors’ models). | 2026-10-07 | |
vik_2025_unelected_representation |
Conjoint table (irw_conjoint). choice = which of two activist profiles made the respondent feel more represented (forced, no opt-out); rating = how represented the respondent felt by each profile, 1-7, higher = more represented (article wording paraphrased; the article calls it both 7-point and six-point, data run 1-7). Attribute values are the authors’ codes for photo/name and tweet manipulations, not displayed text. Respondents got one of 100 pre-drawn design versions; pooled level shares are uniform but shares within a task/profile slot are unequal and attributes are not independent. Germany, Italy, Romania and Sweden pooled (cov_country); N = 8,279 as in the article. | 2026-10-07 | |
vilaca_2024_corruption_crusaders |
Conjoint table (irw_conjoint). Single-profile image factorial (3 leaflets per respondent): rating = ‘Qual a chance de você apoiar esse(a) candidato(a)?’ 1 = Extremamente provável to 5 = Extremamente improvável (lower = more support, raw); rating_security/corruption/unemployment = perceived effectiveness, 1 = Nada efetivo to 5 = Extremamente efetivo. Slogan stored as the authors’ category with an unmapped variant number; occupation, gender and race are the authors’ codings of costume/photo. 1,116 respondents with a rating (authors’ sample 1,010 flagged, article reports 1,052); IP, GPS, postal code, city and free text dropped. | 2026-10-09 | |
villamizar_2024_afro_districts |
Conjoint table (irw_conjoint). Outcome: forced choice of the district the respondent would prefer to vote in (paraphrase; verbatim Spanish question not deposited), no opt-out. Task and profile are inferred from row order. Attribute text is the authors’ short English labels (display language not documented, presumably Spanish); the deposit’s Networks levels do not match the article’s Table 1, so labels were not mapped. 1,200 respondents as in the article. | 2026-10-07 | |
visconti_2021_chile_candidates |
“Who would you select for president?” (candidate 1/2), asked face to face. An undocumented code 3 (222 tasks) is kept as choice 0 on both profiles, an inferred volunteered “neither”, as the authors code it. DK/DA tasks dropped, removing 10 respondents (300 -> 290); unlabelled survey items dropped. | #2887 | 2026-10-07 |
visconti_2022_flood_candidates |
Conjoint table (irw_conjoint). choice = candidate selected (source ‘selected’; question wording not deposited); 104 rows coded 999 (missing) dropped; 161 pairs with neither candidate selected kept as recorded (opt-out unknown). cov_failed_conjoint marks the 10 respondents the author drops; cov_flood_exposed = the author’s treatment (damage level > 2). Spanish level text as in the data labels. | 2026-10-07 | |
vogel_2024_citizen_attributes_s2 |
1,097 mean-imputed fit cells (fractional values equal to the item mean) dropped before shipping; rt is the deposit’s per-item response time in seconds. | 2026-10-11 | |
vogel_2024_citizen_attributes_s3 |
706 mean-imputed fit cells (fractional values equal to the item mean) dropped before shipping; rt is the deposit’s per-item response time in seconds. | 2026-10-11 | |
vossler_2025_farmworker_projects |
Conjoint table (irw_conjoint). Incentivized referendum DCE: ‘Which option do you vote for?’ with a fixed ‘No project’ option (opt-out, all profiles 0). Three randomized formats in trial_mechanism: single binary (one project vs none, 600), sequential binary (9 questions, 300), sequential trinary (Project A vs Project B vs none, 9 questions, 300). Projects are 9 fixed bundles (fixed design); cost $5/$8/$11/$15 random. task = display order; trial_question = question number. Prolific IDs dropped. | 2026-10-08 | |
vuk_2026_participation_legitimacy |
Conjoint table (irw_conjoint_2). Factorial vignettes: each respondent rated two vignettes of a participatory decision process, varying participation type (eparticipation, minipublic), selection mode (selfselection, sortition) and policy topic; the control vignette has no process, coded ‘(not shown)’. Fairness and acceptance are 1-10 ratings (higher = more; wording and anchors not in the deposit); no opt-out. Attribute level text is the short tags stored in the deposit (the vignette wording was not available). Spain and Switzerland are pooled (cov_country); task order is inferred from row order. | 2026-10-14 | |
wagner_2025_provider_choice |
Conjoint table (irw_conjoint). Discrete choice experiment: choice = which of two providers the caretaker would choose (forced choice, no opt-out; wording not deposited). Fixed blocked design in three versions, each showing kindness, quality, name type and one of medicines/fee/time; attributes outside a version are ‘(not shown)’. Name is stored as the authors’ religious/caste category of the name. N = 1,189 as in the article, but the deposit has 4 choice sets per household where the article says 8. | 2026-10-09 | |
wang_2026_green_space_heat |
Conjoint table (irw_conjoint). DCE: choice between two park-visit profiles or a no-visit option (paraphrase; choice = 0 on both profiles when no visit was chosen, 31% of sets); fixed blocked D-efficient design, 6 versions x 8 sets. Attribute levels are the article’s quantities (°C, %, present/absent, minutes); question and Japanese wording not deposited. All 1,571 respondents kept; the authors’ exclusion flags (meaning undocumented) are cov_reject_1..3, and their analysed N (1,325 in the article, 1,365 or 1,003 in the code) differs. | 2026-10-09 | |
wang_2026_moral_elevation |
Study 2 orders the 15 moral-elevation items differently from Study 1; its columns were mapped onto Study 1’s item codes by exact variable-label match (one-to-one, same value labels). ids are s1_/s2_ prefixed. | 2026-10-09 | |
wang_2026_psm |
Same 18 items and value labels in both studies (checked); pooled with cov_study and s1_/s2_ id prefixes. treat is the Study 1 recall condition. | 2026-10-09 | |
ward_2025_korean_candidates |
Conjoint table (irw_conjoint). choice: which of two candidates the respondent supports the most (paraphrase), forced choice, no opt-out, 10 tasks. Attributes randomized without constraints (working paper). The general/presidential election framing arm is not deposited. 22 respondents whose choices could not be matched were dropped (1,983 of 2,005; paper n = 2,006). Labels are the authors’ English translations of Korean text. | 2026-10-08 | |
weatherspoon_2015_family_physicians_effectiveness |
Three respondents (ids 291, 493 and 522) are left out. The item offers only Yes (1), No (2) and Don’t know, but the deposit gives these three 25 answers of 3 to 5, and 522’s remaining three 2s cannot be read as No among them. The paper’s Table 4 also leaves out the 3 to 5 answers. Their frequency-of-use answers stay in weatherspoon_2015_family_physicians_freq. | #2196 | 2026-09-30 |
wehr_2026_housing_densification |
Conjoint table (irw_conjoint). choice = ‘Which of the two proposals would you prefer?’ (forced, no opt-out); rating_accept = whether the respondent accepts each proposal, 1 = yes, 0 = no (wording paraphrased). English level labels stored though most respondents answered in German, French or Italian; densification is the % increase in residents and cost rent the % of units at cost rent; postal codes and free text dropped; 3,500 respondents vs 3,497 in the article. | 2026-10-09 | |
weissmuller_2026_legitimation_de |
Conjoint table (irw_conjoint). Paired text-vignette choice experiment, fixed design of 18 choice sets (5 per respondent, trial_choiceset): choice = which scenario’s group action seems more appropriate/legitimate, forced, no opt-out; rating_responsible, rating_accountable, rating_punish = 1 (strongly disagree) - 5 (strongly agree), asked about the chosen scenario only. Non-chosen levels rebuilt from the authors’ design listing; task order not recorded. | 2026-10-09 | |
weissmuller_2026_legitimation_uk |
Conjoint table (irw_conjoint). Paired text-vignette choice experiment, fixed design of 18 choice sets (5 per respondent, trial_choiceset): choice = which scenario’s group action seems more appropriate/legitimate, forced, no opt-out; rating_responsible, rating_accountable, rating_punish = 1 (strongly disagree) - 5 (strongly agree), asked about the chosen scenario only. Non-chosen levels rebuilt from the authors’ design listing; task order not recorded. | 2026-10-09 | |
weisstanner_2025_symbolic_class |
Conjoint table (irw_conjoint_2). Paired candidates for the leadership of the respondent’s own party (cov_conjoint_party); choice = ‘If you had to decide, which party president would you prefer?’ (Don’t know = NA on both profiles, 1,344 of 7,700 tasks), rating = ‘How likely is it that you will vote for the PARTY in the upcoming national elections if candidate [1/2] becomes party president?’ 0-100 (Don’t know = NA). Attribute text in German from the published questionnaire with placeholders resolved by the script (Sohn/Tochter, ‘gemässigten’ for SVP). Paper says economic and cultural positions were ‘kept together’ (rule not explained; all combinations occur). N = 1,540 in deposit, 1,506 with an answer (paper 1,550). | 2026-10-09 | |
wenzke_2025_att_sd_a |
Semantic-differential items: the deposit gives the adjective pairs but not the object being rated. | 2026-10-09 | |
wenzke_2025_att_sd_b |
Semantic-differential items: the deposit gives the adjective pairs but not the object being rated. | 2026-10-09 | |
west_2024_representation_men |
Conjoint table (irw_conjoint_2). Single-job vignette ratings, five per respondent: rating = willingness to apply (0-100), rating_qualify = feeling qualified (0-100), no opt-out. Levels are the authors’ labels (the instrument is not deposited); the respondent-level framing arm is trial_treatment; Men replication sample. | 2026-10-11 | |
west_2024_representation_women |
Conjoint table (irw_conjoint_2). Single-job vignette ratings, five per respondent: rating = willingness to apply (0-100), rating_qualify = feeling qualified (0-100), no opt-out. Levels are the authors’ labels (the instrument is not deposited); the respondent-level framing arm is trial_treatment; N matches the paper (4,545 women). | 2026-10-11 | |
wicki_2019_passenger_transport |
Conjoint table (irw_conjoint). choice: forced choice between two policy packages (wording not published); rating: support for each package, 1 (strongly opposed) to 7 (strongly support). Countries pooled (cov_country). Attribute text is the authors’ English labels; the car-requirement levels are stored in miles per gallon for all three countries, though the article lists l/100 km for the non-US versions. ‘No tax revenues’ appears only with ‘No new tax’. Attribute order randomized per respondent but not recorded. | 2026-10-08 | |
wicki_2020_vehicle_emissions |
Conjoint table (irw_conjoint). choice = “If you had to decide: Which of the two proposals would you choose?” (forced choice, no opt-out); rating = “How much do you agree or disagree with proposal 1/2?” on 1-7, higher = more agreement (end labels not documented); both are the article’s English of a German/French/Italian survey. Levels are the English text of the article’s Table 1; level shares are unequal (e.g. year 2045 in 25% of profiles), so level weights are observed; no combination of levels is excluded; attribute order was randomized per respondent but not recorded; frame arm in trial_frame. | 2026-10-07 | |
wiesenthal_2026_energy_sharing |
Conjoint table (irw_conjoint). Discrete choice experiment: in each of 10 tasks, choose the renewable energy community offer (choice = 1) or stay with the current electricity supplier (choice = 0; opt-out, no attributes besides the current cost). Fixed blocked design of 40 sets in 4 blocks. Cost level stored as the relative text without the respondent-specific euro amount shown on screen. | 2026-10-09 | |
wilkins_2018_phq9 |
The PHQ-9 was given in each patient’s first or best language by a nurse who spoke it, and the language was not recorded per patient. Most patients were expatriates of many nationalities, so the table pools several language versions. | #2876 | 2026-10-06 |
williamson_2023_preacher_authority |
Conjoint table (irw_conjoint). choice = ‘Who would you personally trust more as an authority on religious matters?’ and choice_listen = ‘If you could only listen to a sermon by one of these preachers, which one would you choose to listen to?’, both forced choice without opt-out; rating = trust in each preacher, stored 0-4 (Not trust at all to Completely trust) as the authors code it. Nationality ‘home country’ was piped per respondent, so Saudi respondents never see the foreign ‘Saudi Arabia’ level. All 14,466 respondents kept; the paper’s N of 12,025 is the Sunni and Salafi subset (cov_sect). | 2026-10-08 | |
williamson_2025_aid_bilendi2024 |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: region (Africa/Asia/Middle East) x government statement (6: democratic, respects rights, both, authoritarian, violates rights, both) x US security threat (none / terrorist group killed dozens of Americans). rating is agreement with supporting military aid, 1 strongly disagree .. 7 strongly agree (HIGHER = more favourable). 10,338 rows of the source carry no treatment and are dropped; the question stem is not in the deposit. | 2026-10-11 | |
williamson_2025_aid_yougov2017 |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent: region x aid cost x government (democratic/respects rights vs authoritarian/violates rights) x US security threat (none / dozens / hundreds killed). rating is agreement with supporting military aid, 1 strongly agree .. 7 strongly disagree (LOWER = more favourable, raw). The question stem is not in the deposit; the security texts are truncated in the .dta labels and completed by analogy. | 2026-10-11 | |
wimmer_2024_rivalry_immigrants |
“Which person do you think would be a better fit to settle in [country]?”: forced choice over 6 pairs, task and profile inferred from row order. Undocumented restrictions are visible (no 21-year-old doctors; language “None” under-sampled). ResponseIds, timestamps and derived rivalry/chauvinism columns dropped; 45,740 respondents deposited vs 46,549 in the article. | #2887 | 2026-10-07 |
woen_2024_tofu_preferences |
Conjoint table (irw_conjoint_2). Single-profile rating conjoint: each respondent scored 16 fixed tofu-snack cards (variant, flavour, category, portion; Indonesian text) on a 1-5 scale stored raw; the deposit gives no question wording or scale labels. Fixed 2^4 full factorial, task = card number in column order (display order not documented). 105 respondents; the demographic sheet holds personal data and was not used. | 2026-10-12 | |
wolbers_2017_class_contact |
QL_13.2 and QL_13.5 are negatively worded and stored already reversed (their value labels run 1 = klopt precies .. 5 = klopt helemaal niet); the other items run 1 = klopt helemaal niet .. 5 = klopt precies. | 2026-10-09 | |
wolbers_2017_conversation |
Criteria are scored 0-2 and gesprek_22 is the global judgement (0-4). Code 9 (‘too little input to rate’, 109 responses) is dropped as a non-response. Criterion 4, a judgement of whether the test leader affected the conversation, is not shipped. | 2026-10-09 | |
wolbers_2017_speak_climate |
QL_13.15 is negatively worded and stored already reversed (1 = klopt precies .. 5 = klopt helemaal niet); the other items run 1 = klopt helemaal niet .. 5 = klopt precies. | 2026-10-09 | |
wolbers_2017_teach_contact |
QL_13.13 is negatively worded and stored already reversed (1 = klopt precies .. 5 = klopt helemaal niet); the other items run 1 = klopt helemaal niet .. 5 = klopt precies. | 2026-10-09 | |
wolf_2017_study2_video_liking |
resp is stored as answered (1 to 5), not reverse-scored. like2, like4 and like6 are negatively worded; score them as 6 - resp before averaging. Done that way, the six-item mean matches the deposit’s own video_like composite exactly for all 264 respondents. The raw mean correlates -0.26 with that composite. | #2337 | 2026-09-30 |
wood_2022_campaign_finance |
Conjoint table (irw_conjoint). choice: ‘If you had to choose between them, which of these candidates would you vote for?’ (Candidate A or B, no opt-out); rating: trustworthiness of each candidate, 1 (Not at all) to 5 (Extremely). Respondents voted in their own party’s primary (trial_primary; independents assigned at random); Democratic candidates never ‘Strongly opposes’ and Republican candidates never ‘Strongly Supports’ the two policies. Contests with missing or duplicated rows dropped (8,870 kept vs 8,931 reported). | 2026-10-09 | |
wozniakjechorek_2026_remote_work |
Conjoint table (irw_conjoint). choice: which of two hypothetical jobs the respondent picked (wording not deposited), forced choice, one pick per task. 10 pairs, 6 attributes; attribute order, restrictions and level probabilities undocumented. One duplicate survey taker reduced to the first response (627 respondents, as in the article); 2 unanswered tasks dropped; IP, GPS and panel IDs removed. | 2026-10-08 | |
wright_2019_obesity_incentives |
Conjoint table (irw_conjoint). Discrete choice experiment: choice between Reward A and Reward B (“which of the two rewards would motivate you more in a family-based weight management program”; wording varies slightly by task), forced, no opt-out. Fixed blocked fractional-factorial design (30 versions of 10 tasks, trial_version). The payment-structure text named the profile’s own value; it is stored with the placeholder [total value]. cov_survey_weight is the sampling weight. Deposit has 305 respondents, article 304. | 2026-10-08 | |
wu2021_basic_empathy |
The eight negatively worded Basic Empathy Scale items (D1, D3, D4, D5, D10, D14, D15, D16) are stored already reverse-scored, as deposited, so higher resp means more empathy on every item. Their option_text follows the stored coding (resp 1 = Completely agree for those items). | #2198 | 2026-10-01 |
wu_2022_dcep_fairness |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2022_dcep_habit |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2022_dcep_national_identity |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2022_dcep_risk |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2022_dcep_social_influence |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2022_dcep_usage |
Eleven groups of respondents (24 rows) share an identical 18-item response vector; almost all are near-constant 5/6 patterns and mostly non-adjacent, consistent with the scale’s concentration, so all 295 respondents are kept. | 2026-10-05 | |
wu_2024_candidate_race_s1 |
Conjoint table (irw_conjoint). Single-profile vignette. 18 ratings: perceived ideology overall/economic/social (1 Extremely Liberal .. 7 Extremely Conservative), priority of 7 issues (1 low .. 3 high), fairness to 8 groups (1 very unfair .. 7 very fair); answer text coded in scale order. Two non-racial positions plus an affirmative-action position (or none), shown in random order (attrpos_). TANF/minimum wage/reparations inference items dropped; 128 respondents without any outcome dropped. | 2026-10-09 | |
wu_2024_candidate_race_s2 |
Conjoint table (irw_conjoint). Five single-profile vignettes per respondent. rating_vote (1 very unlikely .. 5 very likely; wording not in deposit), rating_ideology (1 Extremely Liberal .. 7 Extremely Conservative), prioritize interests of 8 groups (1 none at all .. 5 a great deal). Race restricted (each of four races once, fifth Black or White); district composition and Biden share drawn without replacement across a respondent’s five profiles. Attention-check failures excluded as pre-registered (1,447). District Hispanic/Asian shares may be swapped relative to the appendix. | 2026-10-09 | |
wu_2024_energy_mix |
Conjoint table (irw_conjoint). choice: “Which of these choices do you prefer?” (Choice 1 / Choice 2), forced, no opt-out. 10 pairs of energy mixes, 3 attributes (monthly bill, percentage of renewables, locally produced renewables), attribute order randomized per respondent (attrpos_). Author’s sample: Lucid completes passing two attention checks (1,862; one answered no task). IP, GPS, ZIP and Lucid IDs removed. | 2026-10-08 | |
wu_2026_scotus_nominees |
Conjoint table (irw_conjoint). choice = “Which nominee do you prefer to fill a vacancy on the Supreme Court?”, prefaced in the treatment arms by “Assuming the Supreme Court is split 4-4” or “… composed of a 6-2 Republican [Democrat] majority” (arm in trial_court); forced choice, no opt-out. Partisanship (Independent 10%) and race (White 50%, Asian American 10%) were drawn with unequal weights. 9,359 conjoint completers; the paper’s n = 9,895 includes 536 who did not finish. | 2026-10-07 | |
wunn_2026_research_competence |
wave 1 = before, wave 2 = after the student lab; the deposit’s two side-by-side blocks were joined on the pseudonymous code, and seven post-test codes with no pre-test row were dropped. | 2026-10-10 | |
wunsch_2025_democratic_backsliding |
Conjoint table (irw_conjoint). choice: “Na ktorego kandydata najprawdopodobniej zaglosowal(a)bys?” (which candidate would you most likely vote for), forced, no opt-out; rating: “Jak oceniasz profil kandydata A/B?” 1 = strongly disapprove .. 7 = strongly approve (codebook’s 1-6 is wrong). Attribute order randomized per task (not recorded); candidate gender weighted 35% female. Includes all 2,910 respondents but one with no attributes; the article’s analytic sample (1,979) drops inattentive respondents and speeders: cov_attention_pass_1 / cov_attention_pass_2 carry the authors’ attention flags; covariates blank for 287 respondents not in the survey file. | 2026-10-07 | |
wysienska_2024_just_pay |
Conjoint table (irw_conjoint_2). Rating = amount of earnings (PLN) the respondent thinks is just for a hypothetical worker described by occupation, gender, age, children and job experience; continuous, higher = more pay; no opt-out. Task = row order within respondent (inferred); 1,964 respondents in the file vs 2,001 recruited. | 2026-10-11 | |
xu_2018_authentic_leadership |
Constructs are read from the Chinese subscale headers (no codebook or paper); item codes are reversible transliterations of those headers. 12 rows that repeat another row on all 63 columns (runs of nearby rows from one school, not straight-liners) were dropped as double entries; applies to all four xu_2018 tables. In xu_2018_emotional_exhaustion, exhaustion_10_rev is stored reverse-scored. | 2026-10-07 | |
xu_2026_substitutability |
Observed 1-5 on a 1-7 scale. | 2026-10-10 | |
yamagishi_2016_pd |
The 12 decisions were made in one fixed order (simultaneous, then first mover, then second mover), with the stake rising 300, 800, 1500 within each role, and the second-mover decisions were elicited by the strategy method rather than in response to a real partner’s move. | 2026-10-04 | |
yamaguchi_2024_lottocracy |
Conjoint table (irw_conjoint_2). Factorial vignette, one per respondent, 4 x 2 between subjects: decision model (electoral democracy, lottocracy full, lottocracy upper only, epistocracy) x policy (education, environmental tax). rating is acceptance of the decision, stored raw 1-5 with 1 = most accepting (inferred from the authors’ reversal); level names are the authors’ labels, the vignette text and item wording are not in the deposit. | 2026-10-11 | |
yamao_2025_armed_iraq |
Conjoint table (irw_conjoint_2). Five single-profile tasks per respondent (Iraq, online panel, Jan 2025); a profile is a hypothetical armed non-state actor with four attributes (order randomized, attrpos_ recorded); rating = integer 1-5 stored raw. The deposit has no questionnaire, so the question wording, scale anchors and direction are unknown. Rows with missing rating dropped (some respondents lost entirely); trial_period flags before/after the 17 Jan 2025 ceasefire. | 2026-10-14 | |
yamao_2025_armed_lebanon |
Conjoint table (irw_conjoint_2). Five single-profile tasks per respondent (Lebanon, online panel, Jan 2025); a profile is a hypothetical armed non-state actor with four attributes (order randomized, attrpos_ recorded); rating = integer 1-5 stored raw. The deposit has no questionnaire, so the question wording, scale anchors and direction are unknown. Rows with missing rating dropped (some respondents lost entirely); trial_period flags before/after the 17 Jan 2025 ceasefire. | 2026-10-14 | |
yamao_2025_armed_palestine |
Conjoint table (irw_conjoint_2). Five single-profile tasks per respondent (Palestine, online panel, Jan 2025); a profile is a hypothetical armed non-state actor with four attributes (order randomized, attrpos_ recorded); rating = integer 1-5 stored raw. The deposit has no questionnaire, so the question wording, scale anchors and direction are unknown. Rows with missing rating dropped (some respondents lost entirely); trial_period flags before/after the 17 Jan 2025 ceasefire. | 2026-10-14 | |
yamao_2025_armed_yemen |
Conjoint table (irw_conjoint_2). Five single-profile tasks per respondent (Yemen, online panel, Jan 2025); a profile is a hypothetical armed non-state actor with four attributes (order randomized, attrpos_ recorded); rating = integer 1-5 stored raw. The deposit has no questionnaire, so the question wording, scale anchors and direction are unknown. Rows with missing rating dropped (some respondents lost entirely); trial_period flags before/after the 17 Jan 2025 ceasefire. | 2026-10-14 | |
yamao_2025_iraq_protests |
Conjoint table (irw_conjoint_2). Five forced-choice tasks of two hypothetical protests (Iraq, online panel); attributes place, protesters’ sect, reason, timing, organizer; choice = protest chosen, one per task, no opt-out. Question wording is not in the deposit; attribute levels are the authors’ English tags (LABELS: tags only). Six malformed tasks dropped (916 of 917 respondents kept). | 2026-10-14 | |
yamao_2025_iraq_psl_amendment |
Conjoint table (irw_conjoint_2). Five single-profile tasks per respondent (Iraq, online panel, Arabic): a hypothetical Personal Status Law amendment described by four attributes; rating = integer 1-5 stored raw (the deposit has no questionnaire, so wording, anchors and direction are unknown). Attribute text is the displayed Arabic text from the raw export; 82 rows with a missing rating dropped (1,052 to 1,044 respondents). | 2026-10-14 | |
yan_2023_job_stress |
The deposit labels this block only ‘OS1-OS4’ and ships no codebook; it is identified as the paper’s job-stress measure from the abstract and the item count/range, not from a codebook. | 2026-10-06 | |
yang_2023_emotional_eating_cesd |
The item codes do not follow canonical CES-D numbering for the first two items: CESD_1 is CES-D item 2 (appetite) and CESD_2 is CES-D item 1. CESD_3 onward match their canonical numbers. The survey export put the appetite item in a column of its own ahead of the block. The responses are correct; only those two labels are out of order. | #2218 | 2026-09-30 |
yang_2023_emotional_eating_ders |
DERS_1 is a real item, but in the deposit its column header is the block’s instructions, not the item’s wording, so which DERS item it is cannot be read from the source. Its responses are genuine. | #2218 | 2026-09-30 |
yang_2023_emotional_eating_eesr |
EESR_1 is a real item, but the deposit does not say which emotion it asks about: the survey export gave its column the question stem (“When you have the following emotions, the degree of desire to eat is”) instead of the item’s label. Its responses are genuine, and its item text is blank. | #2218 | 2026-09-30 |
yang_2023_emotional_eating_uppsp |
The deposit’s header for UPPSP_1 is the block’s instructions run together with a label cut off after “I”, so the item’s own wording is not preserved in the source. Its responses are genuine. | #2218 | 2026-09-30 |
yang_2024_charity_regulation |
Conjoint table (irw_conjoint). choice: ‘Of the two charities above, which do you expect the government to adopt the flexible regulation instead of rigid regulation?’ forced choice, no opt-out; rating: ‘To what extent do you support the government to adopt the flexible regulation for Charity A [B]?’ 1-7, higher = more support. Text from the English translation (fielded in Chinese). One task per charity field (trial_field), task number = field. Only Experiment 2 of the questionnaire is deposited. | 2026-10-09 | |
yang_2025_nigeria_tax_health |
Conjoint table (irw_conjoint). choice: ‘Which of these two policies do you prefer?’ (Policy 1 / Policy 2), forced, no opt-out; rating: ‘How much would you be satisfied or dissatisfied with policy 1/2’, 1 Very dissatisfied to 5 Very satisfied, don’t know/refused (106) set to NA. Attribute order randomized per policy and task (attrpos_). Choice taken from the .dta because the deposit’s long csv choice column is misaligned. Survey weight kept. | 2026-10-09 | |
yang_2026_candidate_posts |
Conjoint table (irw_conjoint_2). Paired design: one woman and one man shown as simulated Facebook posts, with the post type (Campaign, Personal Viewpoints, Personal Lives) assigned by one of five fixed pair types (control: both Campaign). choice is the forced vote between the two (no opt-out), rating the 0-10 impression of each candidate (higher = more favourable). Profile order is the authors’ option 1 and 2 of the vote question; cov_gender_code and cov_education_code keep unlabelled codes. 1,322 respondents in the file against 1,332 in the Readme. | 2026-10-14 | |
ylisalo_2025_pledges_finland |
Conjoint table (irw_conjoint). Factorial vignette experiment: 6 vignettes per respondent, one per pledge, each rated 0-10 (how acceptable the non-fulfillment of the promise is, 0 = not at all acceptable, 10 = fully acceptable). Level text is the English codebook labels. 2,466 in the file (paper code 2466); 44 never reached the vignettes and 55 more gave no rating, so 2,367 remain; the paper’s figures keep Finnish citizens only (cov_citizenship). | 2026-10-08 | |
ylisalo_2025_pledges_germany |
Conjoint table (irw_conjoint). Factorial vignette experiment: 6 vignettes per respondent, one per pledge, each rated 0-10 (how acceptable the non-fulfillment of the promise is, 0 = not at all acceptable, 10 = fully acceptable). Level text is the English codebook labels. All 1,359 respondents (paper code 1359). | 2026-10-08 | |
ylisalo_2025_pledges_germany_ptv |
Conjoint table (irw_conjoint). Factorial vignette experiment: 6 vignettes per respondent, one per pledge, each rated 0-10 (propensity to vote for the pledging party, 0 = very unlikely, 10 = very likely). Level text is the English codebook labels. 4,721 rows in the file; 2,093 respondents with vignettes and ratings remain; attention test kept as cov_attention_pass. | 2026-10-08 | |
yoh_2024_journal_choice |
Conjoint table (irw_conjoint). Discrete choice experiment: choice among three hypothetical journals (Card preference per choice; question wording not deposited) with an opt-out ‘Would not choose any of these journals’: 3,510 of 12,365 sets are opt-outs with choice = 0 on all three profiles. Fixed D-efficient design (12 cards, no blocks); task is the card number, not the order shown. 91 respondent-card sets with no answer dropped; N = 1,038 and 12,365 sets match the paper. | 2026-10-09 | |
yuan_2026_vr_chinese |
Pre/post quasi-experiment: wave 1 = pre-test, wave 2 = post-test; treat 1 = VR-taught group. Six three-item blocks (CCE, PA, VOR, EM, SCS, CCA as deposited) shipped as one table. | 2026-10-11 | |
zabelina_2022_caq |
Checkbox instrument. resp follows the authors’ published scoring: 0 = not checked, 1-6 = frequency category (1 time … 20+ times) for items with frequency options, 0/1 otherwise. A domain where the respondent checked nothing at all is missing (dropped), not zero, as the scoring instructions specify. Item 1 of every domain is ‘I don’t have much talent in this area’. | 2026-10-09 | |
zangger_2020_childcare |
Conjoint table (irw_conjoint_2). Discrete choice experiment, one task per respondent: ‘Which option would you recommend your friend?’ among a nursery, a child minder and a private arrangement; the private arrangement has no attributes and is the opt-out (choice 0 on both profiles, 27 of 486 tasks). Profile 1 is always the nursery and profile 2 the child minder, with 5 attributes whose price levels differ by type (nursery 55/85/125, child minder 35/60/85 Swiss Francs). Four scenario factors are trial_ columns. N = 486 of the 540 respondents in the file (the paper reports 540 parents). | 2026-10-11 | |
zeng_2026_foreign_social_media |
Conjoint table (irw_conjoint_2). Paired conjoint of hypothetical foreign social media apps (initial wave, US adults); choice = which app the respondent would most support the U.S. government banning (forced, no opt-out); rating = support for banning each app, 1 Strongly oppose to 5 Strongly support, stored raw. Attribute order is randomized per task (attrpos_); task 7 repeats task 1 with the two profiles swapped. ‘Owned by the government’ never appears with ‘Highly democratic’; level shares are programmed non-uniform. Task 5 is absent for 50 respondents (attribute strings empty). 1,494 respondents; article N not checked. | 2026-10-12 | |
zhang_2020_contact_tracing_app |
Conjoint table (irw_conjoint). One app article per respondent (task = profile = 1), 6 randomized features (expiry sentence absent for a third = (not shown)). Ratings: likelihood to download and to report a positive test (0-100, smartphone owners only), guessed local uptake (0-100), nine consequences (1 Yes/0 No), confidence data protected (0-3), five agree/disagree statements (-2..2); refusals and Don’t know -> NA, no imputation. 646 respondents who never reached the block and 22 with no answer dropped; no deposited weight. | 2026-10-08 | |
zhang_2024_bureaucrat_cooperation |
Conjoint table (irw_conjoint). Three scenarios (trial_scenario), one paired task each: choice = which officer/counter the respondent would rather cooperate with/go to (forced, no opt-out); rating = willingness to cooperate 1-10 and rating_protect = belief the official would protect one’s interests 1-10 (higher = more). Level text is the Chinese questionnaire wording; accent code order inferred from the questionnaire; experiment 4 is a separate table (zhang_2024_hotline_operator). | 2026-10-09 | |
zhang_2024_hotline_operator |
Conjoint table (irw_conjoint). One paired task (the 4th scenario of the survey): choice = which hotline operator the respondent would rather keep raising their request with (forced, no opt-out); rating = willingness 1-10, rating_protect = belief the operator would protect one’s interests 1-10. Level text is the Chinese questionnaire wording; accent code order inferred from the questionnaire. | 2026-10-09 | |
zhang_2024_volunteer_projects |
Conjoint table (irw_conjoint). Paired choice between two volunteer projects, 6 tasks per respondent; choice = source response (exactly one profile chosen per task, no opt-out); question wording not in the deposit. Attribute text is the author’s English value labels; the Chinese display text was not deposited. Covariate codes have no codebook and keep their codes; the author’s derived scales and dummies are dropped. No linked article, so N (658) cannot be checked. |
2026-10-08 | |
zhang_2025_schisto_attitude |
The file has no pupil identifier, so baseline and post-test questionnaires cannot be linked: each questionnaire is its own id and the occasion is cov_timepoint, not wave. | 2026-10-09 | |
zhang_2025_schisto_knowledge |
The file has no pupil identifier, so baseline and post-test questionnaires cannot be linked: each questionnaire is its own id and the occasion is cov_timepoint, not wave. | 2026-10-09 | |
zhang_2025_schisto_practice |
The file has no pupil identifier, so baseline and post-test questionnaires cannot be linked: each questionnaire is its own id and the occasion is cov_timepoint, not wave. | 2026-10-09 | |
zhao_2025_htse |
The paper says items 3, 6 and 10 of the Healthcare Technology Self-Efficacy Scale are reverse-scored, but in the deposit it is HTSE4 and HTSE8 that correlate negatively with the other items. Responses ship as deposited, with nothing recoded. | #2997 | 2026-10-14 |
zhao_2025_ssrs |
Item 5 of the SSRS is deposited only as the sum of its five sub-items (8-20) and is not shipped. Items 6 and 7 are counts of support sources rather than ratings, so the table mixes two response formats. The deposit has one pair of rows identical across all 51 columns; both are kept as separate ids. | #2997 | 2026-10-14 |
zhirkov_2022_immigrant_admission |
Each profile rated 0 (“Definitely not admit”) to 10 (“Definitely admit”); no choice question. Attributes fully randomized, attribute row order randomized per respondent (kept in attrpos_). 7 missing ratings omitted; N = 929 matches the paper. | #2887 | 2026-10-07 |
zhirkov_2025_illegal_undocumented |
Conjoint table (irw_conjoint). choice = the profile the respondent identified as the “illegal” (trial_condition = illegal) or “undocumented” (trial_condition = undocumented) immigrant, a perception judgement, not a preference; verbatim instruction not deposited. No opt-out documented; 40 of 6,498 tasks with no chosen profile were dropped, removing 3 of 1,083 respondents. Police record and benefits levels were not equally likely (about half No record / None; 2 labelled police-record levels never occur). Attribute text is the authors’ value labels. | 2026-10-07 | |
zhirkov_2025_immigrant_illegality |
Conjoint table (irw_conjoint). One profile per task: respondents rated each hypothetical immigrant by the likelihood of belonging to an “illegal/undocumented immigrant” group, 0 = Extremely unlikely to 10 = Extremely likely (higher = judged more likely undocumented). No choice question. Levels independently randomized, but police record and government benefits are no/yes with probability 1/2 each, the yes half split equally across specific crimes (3) and programmes (4). The article reports 935 respondents (after dropping 79 straight-liners); 7 of them have no conjoint rows in the deposit, so the table has 928. | 2026-10-07 | |
zhirkov_2025_immigrant_regions |
Conjoint table (irw_conjoint). One immigrant profile per task; respondents guessed the immigrant’s world region (Africa, Asia, Europe, Latin America or the Middle East). The single five-option answer is stored as rating_africa .. rating_middle_east, 1 = that region was chosen (exactly one per row). No choice among profiles. Benefits and police record are ‘none’ with probability 1/2, otherwise one of several specific labels; other attributes uniform and independent. N = 1,979 as in the article. | 2026-10-08 | |
zhirkov_2025_welfare_stereotypes |
Conjoint table (irw_conjoint). rating is how typical the profiled person is of welfare recipients, 0-10 (higher = more typical; a stereotype-content rating, not a preference), one profile per task, 30 tasks; no choice question. Attribute order randomized once per respondent (attrpos_*); levels independently drawn. The table keeps all 1,964 respondents; the article analyses 1,271 non-Hispanic white respondents (cov_ethnicity == 1 and cov_hispanic == 1) with non-constant ratings. 72 missing ratings omitted. | 2026-10-07 | |
zhou_2024_eq5d5l_parent |
The parent’s own health; id is the child’s code (one parent per child). | 2026-10-09 | |
zhou_2024_eq_hwb_s |
The parent’s own wellbeing; id is the child’s code (one parent per child). | 2026-10-09 | |
zhou_2026_zainichi_hiring_s1 |
Conjoint table (irw_conjoint). choice = which of two applicants should be hired (forced, no opt-out). Each respondent did a private-company set and a public-school set (trial_job), in random order that is not recorded; both sets show the same 7 applicant pairs (trial_pair), and the first pair of each set is repeated at its end (trial_repeat_of, profiles not swapped). Names never repeat within a pair. Attribute text is Japanese as displayed; names carry ethnicity and gender (see crosswalk). Platform IDs and free-text answers dropped. | 2026-10-08 | |
zhou_2026_zainichi_hiring_s2 |
Conjoint table (irw_conjoint). choice = which of two applicants should be hired (forced, no opt-out). Each respondent did a private-company set and a public-school set (trial_job), in random order that is not recorded; both sets show the same 7 applicant pairs (trial_pair), and the first pair of each set is repeated at its end (trial_repeat_of, profiles not swapped). Names never repeat within a pair. Attribute text is Japanese as displayed; names carry ethnicity and gender (see crosswalk). Platform IDs and free-text answers dropped. | 2026-10-08 | |
zhu_2023_corruption_competence |
Conjoint table (irw_conjoint_2). One vignette per respondent about a local official, 2x2x2 between subjects (integrity corrupt/honest, competence, government level township leader/municipal mayor); level text is paraphrased from the authors’ do file and appendix tables because the vignette wording is not deposited. Three yes/no questions about the same official (support demoting, promoting, reporting corruption), stored as choice_* with opt_out = yes. No respondent id: id is the row number of 5,527 respondents. | 2026-10-14 | |
zia_2026_ai_decision_support |
Conjoint table (irw_conjoint). Discrete choice experiment: “Which system would you prefer (Please choose only one option)?” among AI-DSS 1-3 or “None of these”; the opt-out is not a profile, so choice is 0 on all three profiles in 995 of 3,823 tasks. Fixed design of 5 choice sets shown to everyone, with collinear levels; task is the choice-set number, not the order shown (randomized, not recorded). 20 respondents answered fewer than 5 sets. Platform IDs re-keyed; free-text specialization dropped. | 2026-10-08 | |
ziller_2021_state_surveillance |
Conjoint table (irw_conjoint_2). One vignette per respondent (3 x 3 x 2: policy scope, data security, terror salience; a ‘not mentioned’ level is stored as ‘(not shown)’); rating = support for the surveillance measure and rating_trust_gov = trust in government, both 11 points coded 1-11 (wording and direction not in the deposit). Four countries pooled (cov_country). Level text is the authors’ short labels. N = 6,168 respondents (6,169 in the file, one with both outcomes missing). | 2026-10-11 | |
zou_2020_ev_purchase |
Conjoint table (irw_conjoint). Discrete choice experiment: choice between a gasoline car (profile 1, fixed price 0.85 x budget, 2 per 100 miles, 400 miles; charging rows not shown) and an electric version (profile 2), forced, no opt-out; question wording paraphrased. Attribute text rebuilt from the deposit’s numbers with the templates of the authors’ screenshot; the text of 0-minute, 60-minute and not-available levels is unverified. Task order is row order (display order not recorded). New- and used-car arms in trial_car_type. Home ZIP codes dropped. | 2026-10-08 |