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.

15 notes covering 15 tables. Start typing to search.

Table Note Issue Added
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
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 2026-09-28
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 2026-09-28
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
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
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
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