19,814 rated or chosen profiles from 9,907 respondents, in up to 1 tasks of 2 profiles varying on 5 attributes.
| Description | US Prolific respondents (n = 9,907) each chose which of two hypothetical immigrants (one college-educated, one with no formal education; profession, origin, reason and prior trips varied) they would admit, with contexts assigned by an adaptive bandit design plus two validation phases. |
|---|---|
| Reference | Gosciak, J., Molitor, D., & Lundberg, I. (2026). Adaptive randomization in conjoint survey experiments. Political Analysis, 1-28. |
| DOI | 10.1017/pan.2026.10040 |
| Licence | CC0 1.0 |
| Source data | https://doi.org/10.7910/DVN/7EBDDY |
These notes describe the source data, which the IRW reproduces as released. All notes are listed on the Data Notes page.
| Respondents | 9,907 |
|---|---|
| Rows (one per respondent, task and profile) | 19,814 |
| Tasks per respondent (most) | 1 |
| Profiles per task | 2 |
| Attributes | 5 |
| Tasks with no profile chosen | 0 |
| Topic | immigration asylum |
|---|---|
| A profile represents | person |
| Country | US |
| Language shown | en |
| Language of stored levels | en |
| Randomization restrictions | yes (stated by the source): Profiles come as 16 fixed pairs (arms): within a pair the 4 context attributes share one value (each profile a different signal of it) and education is always College degree vs No formal education; the college profile always carries the same signals in an arm (metadata files). Arms assigned by Thompson sampling: warm-up ~equal, then adaptive max and min searches, then validation of arm 7 (max) and arm 10 (min) only. |
| Task order | recorded in the deposit |
| Profile position | inferred from row order |
Because the attribute levels were not randomized independently, an estimator has to account for the restriction; a plain difference in means across levels can mislead. Task order or profile position was rebuilt from row order; analyses of position or task order should leave it out.
Outcomes:
| choice | Which of these two immigrant profiles would you select for admission to the United States? [choice, forced] |
|---|
attr_educationattr_originattr_prior_tripsattr_professionattr_reasonchoicecov_agecov_check_correctcov_commitmentcov_garbagecov_gendercov_hispaniccov_raceidprofiletasktrial_armtrial_phase
The layout (one row per respondent, task and profile; choice/rating outcomes; attr_ columns holding each attribute level as displayed) is defined in the IRW conjoint standard. What the outcomes asked is under Design above.
Or load it directly in R or Python (both need a free Redivis account; the first call opens a browser to sign in):
# R
install.packages("remotes")
remotes::install_github("itemresponsewarehouse/Rpkg")
library(irw)
df <- irw_fetch("gosciak_2026_immigrant_admission", source = "conj")
# Python
pip install irw
import irw
df = irw.fetch("gosciak_2026_immigrant_admission", source="conj")
@article{Gosciak_2026, title={Adaptive Randomization in Conjoint Survey Experiments}, volume={34}, ISSN={1476-4989}, url={http://dx.doi.org/10.1017/pan.2026.10040}, DOI={10.1017/pan.2026.10040}, number={4}, journal={Political Analysis}, publisher={Cambridge University Press (CUP)}, author={Gosciak, Jennah and Molitor, Daniel and Lundberg, Ian}, year={2026}, month=Apr, pages={541–568} }
| IRW version | v584 |
|---|---|
| Redivis dataset | irw_conjoint v4.2 |
| Redivis dataset DOI | 10.57761/qkye-q530 |
| Manifest pin for this IRW version | v4.2 |
| Metadata source | irw_meta v44.0 |