JH John B. Holbein

Tool · Experimental research on race

Pick validated names for a study of racial discrimination.

Researchers studying race and discrimination — in correspondence audits, conjoint experiments, vignettes, and survey experiments — often use names to signal a person's race. But names also signal other things — class, citizenship, education — and those bundled signals can confound a clean estimate of race. This tool helps you choose names that fit the comparison you want, with the trade-offs visible.

The data come from Crabtree, Kim, Gaddis, Holbein, Guage, and Marx (2023) — 600 names evaluated by 4,026 U.S. respondents on perceived race, citizenship, income, and education.

Two questions to ask yourself before picking names. (1) What racial comparison are you testing? (2) Do you want the racial signal to be as strong as possible (and accept that other attributes will come along for the ride), or do you want race held apart from class/citizenship/education? The first choice is fine when bundled treatments are what you want; the second is what you want for a clean causal estimate of race alone.

Your name list

Loading the 600-name dataset…
Data & citation. Crabtree, C., Kim, J. Y., Gaddis, S. M., Holbein, J. B., Guage, C., & Marx, W. W. (2023). Validated names for experimental studies on race and ethnicity. Scientific Data, 10:130.
Repository: github.com/jaeyk/validated_names · R package: validatednamesr.
Please cite the source paper when you use these names in published work.