What's In a Name? 
An Experimental Analysis of Signaling Race, Ethnicity, and Gender Using Names
Matthew Hayes & Elizabeth Mitchell

A growing body of research uses names to cue experimental subjects about race, ethnicity, and gender. Yet, researchers have not explored the myriad of characteristics that might be signaled by these names. We developed two large, publicly available databases of the attributes associated with common American first names. The first provides perceived racial distinctiveness scores for 1,000 names. The second includes ratings of 89 names on 21 additional characteristics, including personal traits as well as nonracial group memberships. We show that the traits associated with first names vary widely, even among names associated with the same race and gender. Researchers using names to signal group memberships are thus likely cuing a number of other attributes as well. We demonstrate the importance of name selection by replicating DeSante (2013). We conclude by outlining two approaches researchers can use to choose names that successfully cue race (and gender) while minimizing potential confounds. 

Our recommendations can be found in our main text, here: bit.ly/3eq8lYf

Tables with the most racially distinctive names can be found in our supplemental information, here: bit.ly/3d5O8a8

Data available here: https://doi.org/10.7910/DVN/HEOSVW

Accuracy of racial perceptions

Differences in perceived traits by race-gender group

Differences on selected traits within race-gender groups