Data & Code: Beyond CVs: A video-based experimental approach to study ethnic evaluation biases in hiring

The published Stata syntax file (do-file) and dataset can be used to replicate the results reported in the cited article. The dataset contains only those questionnaire variables that are required to replicate and verify the analyses presented in the article. Variables concerning respondents’ professional careers or their own ethnic backgrounds, for example, <u>are not included in the dataset </u>. <u>Journal Abstract: </u> Ethnic biases in hiring are most commonly studied through correspondence tests that examine discriminatory behavior and text-based factorial survey experiments that focus on evaluations of applicants. To move beyond written application materials, we explore the potentials and limitations of a complementary approach: a factorial survey experiment in which hypothetical applicants are presented in two formats – written CVs and 90-second video applications. The videos approximate more interactive stages of the hiring process, allowing for the integration of visual and auditory cues, while maintaining comparability between applicants of different ethnic backgrounds (German, Asian, and African). Deepfake technology ensures that all videos feature an identical audio track, characterized by accent-free, fluent, and high-quality German. In our study, managers (n=1,289) with personnel responsibility in German companies evaluate applicants for a marketing position to examine how prejudice level and the availability of applicant information shape hireability assessments. Our findings show that providing additional information through video applications can enhance evaluations of ethnic minority candidates. However, the influence of respondents' level of prejudice does not differ significantly between the two hiring stages. Overall, the proposed approach shows great potential for future research but also faces limitations and concerns, particularly regarding social desirability. The published Stata syntax file (do-file) and dataset can be used to replicate the results reported in the cited article. The dataset contains only those questionnaire variables that are required to replicate and verify the analyses presented in the article. Variables concerning respondents’ professional careers or their own ethnic backgrounds, for example, <u>are not included in the dataset </u>. <u>Journal Abstract: </u> Ethnic biases in hiring are most commonly studied through correspondence tests that examine discriminatory behavior and text-based factorial survey experiments that focus on evaluations of applicants. To move beyond written application materials, we explore the potentials and limitations of a complementary approach: a factorial survey experiment in which hypothetical applicants are presented in two formats – written CVs and 90-second video applications. The videos approximate more interactive stages of the hiring process, allowing for the integration of visual and auditory cues, while maintaining comparability between applicants of different ethnic backgrounds (German, Asian, and African). Deepfake technology ensures that all videos feature an identical audio track, characterized by accent-free, fluent, and high-quality German. In our study, managers (n=1,289) with personnel responsibility in German companies evaluate applicants for a marketing position to examine how prejudice level and the availability of applicant information shape hireability assessments. Our findings show that providing additional information through video applications can enhance evaluations of ethnic minority candidates. However, the influence of respondents' level of prejudice does not differ significantly between the two hiring stages. Overall, the proposed approach shows great potential for future research but also faces limitations and concerns, particularly regarding social desirability.