Sixteenth International Conference on The Image

Abstract

Correspondence studies have been used widely in the social sciences to examine discriminatory behavior, including bias in hiring practices based on applicants’ perceived race/ethnicity or gender. In labor research, correspondence studies can be used to estimate the causal effect of a specific applicant characteristic on the probability of being contacted for a job opportunity. To generate valid results, the “correspondence” (i.e., CVs) for fictional job applicants should be credible. If, for example, the educational experience indicated on a particular CV is incommensurate with what would be expected for a specific job posting, the application may be discarded by a prospective employer. Few studies have examined the validity of the application materials generated in correspondence studies. We are conducting an experimental study to test the validity of the CVs generated for fictional applicants. We recruited participants from Bengaluru, India, one of the planned sites for a future correspondence study. We developed CVs for fictional applicants and collected corresponding CVs from online repositories representing real applicants. Each participant will receive 10 CVs, half real and half generated, matched to specific job postings. Applicants will be asked to complete a structured questionnaire for each CV, including information about the quality of the candidate and if the CV represents a real or fictitious applicant. Our analyses will assess if participants can systematically distinguish between real and fictitious CVs. This methodology can be adapted to examine the validity of materials used for conducing correspondence studies in the social sciences more generally.