A meta-analysis on bias in technology-mediated interviews

The present meta-analysis examines the occurrence and magnitude of bias in technology-mediated employment interviews, a selection format that has become increasingly common since the COVID-19 pandemic. While employment interviews are widely used and can yield valid hiring decisions when structured, they remain vulnerable to stereotyping and bias based on job-irrelevant applicant characteristics. Technology-mediated interviews (e.g., telephone interviews, videoconference interviews, asynchronous video interviews) may introduce additional biasing cues, such as reduced interpersonal information or contextual visual signals, potentially amplifying biased evaluations. This meta-analysis synthesizes empirical evidence on whether applicants with socially stigmatized characteristics (e.g., gender, age, race, disability, accent, physical appearance, socioeconomic status) receive systematically different interview evaluations compared to non-stigmatized applicants. We test whether bias effects are present overall (H1) and whether they are reduced under conditions of greater interview structure, including higher standardization (H2), greater job-relatedness of questions (H3), and a larger number of interview questions (H4). In addition, we examine whether bias is stronger for stigmatized attributes perceived as less controllable or changeable (RQ1), and whether technology-specific cues (e.g., video backgrounds) influence judgments differently than cues common across interview formats (RQ2). Finally, we investigate whether the simultaneous presence of multiple stigmatized characteristics leads to stronger negative evaluation effects (H5). By quantitatively integrating findings across diverse bias types and interview formats, this meta-analysis aims to clarify the extent to which technology-mediated interviews compromise fairness in personnel selection and to identify conditions under which biased evaluations are minimized.