Cognitive and Emotional Consequences of Algorithmically Mediated Video Interviews

Le résumé fourni par la source

Algorithmic evaluation systems are increasingly used in hiring, yet their psychological impact on candidates remains weakly understood. While such systems promise greater consistency and objectivity, emerging research suggests they may also generate stress and discomfort during evaluative interactions. We propose a dual-pathway framework distinguishing candidates’ cognitive appraisals, i.e., perceived uniqueness, from their affective responses, i.e., anxiety, and argue that these pathways are shaped by distinct features of algorithmic evaluation design. Using a 2×2 experiment, we manipulate evaluation structure, i.e., single-agent vs. human–AI dyad, and evaluative agency, i.e., human-primary vs. AI-primary, in a standardized AI-enabled interview context. Results show that evaluation structure primarily influences cognitive appraisals: candidates viewed human–AI dyads as more standardized than single-agent systems, particularly AI-only evaluation. In contrast, evaluative agency shaped affective responses: AI-primary conditions heightened anxiety and discomfort relative to human-primary conditions, even when the interview process remained constant. These findings reveal a core tension in algorithmic hiring design: features that enhance cognitive legitimacy may not mitigate, and may even exacerbate, affective anxiety. Our study contributes to research on algorithmic decision-making by introducing a structural–agential framework for evaluating algorithmic systems and focusing on the trade-offs that hiring organizations must navigate when integrating AI into evaluative roles. We discuss implications for human–AI collaboration, fairness in technologically mediated judgment, and the psychological design of hiring systems.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.