Fairness in hire, what talents admire. Scrutinizing fairness perceptions of AI and transparency in asynchronous video interviews.

Sammanfattning: In pursuit of enhancing time- and cost-efficiency, organizations have been adopting asynchronous video interviews (AVIs) integrated with artificial intelligence (AI) to optimize and to automate recruitment processes. The roles of AI decision agents and transparency in shaping job applicants' fairness perceptions in AVIs remain largely unexplored despite the growing prominence of this novel interview format. The aim of this thesis is, therefore, to bridge the research gap by investigating applicants' procedural fairness perceptions in this technology-mediated interview assessment, taking into account the influences of different types of decision makers, the disclosure of decision makers, and the provision of explanations. An online scenario-based between-subject experiment was conducted and Gilliland's justice model was employed to measure applicants' fairness perceptions in the recruitment process. Based on the quantitative analysis of 288 observations, the findings revealed that (1) AVIs rated by humans are perceived procedurally fairer than those rated by AI, (2) transparency regarding decision makers does not significantly influence applicants' perceptions of procedural fairness, and (3) the provision of explanations has a moderating effect on procedural fairness perceptions only when the decision makers are humans. The research yields several important implications for organizations to mitigate applicants' potential negative perceptions and reactions and to facilitate a progressive transition from human to AI delegation in AVIs.