A meta-analysis on reactions to algorithmic decision-making in human resource management

In this meta-analysis, we assess reactions to algorithmic decision-making (ADM) in human resource management. Drawing on the computers-are-social-actors (CASA) paradigm, we offer a theoretical framework to explain the relationships between ADM and (1) system-related reactions (e.g., perceptions of justice, trust) as well as between ADM and (2) organization-related reactions (e.g., organizational attractiveness, turnover intention). Additionally, we examine boundary conditions (e.g., type of stakeholder, type of interaction, type of task) of these relationships. Our meta-analysis includes 365 effect sizes from 73 samples in 53 studies ( N = 24,578) and reveals that ADM is negatively associated with system-related (i.e., justice, fairness, trust, trustworthiness) and organization-related (i.e., organizational attractiveness, job pursuit intention) reactions, which depend on boundary conditions (i.e., type of interaction, extent of decision). We provide an empirical and theoretical synthesis of primary studies and outline avenues for future research.