Witnessing robot mistreatment: A dual-path model of behavioral contagion and empathy in customer observers

Abstract

This study investigates how focal customers respond when witnessing another customer mistreat a service robot, extending research that has primarily focused on direct customer–robot interactions. Drawing on social learning theory and moral emotions theory, we propose a dual-path model in which observed mistreatment simultaneously normalizes incivility and elicits empathic concern, each shaping distinct forms of behavioral intention. Two experimental studies using video-based customer–robot interactions tested this framework. Study 1 (N = 200) examined the effects of observed mistreatment on perceived permissibility and perceived victimhood, and their respective roles in predicting incivility intentions and prosocial behavior intentions. The latter pathway was tested through sequential mediation via perceived victimhood and empathy. Study 2 (N = 320) tested the moderating effects of robot anthropomorphism (manipulated) and moral identity (measured) using a 2 (Observed Behavior: Mistreatment vs. Neutral) × 2 (Robot Anthropomorphism: High vs. Low) between-subjects design. Observed mistreatment activated two distinct mechanisms: the contagion pathway, mediated by perceived permissibility, increased incivility intentions; the empathy pathway, sequentially mediated by perceived victimhood and empathy, increased prosocial behavior intentions. Robot anthropomorphism strengthened empathic responses and reduced incivility, while moral identity weakened contagion effects and amplified moral concern. These findings demonstrate that focal customers’ moral and normative appraisals of robot mistreatment can lead to both deviant and prosocial behavioral intentions. The study contributes to emerging research on third-party responses in technologically mediated service contexts and offers practical implications for the ethical deployment and design of service robots in customer-facing environments.