The psychology of user reliance on anthropomorphic robots: Affective dissonance, cognitive load, and enactive sense-making.

Abstract

Guided by Enactive Cognition Theory, this study examines how enacted anthropomorphism relates to users' reliance on service robots through affective dissonance and cognitive load. A sequential mixed-method design comprised a pilot phase for construct refinement, a quantitative study of 311 users with prior service-robot experience using structural equation modeling, and a post-hoc qualitative study with 19 hospitality industry professionals. Results show that enacted anthropomorphism is positively associated with reliance and negatively associated with affective dissonance, while unexpectedly increasing cognitive load. Affective dissonance negatively relates to reliance and constitutes a significant indirect pathway between anthropomorphism and reliance. In contrast, although cognitive load negatively relates to reliance, it does not significantly mediate the anthropomorphism-reliance relationship. Qualitative findings contextualize these results by showing that human-like cues can facilitate social connection while also generating discomfort or interpretive demands when expectations exceed robot capabilities. The study extends Enactive Cognition Theory by demonstrating that anthropomorphic human-robot interaction can simultaneously involve social facilitation, emotional coherence, and cognitive effort, positioning reliance as an outcome of ongoing enactive sense-making.