Social Robots and Performance Errors: Level of Empathy Distinguishes Changes in Trust

Abstract

Much work has been done to engineer robots’ mechanical capabilities to best suit the general demands of their users and tasks. However, minimal research has addressed the impact of individual differences on perceptions of robot trustworthiness. These conclusions can provide guidance to optimize adaptive robotic systems in education, healthcare, and industry settings. This study examined the relationship between personality and human robot interaction in two contexts: (1) error-free and (2) errors. Assessment of individual differences were achieved via the Interpersonal Reactivity Index (IRI) (Davis, 1980) and robot trust assessed using the Multidimensional Measure of Trust (MDMT) (Ullman & Malle, 2018). This project provided a novel contribution in the field of human-robot interaction, highlighting the influence of technological failure on trust impressions of a social robot. Additionally, we sought to understand the degree to which empathy levels mediate these changes in trust.