“I Feel You” Versus “I Understand You”: How Distinct Empathic Strategies Shape Robot Personality and Trust for Older Adults

Abstract

Socially assistive robots (SARs) and voice-based conversational agents may support older adults, but adoption depends on trust. This study examined whether affective and cognitive empathic dialogue shape perceived robot personality and whether trust reflects personality similarity or complementarity. In a between-subjects online experiment, 341 U.S. adults aged 65+ first rated their own Big Five traits, then viewed animated robot interaction videos in affective empathy, cognitive empathy, or control conditions and rated the robot’s traits and trust. Dialogue strategy significantly affected all five perceived robot traits. Affective dialogue produced the highest Conscientiousness ratings, while both empathy conditions increased the remaining perceived traits relative to control. Trust showed a mixed pattern: similarity in Agreeableness, Conscientiousness, and Openness predicted higher trust, whereas Extraversion dissimilarity predicted higher trust. These findings inform trust-oriented SAR and voice-agent design by linking empathic wording to trait-specific personality attributions in home healthcare contexts for older adults.