This study explores the paradoxical impact of fraud alert features on user satisfaction in age-friendly applications and examines how the personality traits Openness and Neuroticism moderate this relationship.
Drawing on 7,516 user reviews of an age-friendly music application collected from April 2021 to November 2024, we employ large language model–based (LLM-based) personality inference to identify users' psychological traits. Ordinary least squares regressions with brand fixed effects are applied to test the direct and moderating effects of fraud alert feature within a self-efficacy theoretical framework.
Results indicate that fraud alert features are significantly associated with lower user satisfaction, contradicting the common belief that security innovations are universally beneficial. Users high in Openness experience less dissatisfaction by viewing alerts as learning opportunities, while those high in Neuroticism show stronger dissatisfaction due to anxiety and perceived competence threats. These personality effects are more pronounced in contexts with fewer likes and lower review activity.
This study is among the first to integrate personality psychology with self-efficacy theory to explain heterogeneous responses to security features in age-friendly software. It introduces a scalable LLM-based personality measurement method and highlights the need for personality-aware, context-sensitive design strategies to balance protection and satisfaction in digital aging solutions.
