Trust and AI in Wearable Health Monitoring: Evaluating Usability, Explainability, and Data Visualization

Abstract Advancements in artificial intelligence (AI) and large-scale data analysis have reshaped decision-making in personal health monitoring, enabling wearable devices, such as smartwatches, to provide continuous physiological insights and personalized health recommendations. However, user trust, satisfaction, and acceptance of wearable-generated health insights are critical for their effective adoption. The study conducted a follow-up survey with participants who previously took part in an experiment on data integrity using the Apple Watch as a focal product. This survey study investigates the influence of product usability, data visualization preferences, the impact of contextual information, and transparency in AI-generated recommendations on user confidence in health data and AI recommendations. Results indicate that while users report high satisfaction with the device’s usability, participants preferred a detailed visual summary over the device interface display, emphasizing the importance of clear and effective data visualization. However, the inclusion of contextual information, such as environmental factors, had limited influence on user confidence, suggesting that users prioritize data clarity and accuracy over additional contextual layers. Furthermore, AI-generated recommendations with explanations slightly increased trust and adherence, but the effect was not statistically significant, implying that users may already have an inherent level of trust in AI-driven insights. These findings highlight the need for more user-friendly interfaces and transparent AI explanations to enhance trust and engagement in wearable health technologies. By integrating Explainable AI (XAI) and optimizing data visualization, wearable health systems can enhance user confidence and support informed decision-making in digital health technologies.