Abstract
Persuasive mental health applications (apps) are effective tools for promoting behavior change. Tailoring persuasive interventions can boost their effectiveness. Research has shown that there is a relationship between individuals' personality traits and their susceptibility to various features of a persuasive app. The aim of this paper was twofold. First, we explore the relationships between personality and features of a persuasive app for promoting mental and emotional well-being. Second, we explore possible domain-dependent variability on the relation between personality and persuasive features across various domains (e.g., habit formation, fitness, risky behaviors change, transportation habits). Specifically, to advance research in the area of personality and computing app design, first, we reviewed 103 mental health apps from the app stores and uncovered the various persuasive features employed in their design. Second, we conducted focus-group studies of 32 participants to uncover more insights about the mental health app features. Finally, we implemented the common features that emerged from these two studies in persuasive mental health app prototypes and conducted a large-scale study of 561 users to evaluate their persuasiveness depending on people's personalities. The results of the large-scale study show that people's personality traits play a significant role in the perceived persuasiveness of different features-the app features that appeal to various individuals. Conscientious people tend to be motivated by apps that offer relaxation audios, encouragement, suggestions and trusted information, contact for help; Neurotic people are more inclined to apps that provide some relaxation exercises and audios, social support, and apps with a clear privacy policy; People who are more open to experience tend to be motivated by self-monitoring, relaxation exercises and audios, and social support. Finally, we provide a comparative analysis of the relationship between personality traits and persuasive features across various domains to show domain dependency of persuasive feature effectiveness and offer design guidelines for tailoring persuasive and behavior change apps based on an individual's personality.