The Good, the Bad, and the Missing: Topic Modeling Analysis of User Feedback on Digital Wellbeing Features

Abstract

Digital wellbeing features could potentially help users mitigate unintended effects of IT use such as smartphone addiction. However, knowledge about users’ perceptions of such features is still scarce. To bridge this gap, we applied structural topic modeling to analyze user reviews of 93 digital wellbeing apps from the Google Play Store. Our findings revealed three promising research areas: (1) mitigation mechanisms associated with self-monitoring, goal advancement, and change UI features, (2) the relationship between restrictiveness of block features, user characteristics, and addiction levels, and (3) the association of gamification with other features to promote behavior change. We also highlight the advantages of using structural topic modeling to analyze a large body of app reviews. Finally, we provide developers of digital wellbeing apps with feature requests extracted from the reviews.