Analyzing and Predicting Self-Regulatory Behaviors in Wake-Up Task Alarm

Abstract

Many popular mobile alarm apps offer wake-up task alarms that do not allow the user to dismiss an alarm until they complete a specific task. Understanding mobile alarm usage and wake-up tasks can provide useful implications on human self-regulatory behaviors and inconvenient interaction design. This study aims to understand user behaviors over the mobile alarm using the wake-up tasks based on a large number of Alarmy app usage logs. Alarmy is the most downloaded alarm clock app from the App Store in 97 countries and provides four different wake-up tasks: (1) touching a button (normal_task), (2) taking a picture (picture_task), (3) shaking the device (shake_task), and (4) solving math problems (math_task). This study collected 42,909,263 Alarmy usage logs about setting and dismissing alarms from 211,273 users who had used the app for more than one month from January to May in 2018. Specifically, the Alarmy log data were analyzed in three perspectives: (1) Extracting individual pattern, (2) Comparing the usage contexts among the wake-up tasks, and (3) Modeling hard wake-up taskers. The study results are well linked with the Fogg Behavior Model (B.J. Fogg, 2009), which explains human behavior by the three components (1) motivation, (2) ability, and (3) prompts. Finally, the results provide the potential of using wake-up tasks for mobile crowdsourcing.