Mandryk RL, Birk MV. (2019). The potential of game-based digital biomarkers for modeling mental health. JMIR Mental Health. 6(4): e13485. doi: 10.2196/13485
In this paper, the authors discuss the potential for data collected from video game use to be used as biomarkers for mental health. Games developed for research as well as commercial video games can provide researchers with different kinds of data, including players’ gaming behavior (e.g. login frequency, play time, scores), interactions with other players (e.g. chat logs from within the game), activity on gaming-focused social media platforms (e.g. chatting on Discord, streams on Twitch.tv), and physiological indicators while playing (e.g. eye movement, heart rate, facial expression). A number of game developers and social media companies provide access to user data through application programming interfaces (APIs), but some data may need to be collected using special equipment (e.g. eye tracking) or software (e.g. mouse movement tracking). Researchers can derive biomarkers for mental health from different domains of game interaction including user behavior (e.g. play time and style patterns, player preferences for games and roles), cognitive performance (e.g. short-term memory, pattern recognition, decision-making), affective and emotional indicators (e.g. vocal tone and language, posture, eye movement), motor input (e.g. interactions with mice, keyboards, controllers, and phone screens), and social interactions (e.g. multiplayer game interactions, in-game verbal utterances over microphone, in-game and third-party chat interfaces, social media interactions). The authors note that more research is needed to connect digital biomarkers derived from gaming behavior to mental health indicators and that there are many ethical and legal issues that need to be considered.