What young people’s device usage reveals about their mental health
Imagine a scenario where your keystroke patterns or the words you type into your device serve as an early warning system for mental health concerns. This could enable preventative or first-response mental health services to intervene sooner. It may sound like a pipe dream, but electronic devices can help address the rising number of young people facing mental health crises.
I recently got a sneak preview of the latest episode of the Psych Attack podcast, Digital phenotyping: Using smartphone metadata to predict mental health symptoms. In this episode, Dr. Jasmine B. MacDonald poses thought-provoking questions to Dr. Taylor A. Braund, delving into valuable insights about young people’s device usage and what it reveals about their mental health.
While there is a lot of negative press about adolescents and social media use, Taylor has given us a strong counter argument, highlighting some of the hidden benefits that can come from our online natives (teenagers) device usage.
This post provides a snapshot of what I have learnt while listening to the episode.
What is a phenotype?
Taylor described a phenotype as akin to a digital signature—a unique collection of data reflecting an individual’s device usage. This data encompasses various aspects, including:
- Linguistic data: the words we type.
- GPS data: our movements in terms of location and timing.
- Gyroscope data: information about the speed and direction of phone movement.
- Camera or screen captures: the content we include in our photos.
- Typing pace and patterns: insights into how quickly we type and our patterns of pauses.
How can digital phenotypes be used in real-world settings?
Taylor provided great examples of how digital phenotypes can be applied in real-world contexts. Some of these applications were surprising and prompted me to re-evaluate my own work in identifying well-being concerns among young people. The insights gained from analysing this data can enhance our understanding of mental health and inform interventions.
Linguistic data (i.e. words)
Taylor referred the audience to research that showed the words we use in our communication could be an early predictor of rumination (one of the symptoms of Major Depressive Disorder) (Edwards & Holtzman, 2017).
Information within still photos or video footage
Taylor highlighted recent research where machine learning has been used to analyse facial movement patterns in camera and video footage to identify individuals with depressive symptoms, potentially aiding in mental health diagnosis (Islam & Bae, 2024; Nepal et al., 2024).
GPS data
A recent review on digital phenotyping for stress, anxiety, and mild depression found that students experiencing these symptoms showed distinct behavioural patterns captured by their devices. These students visited fewer locations, were more sedentary, and had increased phone usage. This correlation provides a new way to capture and understand behaviours (Choi et al., 2024).
Findings from the Future Proofing Study
In The Black Dog Institute’s Future Proofing Study, Taylor’s recent research analysed typing pace and frequency among 943 Australian adolescents. His findings suggest that young people experiencing #depression tend to type less frequently on their smart phones, but when they do use their phones they type faster than other young people.
For young females, slower keystrokes and higher word frequency are associated with mental health symptoms, while males exhibited the opposite pattern. Understanding these patterns of differences gets us one step closer to using phenotypes to support the mental health of young people.
For more details, you can access the study here (Braund et al., 2023) and the full conversation between Taylor and Jasmine is available on all major podcast platforms (audio) and on YouTube (video).
A reflection on research pathways
This episode of Psych Attack also provided insights into the life of a researcher. Taylor gave a brief recount of his journey (including some international travel) that has ultimately led him to work with the Black Dog Institute Future Proofing project team, investigating the current state of adolescent mental health. For me this highlighted the career pathways outside of academic settings for researchers. I am looking for to see where my journey takes me!
Kirstie Northfield
Kirstie Northfield, a researcher at the School of Psychology at Charles Sturt University, is dedicated to identifying who, besides adolescents themselves, can reliably assess the well-being of adolescents. Her research focuses on evaluating the accuracy of parents, teachers, and peers in making well-being support referrals and understanding the characteristics of a reliable judge of adolescent well-being. In addition to her research, Kirstie teaches positive psychology, emphasising resilience and emotional intelligence. To maintain her own emotional balance, she enjoys running short distances, often using this activity to explore her favourite holiday destinations.
References
Braund, T. A., O’Dea, B., Bal, D., Maston, K., Larsen, M., Werner-Seidler, A., Tillman, G., & Christensen, H. (2023). Associations Between Smartphone Keystroke Metadata and Mental Health Symptoms in Adolescents: Findings From the Future Proofing Study. JMIR Mental Health, 10, e44986-e44986. https://doi.org/10.2196/44986
Choi, A., Ooi, A., & Lottridge, D. (2024). Digital Phenotyping for Stress, Anxiety, and Mild Depression: Systematic Literature Review. JMIR Mhealth Uhealth, 12, e40689. https://doi.org/10.2196/40689
Edwards, T. M., & Holtzman, N. S. (2017). A meta-analysis of correlations between depression and first person singular pronoun use. Journal of research in personality, 68, 63-68. https://doi.org/10.1016/j.jrp.2017.02.005
Islam, R., & Bae, S. W. (2024). FacePsy: An Open-Source Affective Mobile Sensing System – Analyzing Facial Behavior and Head Gesture for Depression Detection in Naturalistic Settings. Proceedings of the ACM on human-computer interaction (pp. 1-32). ACM. https://doi.org/10.1145/3676505
Nepal, S., Pillai, A., Wang, W., Griffin, T., Collins, A. C., Heinz, M., Lekkas, D., Mirjafari, S., Nemesure, M., Price, G., Jacobson, N., & Campbell, A. (2024). MoodCapture: Depression Detection using In-the-Wild Smartphone Images. 1-18. https://doi.org/10.1145/3613904.3642680
Photo
Photo by freestocks on Unsplash