Understanding perceived barriers to treatment from web browsing behavior
- Stephen M. Schueller ,
- Diana M. Steakley-Freeman ,
- David C. Mohr ,
- Elad Yom-Tov
Journal of Affective Disorders | , Vol 267: pp. 63-66
Background
The expanding amount of information available from our use of technologies has led researchers to explore how this information can aid in the detection of mental health issues. We expand on past work in this area by exploring how browsing histories might be able to predict perceived barriers to psychological treatment.
Methods
We obtained 10 days of browsing history data for 255 respondents as well as assessments of Perceived Barriers to Psychological Treatments and depression, the Patient Health Questionnaire.
Results
We found that browsing histories enabled high performance classification of people with high levels of perceived barriers to psychological treatments (AUC average of 0.86).
Limitations
Our high classification accuracy does not help understand why different features within the browsing histories are useful to classify people according to browsing history. We also look at people who decided to contribute their browsing history but the use of this data more generally presents additional ethical questions.
Conclusions
Browsing histories might be useful to classify people’s barriers to seeking psychological treatment. It is clinically relevant to find those who perceive barriers to seeking treatment to better design ways to address those concerns and help them find treatment.