We present a machine learning approach that uses data from smartphones and fitness trackers of 138 college students to identify students that experienced depressive symptoms at the end of the semester and students whose depressive symptoms worsened over the semester. Our novel approach is a feature extraction technique that allows us to select meaningful features indicative of depressive symptoms from longitudinal data. It allows us to detect the presence of post-semester depressive symptoms with an accuracy of 85.7% and change in symptom severity with an accuracy of 85.4%. It also predicts these outcomes with an accuracy of >80%, 11–15 weeks before the end of the semester, allowing ample time for pre-emptive interventions. Our work has significant implications for the detection of health outcomes using longitudinal behavioral data and limited ground truth. By detecting change and predicting symptoms several weeks before their onset, our work also has implications for preventing depression.

Online mental health interventions are increasingly important in providing access to, and supporting the effectiveness of, mental health treatment. While these technologies are effective, user attrition and early disengagement are key challenges. Evidence suggests that integrating a human supporter into such services mitigates these challenges, however, it remains under-studied how supporter involvement benefits client outcomes, and how to maximize such effects. We present our analysis of 234,735 supporter messages to discover how different support strategies correlate with clinical outcomes. We describe our machine learning methods for: (i) clustering supporters based on client outcomes; (ii) extracting and analyzing linguistic features from supporter messages; and (iii) identifying context-specific patterns of support. Our findings indicate that concrete, positive and supportive feedback from supporters that reference social behaviors are strongly associated with better outcomes; and show how their importance varies dependent on different client situations. We discuss design implications for personalized support and supporter interfaces.

Collective intelligence (CI), a group's capacity to perform a wide variety of tasks, is a key factor in successful collaboration. Group composition, particularly diversity and member social perceptiveness, are consistent predictors of CI, but we have limited knowledge about the mechanisms underlying their effects. To address this gap, we examine how physiological synchrony, as an indicator of coordination and rapport, relates to CI in computer-mediated teams, and if synchrony might serve as a mechanism explaining the effect of group composition on CI. We present results from a laboratory experiment where 60 dyads completed the Test of Collective Intelligence (TCI) together online and rated their group satisfaction, while wearing physiological sensors. We find that synchrony in facial expressions (indicative of shared experience) was associated with CI and synchrony in electrodermal activity (indicative of shared arousal) with group satisfaction. Furthermore, various forms of synchrony mediated the effect of member diversity and social perceptiveness on CI and group satisfaction. Our results have important implications for online collaborations and distributed teams.