CURRENT PROJECTS

EASE

Effortless Assessment of Stressful Experiences

This project is investigating the role of mobile sensing using smart phones in detecting changes in stressful life experiences, using an academic stress paradigm. Data derived from naturalistic phone use include natural language use, facial expression, acoustic voice  and geographic movement.

F.LUX

The Effect of Screen Time on Sleep

This project examines the effects of automated diurnal variation in electronic screen temperature (i.e., automatically filtering blue light on screens in the evening) on sleep quality using a blinded, randomized controlled design.

PRISM

Psychological Risk In Social Media

We are examining automated methods to detect social media accounts (Twitter, Instagram) the are associated with high levels of psychological stress and mental health risk. These accounts are then characterized in terms of their patterns of network connectivity, and which of these characteristics predict persistent versus transient psychological distress. These studies will provide a foundation for building social network based interventions.

CAPS

Continuous Assessment of Psychological Symptoms

This project is using sensing via mobile and wearable computing to track psychological symptom change in a clinical population engaged in psychological treatment. Data derived from naturalistic phone use and wearables includes natural language use, facial expression, acoustic voice, sleep, physical activity, sleep  and geographic movement.

PROM

Promoting Healthy Adolescent Relationships With Social Media

The PROM study is developing a social media intervention to promote mental health by enhancing healthy sexual and romantic relationships in teenagers. Using an instructional design approach to behavior change, the intervention will target 4 phases of romantic relationships: choosing a partner, maintaining a relationship, choosing whether and when to end a relationship, and dealing with breakups. Online methods including video modelling and text based practice and coaching will promote teens using decision making, self care, and relationships skills.

SENSE +

Integrating mobile sensing into personalized adaptive sleep Interventions

Building on our previous work developing effective sleep enhancement interventions for adolescents, the SENSE + project is integrating individualized assessment of sleep patterns using wearable devices in order to provide adaptive personalized sleep improvement that incorporates just-in-time nudges delivered via mobile devices to enhance behavior change.

MULTI-MODAL

Automatic Multimodal Affect Detection During Interpersonal Interactions

Observational methods of measuring affective behavior have provided critical insights into emotion, socio-emotional development, and psychopathology. A persistent barrier to their wide application is that they are labor-intensive to learn and to use.  Our interdisciplinary team of behavioral and computer scientists will develop and validate a fully automated system for measuring affective behavior from multimodal (face, gaze, body, voice, and speech) input for research and clinical use.

​

Grant funded by National Institute of Mental Health R01MH096951        Cohn (PI)            2017-2022

G-MAP

Geolocating Movement and Psychopathology

This project is examining how geographical movement and location relate to mental and physical health outcomes using passive sensing technology on smartphone devices. Researchers have previously examined the effects of fixed geographical location on health outcomes, but there are few studies that integrate variation in day-to-day geographical location into a comprehensive predictive health model.

BIG DATA

A framework for "Big Data" mental health research on Twitter

The goals of this project are to study relationships and patterns among online behaviors to identify a set of critical metrics that mark important differences among users; to identify and create metrics based on online behavior that are reliable and valid markers of important mental-health variables; and conduct a proof-of-concept study where use our newly identified metrics of online behavior to study community responses to a particularly devastating category of mass trauma events, mass shootings.(NIMH R21 Grant, PI: Sanjay Srivastava)