PROJECT SUMMARY/ABSTRACT
Suicide rates among adolescents have increased over the past decade. During this same time period, social
media sites (e.g., Instagram, Snapchat, Facebook) have become ubiquitous, with nearly 97% of youth now
using social media. Social media offers both risks and benefits for adolescent development. However, there is
likely a vulnerable subset of youth for whom social media use influences suicidal thoughts and behaviors
(STBs). The primary goal of this study is to examine how and for whom social media use contributes risk for
STBs, so that we may better identify and intervene with youth most at risk. A multi-method approach,
integrating computational psychiatry and experimental techniques, will be used to investigate the specific
mechanisms by which social media use affects STBs. Data will be collected from both an existing dataset (n ~
2,000) and a recruited sample (n = 90 adolescents, ages 14-17, with lifetime history of STBs). To identify
specific, directly observable social media experiences that heighten risk for STBs, digital trace data (i.e., social
media content and meta-data) will be collected directly from adolescents’ social media pages and analyzed
using machine learning techniques. In addition, an innovative laboratory paradigm with eye-tracking will be
used to examine cognitive and affective responses to social media use that increase vulnerability to STBs.
Participant STBs will be assessed at baseline and 6-month follow up, with suicide-related hospital visits
tracked over 12 months. The aims of the study are well-aligned with the candidate’s long-term career goal of
becoming an independent investigator of adolescent social media use and mental health. To purse this
research agenda going forward, the candidate proposes training in four key areas: (1) machine learning
methods for analyzing digital trace data, (2) experimental eye tracking/pupillometry paradigms, (3) mechanism-
based translational research in adolescent suicide risk, and (4) professional development, with an emphasis on
training in effective interdisciplinary collaboration and the ethics of social media research. A team of mentors
and consultants from multiple disciplines will facilitate the candidate’s training, including experts in youth
suicide risk, machine learning and natural language processing, and experimental peer interaction paradigms.
The diverse resources available at Brown University and Rhode Island Hospital will ensure successful
completion of the project and foster the development of the candidate’s research program. The proposed study
has the potential to contribute significantly to our understanding of the role of social media in adolescent
suicide risk. Completion of the proposed research and training goals will uniquely position the candidate to
become a leader in the field of adolescent social media use.