About the Study
A substantial and growing body of research suggests that social media use can be harmful to young people’s mental and physical health. Several recent reviews have linked social media use more specifically to body image concerns, disordered eating behaviors, and dietary intake. While quantitative research on time spent on social media and adolescent body image and disordered eating suggests an adverse relationship, an important gap in the literature is the content that young people view. This may be a critical link between social media and weight-related health.
Social media can also be source of connection and support, and anecdotal evidence suggests that many youth intentionally manipulate their social media algorithms to maximize exposure to certain content (by interacting with similar posts) and minimize other content (by avoiding similar posts). Hashtags such as #bodypositivity, #bodyneutrality, and #bodyacceptance and numerous influencers point to social media corners with countercultural messages about weight, shape, eating, and activity; purposely accessing these and related accounts may allow users to create a social context that is supportive rather than stigmatizing. Research on exposure to body-positive content is emerging, but with mixed results: some studies show that exposure is associated with improvements to body image, while other studies find no associations. This work has important limitations. Much of the research on social media exposures has used experimental designs with female college student samples; while useful, it does not reflect naturalistic exposure in daily life. Other research relies on self-report of viewing different material, which introduces considerable bias (including recall, social desirability, estimation). Novel, objective assessment of exposure to social media content is needed to advance this line of research.
The present study is designed to launch a new line of research that expands upon our earlier Project EAT work with media content and delves into the question of how to “see what youth are seeing” with regards to weight-related messages in the social media space. As a long-standing Project EAT investigator who has expanded the study in novel ways, and a new colleague with expertise in social media, Screenomics, and machine learning, we seek to establish a new collaboration that will advance both our lines of research. We aim to develop new, technology assisted methods that can be brought to scale in future studies, linking novel exposures to physical, mental, and behavioral health outcomes in adolescents. Specifically, we propose the following:
- Aim 1: Use Screenomics for data collection of content viewed on smartphones among 50 Project EAT-connected adolescents and young adults (ages 13-21);
- Aim 2: Review 1000 captured screenshots per participant to establish ground truth data for support and stigma in weight-related content;
- Aim 3: Create machine learning/deep learning (DL) models for automated labeling of all social media screenshots;
- Aim 4: Create novel variables reflecting both supportive and stigmatizing weight-related content and assign exposures scores to each participant.
- Aim 5: Conduct preliminary analysis using new variables, including descriptive statistics of exposure to supportive and stigmatizing content and comparisons across demographic categories.
We expect findings will contribute to future interventions aiming to expand positive and supportive social media messaging and strategies to limit exposure to negative weight-related content.