Date of Award

2026

Document Type

Dissertation

Degree Name

Philosophy (Ph.D)

Department

Psychology

First Advisor

William Chaplin

Second Advisor

Raymond DiGiuseppe

Third Advisor

Melissa Peckins

Abstract

The growth of social media use has been accompanied by evidence linking social media to negative mental health outcomes. The current study focuses on an underexplored area of social media: how individuals describe, structure, and interpret their interpersonal worlds in online and offline contexts, and how these structures relate to mental health outcomes. In addition, we assessed mental health outcomes separately in the social media and in person domains. Drawing on Personal Construct Theory and the eight octants of the Interpersonal Circumplex, we proposed four hypotheses, using both person-centered and variable-centered approaches. We hypothesized 1) individuals will rate those in the social media life as more extraverted and agreeable, 2) those with a two-factor structure in their in-person and social media lives will have better mental health outcomes 3) individuals with more consistent social media and in-person identities will have better mental health outcomes, and 4) that sex, age, time on social media, and number of media sites visited would moderate the relationship between the structural variables and mental health outcomes. We tested these hypotheses in 167 adults (124 women, 43 men) who completed a modified repertory grid test (organized into two data matrices) and 4 outcome measures assessing anxiety, anger, depression, and self-esteem. Results: For our first hypothesis, we found that people rated those in their social media worlds as more dominant, arrogant, and extraverted. For our second hypothesis, factor structure was not associated with mental health outcomes. For the third hypothesis, we found that the higher the within subject correlation across the eight adjectives between the two domains the lower the scores were on the negative mental health outcomes. For our final hypothesis, we generally found that the moderators affected the relationship between the structural variables and the mental health outcome measures. Conclusion: Overall, three of the four of our hypotheses were generally supported and suggest that differences in how we describe and structure our social media and in-person lives are meaningfully associated with psychological well-being. This highlights the importance of considering cross-context identity structure when examining the psychological implications of social media use.

Download

Share

COinS