Date of Award

December 2020

Degree Type

Thesis

Degree Name

Master of Arts

Department

Sociology

First Advisor

Celeste Campos-Castillo

Committee Members

Kenny Joseph, Noelle Chesely

Keywords

ACT, bodybuilding, computational social science, pro-ana, social media

Abstract

This thesis examined the relationship between pro-ana and bodybuilding social media content to understand the similarities between these populations’ identities and help inform social media content policies. Two interrelated studies were used to investigate this relationship: Study 1 used computational methods which compared the content through machine classification of pro-ana and bodybuilding social media posts on Twitter and Study 2 fielded an online survey experiment to compare the perceptions and human classification of content from these populations on Instagram. The findings from both studies broadly revealed that pro-ana and bodybuilding identities are similar, at least in social media content, which raises concern for the current state of social media censorship policies. The results of this thesis highlight the critical need for social media censorship policies to be cognizant of different populations expressing the same content, creating discrepancies when only one is censored.

Included in

Sociology Commons

Share

COinS