Pronoun Relations in Social Media (PRISM): Investigating the Connection Between Digital Popularity and Perception

Overview

Parasocial relationships are one-sided, unreciprocated emotional bonds that develop despite having never interacted directly. Their prevalence has increased via the rise of influencers, who create aspirational content interwoven with sponsorships. From a commercial aspect, these relationships can be indicative of a favorable market or successful creator, given that those with parasocial relationships exhibit trust, loyalty, and impressionability: all traits that sell products faster. On the other hand, these relationships can have serious effects on one's psychological well-being, given the expectation of a reciprocity that rarely comes. The directness of language in addressing the creator through comments can be indicative of a parasocial relationship, i.e. the more direct “you” vs. the indirect “he/she/they”. The relative abundances of such pronouns in TikTok comments were compared against various account metrics: the popularity and activity level of both the poster and commenter, as well as the popularity of the post. A web scraper was used to record public, non-restricted data. A Python program then performed correlational analysis (Spearman's correlation and Logistic or Linear Regression) between each feature and 3rd person vs 2nd person pronoun use. A statistically significant (p<0.001) moderate correlation (r=0.35) was found between post popularity (likes, comments, views) and the ratio of 3rd to 2nd person pronouns, indicating that more popular posts tend to have higher 3rd person pronoun usage. This connection provides a framework against which parasocial relationships can be identified: both for finding favorable markets or creators, and for recognizing the psychological effects of parasocial relationships on linguistics and, in turn, perception.

Competition history

• ISEF 2025 Behavioral and Social Sciences · Entry BEHA030

Resources

• ISEF abstract page ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.