An Analysis of Linguistic Factors Influencing the Number of Evaluations on Fashion Posts in SNS

Abstract

This study empirically examined the impact of linguistic features on user engagement with fashion-related content on social networking sites, focusing specifically on “Outfit of the Day” (OOTD) posts. The collected posts were categorized into high-like and low-like groups based on the number of likes, and their textual content was analyzed using LIWC 2022. Linguistic features were organized into three dimensions: self-expression, social interaction, and contextual framing. Independent samples t-tests revealed statistically significant differences across several linguistic categories. To control for multiple comparisons, a False Discovery Rate (FDR) correction was applied. Subsequently, logistic regression analyses were conducted using the significant variables to assess their predictive power for engagement. The results indicated that positive emotional words, body-related terms, informal expressions, and temporal and spatial references were positively associated with high engagement. In contrast, achievement-oriented vocabulary, overly emotional language, and second-person pronouns were found to be negatively associated. These findings offer new insights into the role of linguistic framing in the context of digital fashion communication.