The data takes the form of observational summaries of comment sections rather than direct access to raw comment text. Since individual remarks were analyzed, the analysis may underestimate the linguistic specificity of follower language; exact word choice, punctuation, emojis, and their interactional nuances cannot be examined with full precision. The approximately six-week observation period represents a time-limited snapshot of a relational process; longer-term dynamics, including the development and decay of parasocial relationships over months or years, fall outside the scope of this window. While the 12-influencer sample enables comparative analysis across content styles, the findings do not generalize to the full population of TikTok fitness creators.Â
External validity is limited by the fact that the sample may not be representative in terms of follower count range, geographic origin, or content niche. This study does not account for private or cross-platform relational practices, such as direct messages, story interactions, or subscription community exchanges, which carry qualitatively distinct modes of parasocial interaction. Observable comment behavior may therefore not capture the full spectrum of influencer-follower relational practice.
Observational coding introduces internal validity concerns around the consistency and reliability of category assignments, particularly for emotionally nuanced registers that resist clean classification