Tied In on TikTok: Tie Strength and Emotional Dynamics in Algorithmic Communities

Abstract

Whether genuine communities can form on algorithmically-driven short-form video platforms like TikTok remains an open question, given that user interactions are often brief, dispersed, and difficult to trace. Building on theories of tie strength and online community formation, we examine whether eating disorder (ED) discourse on TikTok exhibits behavioral and emotional signatures of strong ties, including more frequent, reciprocal, and affectively intense interactions. In this paper, we analyze 43,040 ED-related TikTok videos and over 560,000 comments, alongside a Non-ED comparison dataset. We find that at the user-pair level, greater interaction frequency is associated with increasingly positive emotional expression, a pattern that is amplified in ED-related conversations. This trend is also reflected linguistically, with pairs that interact more frequently exhibiting more of a positive tone. At the same time, how a relationship starts matters: pairs that begin with positive exchanges usually stay mostly positive as they continue interacting, while pairs that begin negatively may add some positive exchanges over time but rarely become mostly positive. To contextualize these dynamics, we classify ED videos into three content types (Pro-Recovery, Pro-ED, and ED Experiences) and find that each exhibits distinct emotional interaction patterns. These findings suggest that dense, emotionally structured relationships can emerge within ED discourse on TikTok. More broadly, our work provides one of the first empirical demonstrations of how community-like relational dynamics form and persist on algorithmically driven short-form video platforms.

Figures (9)

• Figure 1: Distribution of emotional tone across interaction bins for Non-ED (1a) and ED (1b) commenters. Each bar represents the proportion of negative (red), neutral (gray), and positive (green) comments within a given interaction frequency range. Compared to the Non-ED baseline, ED commenters show a sharper decline in negativity and a stronger amplification of positivity as interaction frequency increases, particularly in the 20+ bin.
• Figure 2: Standardized residuals from Chi-Square Tests of Independence for Non-ED commenters (2a) and ED commenters (2b). Positive values (red) indicate overrepresentation of tone categories relative to expectation, while negative values (blue) indicate underrepresentation. ED commenters show a stronger amplification of positivity and sharper underrepresentation of negativity at higher interaction bins (20+), compared to the more moderate shifts observed in Non-ED commenters.
• Figure 3: Tone score distributions across user pair relationships by interaction bin for ED TikTok content. Violin plots show the distribution of relationship-level tone scores within each bin, with black dots marking medians and the dashed line indicating neutrality. Distributions shift toward higher values as interaction frequency increases, reflecting more positive and more stable emotional expression.
• Figure 4: Relationship-level interaction patterns by initial emotional tone. Bars (left y-axis) show the percentage of relationships that are mostly positive at different interaction levels, while lines (right y-axis) show the average share of positive interactions.
• Figure 5: Standardized residuals from Chi-Square Tests of Independence for emotional tone (negative, neutral, positive) across interaction bins in ED-related TikTok content. Panels show differences by content type: (5a) Pro Recovery Content, (5b) Pro-ED, (5c) ED Experiences. Recovery content exhibits strong amplification of positivity and suppression of negativity at high interaction frequencies, while Pro-ED remain near neutral with little systematic change. ED Experiences show early supportive responses that diminish with more interactions.