Online posting effects: Unveiling the non-linear journeys of users in depression communities on Reddit

TL;DR

The study investigates how online self-disclosure in depression-themed Reddit communities relates to changes in expressed well-being, revealing non-linear, spiral-like user journeys rather than a linear recovery path. It combines psycholinguistic profiling (Plutchik emotions via the NRC Lexicon; PAD via the VAD Lexicon; VADER sentiment; Taboo Rate; TextBlob Subjectivity) with social-exposure modeling and conditional Markov chains on monthly snapshots to identify four states ($k=4$) and compute transition probabilities. Using two null models (Cluster and Temporal) to test significance ($p<0.01$), the authors show that exposure to high-distress and fluctuating content can drive transitions in both directions, with some transitions reaching as high as $0.41$ and affecting up to $87\%$ of active users. The work contextualizes these states within the Patient Health Engagement framework and discusses practical implications for online moderation, digital interventions, and supportive design to better accommodate the nonlinear nature of online mental health journeys.

Abstract

Social media platforms have become pivotal as self-help forums, enabling individuals to share personal experiences and seek support. However, on topics as sensitive as depression, what are the consequences of online self-disclosure? Here, we delve into the dynamics of mental health discourse on various Reddit boards focused on depression. To this aim, we introduce a data-informed framework reconstructing online dynamics from 303k users interacting over two years. Through user-generated content, we identify 4 distinct clusters representing different psychological states. Our analysis unveils online posting effects: a user can transition to another psychological state after online exposure to peers' emotional/semantic content. As described by conditional Markov chains and different levels of social exposure, users' transitions reveal navigation through both positive and negative phases in a spiral rather than a linear progression. Interpreted in light of psychological literature, related particularly to the Patient Health Engagement (PHE) model, our findings can provide evidence that the type and layout of online social interactions have an impact on users' "journeys" when posting about depression.

Figures (10)

• Figure 1: Psycholinguistic cluster profiles in Reddit Depression discourse.(A-E) Bar charts of cluster centroids values: Plutchik’s Primary Emotions, PAD Emotional Dimensions, VADER Sentiment, Textblob Subjectivity, Taboo Rate.
• Figure 1: For each cluster (a-d), bar charts display the 10 most frequent topics (in descending order) extracted from users' texts using BERTopic. Each bar chart visualizes the frequencies of the top 8 representative words for each topic (based on TF-IDF) along the x-axis. These words can be unigrams, bigrams, or trigrams. Topics are labeled using single terms that describe their content, such as "job", "work", "school", "college" for Education/work topic. Notice that we label as "Other" those extracted topics (in grey) that are composed of words too general to be categorized with a label.
• Figure 2: Social influence and cluster transitions in Reddit Depression interactions.A) Schematic examples characterizing the four modeled levels of social exposure. Nodes identify users (colored by their profile cluster), and edges/sets identify one-to-one interactions/discussion contexts. B) Visual representation of Markov transition matrices among user clusters conditioned on group-contact typology and filtered based on temporal and volume statistical significance. Rows identify the conditioning profile cluster, and columns the assumed interaction model. In each graph: nodes identify profile clusters; directed edges identify statistically significant transition given the conditioning social interaction with its observed probability; users affected identify the percentage of the online population affected by the identified patterns.
• Figure 2: (a) Distribution of the clusters over time; (b) Distribution of the user permanence: The number of active months on the x-axis indicates the number of snapshots in which users are active on the platform;
• Figure 3: Visual representation of Markov transition matrices among user clusters given the metapopulation only filtered based on volume statistical significance.