Unpacking the Layers: Exploring Self-Disclosure Norms, Engagement Dynamics, and Privacy Implications

TL;DR

This study advances the understanding of online self-disclosure by introducing a multi-label classifier for 11 disclosure types and applying it to a large Reddit corpus. It reveals widespread self-disclosure, with about half of users sharing at least one type in 10% of their posts, and shows that disclosures co-occur and modulate engagement differently across types. The work further demonstrates that interactions from disclosure-oriented communities increase the likelihood of future disclosures and delivers InsightWatcher, a browser plugin to help users manage inadvertent disclosures in real time. These findings highlight notable privacy risks, including exposure of close contacts, and offer a practical tool to mitigate such risks while informing platform design and moderation strategies.

Abstract

This paper characterizes the self-disclosure behavior of Reddit users across 11 different types of self-disclosure. We find that at least half of the users share some type of disclosure in at least 10% of their posts, with half of these posts having more than one type of disclosure. We show that different types of self-disclosure are likely to receive varying levels of engagement. For instance, a Sexual Orientation disclosure garners more comments than other self-disclosures. We also explore confounding factors that affect future self-disclosure. We show that users who receive interactions from (self-disclosure) specific subreddit members are more likely to disclose in the future. We also show that privacy risks due to self-disclosure extend beyond Reddit users themselves to include their close contacts, such as family and friends, as their information is also revealed. We develop a browser plugin for end-users to flag self-disclosure in their content.

Figures (7)

• Figure 1: Cumulative distributions of self-disclosure.
• Figure 2: Correlation between different self-disclosure types that occur commonly per user. Most users who share Age also share Gender and Relationship-related self-disclosures.
• Figure 3: Number of users (top x-axis) and posts (bottom x-axis) for each self-disclosure type.
• Figure 4: Cumulative distributions of per-user mean engagement values, per disclosure type.
• Figure 5: Regression results for engagement. Each panel is a different regression model. The y-axis and x-axis show confounding factors and corresponding estimate, respectively.