Tracking patterns in toxicity and antisocial behavior over user lifetimes on large social media platforms

Area of Science:

• Computational Social Science
• Digital Communication Analysis
• Online Behavior Studies

Background:

• Growing concern regarding toxic and antisocial behavior on social media platforms.
• Need for large-scale, longitudinal analysis of online toxicity trends.

Purpose of the Study:

• To analyze large-scale user toxicity trends on Reddit and Wikipedia over 14 years.
• To investigate individual user toxicity evolution and platform-specific toxicity patterns.
• To examine toxicity in discussions related to shared content.

Main Methods:

• Analysis of nearly 500 million comments from Reddit and Wikipedia spanning 14 years.
• Utilized two distinct toxicity proxies to quantify online antisocial behavior.
• Examined user toxicity levels across their platform lifecycles and by activity level.

Main Results:

• A significant trend reversal observed: users became less toxic pre-2013 but more toxic post-2013.
• Reddit's most toxic behavior is linked to highly active users; Wikipedia's is linked to least active users.
• Toxicity trends in content-based discussions mirrored user-based toxicity trends.

Conclusions:

• Online toxicity is evolving, with a notable increase in recent years.
• Platform architecture and user engagement models influence toxicity manifestation.
• Understanding these dynamics is crucial for developing effective moderation strategies.