Competitive Cancelation Dataset: YouTube Responses to Dramageddon (2019)
Description
This dataset contains anonymized YouTube comment data associated with the 2019 online controversy known as Dramageddon, involving beauty influencers James Charles, Tati Westbrook, and Jeffree Star. The dataset was created for research on online hostility, cancel culture, and competitive communication dynamics among influencers. The dataset includes public user comments collected from 14 YouTube videos posted during May–June 2019, including primary source videos from the influencers involved and reaction videos from commentary channels. A total of ~15,000 comments were collected using the YouTube Data API v3. All comments are anonymized and contain no personally identifiable information. Each comment record is enriched with metadata and derived variables, including: - Sentiment score (range −1 to +1) - Toxicity score (probability 0–1) - Cancel behavior classification (cold, cool, hot) - Moral language category - Engagement metrics (likes, reply depth) - Time of posting - Video-level metadata (creator, phase of controversy) This dataset supports research in computational social science, communication studies, digital sociology, and platform governance. It has been used in studies on cancel culture, moral contagion, algorithmic amplification, and influencer reputation dynamics. This dataset contains only publicly available YouTube comments retrieved in accordance with the YouTube Terms of Service. All usernames, channel IDs, and profile references were hashed or removed during preprocessing to ensure anonymization. No attempts were made to identify or contact any YouTube users. The dataset is provided strictly for research purposes. Users must agree to comply with ethical guidelines for internet research (AoIR 2019) and cite the dataset appropriately.