Depression Video Quality And Engagement Metrics From TikTok
by yanbin lin·Updated 3mo ago
Available on 1 platform
Sign in to view source links and access this dataset
Description
A dataset contains extracted data from a study evaluating depression-related videos on TikTok. It includes video characteristics, publisher types, engagement metrics, and quality scores from mDISCERN, JAMA, and GQS evaluations. Two independent reviewers assessed the videos, with the data supporting analysis of video quality, reliability, and educational value.
Use Cases
Analyze correlations between engagement metrics (likes, comments, shares, collections) and quality evaluation scores (mDISCERN, JAMA, GQS).
Compare video characteristics and publisher types to identify patterns in content quality and audience engagement.
Train a model to predict quality scores (mDISCERN, JAMA, GQS) based on publisher types and video characteristics.
Investigate the reliability and educational value of depression-related content by statistically analyzing reviewer scoring results.
Strengths
Includes multiple quality assessment frameworks: mDISCERN, JAMA benchmark criteria, and Global Quality Scale (GQS).
Data was evaluated by two independent reviewers, supporting reliability for statistical analysis.
Contains both content metrics (video characteristics, publisher types) and audience engagement metrics (likes, comments, shares, collections).
Limitations
Sample size and temporal coverage are unknown, limiting generalizability.
Focus is exclusively on TikTok (Douyin) for depression content, lacking comparative data from other platforms mentioned in the title (Bilibili).
Potential for subjective bias in human-scored quality evaluations, despite dual review.
Provenance
Source
figshare, author yanbin lin.
Collection Method
Extracted data from TikTok (Douyin) videos, evaluated by two independent reviewers using specified quality scales.
Time Range
null
Freshness
Last updated March 2026.
Geography
null
License is CC BY 4.0. The dataset title mentions a comparative analysis with Bilibili, but the description only references data from TikTok (Douyin).