A cross-cultural comparison of educational video quality on college students' anxiety and depression: a cross-sectional content analysis of YouTube and Bilibili.

OBJECTIVE

To systematically compare the quality of educational videos about anxiety and depression among university students on YouTube and Bilibili, and to provide evidence-based guidance for cross-cultural digital mental-health education.

METHODS

Before 20 November 2025, we searched YouTube and Bilibili with English and Chinese keywords and collected the first 100 videos returned by default ranking on each platform. After applying inclusion and exclusion criteria, the remaining videos were evaluated by a third assessor in a double-blind manner using the Video Information and Quality Index (VIQI), the Global Quality Score(GQS)and the modified DISCERN (mDISCERN) scales to assess scientific accuracy, safety and educational value. Platform differences were analyzed with non-parametric tests and correlation analyses.

RESULTS

The final sample comprised 80 YouTube and 77 Bilibili videos. Median views, likes, and comments were markedly higher on Bilibili (p < 0.05). Verified accounts supplied 43.75% of YouTube content but only 28.57% of Bilibili content; licensed mental-health professionals appeared in fewer than 6% of videos on either platform. YouTube favoured television-style or documentary formats, whereas Bilibili relied heavily on single-speaker narratives and animations. YouTube outperformed Bilibili on overall VIQI, GQS, and mDISCERN scores (p < 0.01). On Bilibili, high user engagement correlated moderately to strongly with quality, yet absolute quality scores remained low.

CONCLUSION

Platform architecture, not popularity, drives content quality. YouTube's longer, institution-produced videos set the benchmark, whereas Bilibili trades scientific rigor for real-time chat and high engagement. Both sites remain short of licensed professionals. To prevent digital platforms from amplifying student anxiety, we recommend (a) embedding a quality-weighted algorithmic boost and (b) a sustained "verified expert + student co-creation" pipeline that disseminates evidence-based content at scale.