Examining algorithmic biases in YouTube’s recommendations of vaccine videos
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
OBJECTIVE: This research examines how YouTube recommends vaccination-related videos. MATERIALS AND METHODS: We used a social network analysis to evaluate how YouTube recommends vaccination related videos to its users. RESULTS: More pro-vaccine videos (64.75%) than anti-vaccine (19.98%) videos are on YouTube, with 15.27% of videos being neutral in sentiment. YouTube was more likely to recommend neutral and pro-vaccine videos than anti-vaccine videos. There is a homophily effect in which pro-vaccine videos were more likely to recommend other pro-vaccine videos than anti-vaccine ones, and vice versa. DISCUSSION: Compared to our prior study, the number of recommendations for pro-vaccine videos has significantly increased, suggesting that YouTube's demonization policy of harmful content and other changes to their recommender algorithm might have been effective in reducing the visibility of anti-vaccine videos. However, there are concerns that anti-vaccine videos are less likely to lead users to pro-vaccine videos due to the homophily effect observed in the recommendation network. CONCLUSION: The study demonstrates the influence of YouTube's recommender systems on the types of vaccine information users discover on YouTube. We conclude with a general discussion of the importance of algorithmic transparency in how social media platforms like YouTube decide what content to feature and recommend to its users.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it