Detecting Disinformation about Health in Social Media: A Review of Four Inductive Studies
Bibliographic record
Abstract
Social media enable fast and widespread dissemination of information, both honest and dishonest. Misinformation is generally spread unintentionally. Disinformation, on the other hand, is intentionally dishonest and is designed to harm individuals and organizations, which usually benefts the sender fnancially. Health-related disinformation can be especially dangerous if people act on its claims. How do people determine if health content on social media is honest or if it contains disinformation? This review considers four papers about inductive research studies, where participants were exposed to actual social media posts about 10 health topics, ranging from weight loss to COVID-19 vaccines. Some of the posts were honest and some were dishonest. Participants in all four studies were asked to evaluate the veracity of the posts that they saw and to provide the reasons for their evaluations. Two studies were online surveys. The other studies were conducted in the lab, where the eye movements of participants were recorded with an eye tracker. The key fndings from the review were: (1) People were relatively good at detecting health-related disinformation, with detection success rates ranging from 66% to 90%; (2) People most frequently cited the quality of the source of a post as the reason they decided it was honest; (3) Variables key to successful detection were need for cognition and gender (and to a lesser extent, political aÿliation, education, and age); (4) In the eye tracking studies, the most common determinants of fxations on particular parts of a post were need for cognition, gender, and the veracity of the post; (5) The most important measure of fxation that infuenced detection success was number of fxations; and (6) Overall, need for cognition was the key factor in successful disinformation detection.
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.
How this classification was reachedexpand
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 | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.003 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".