Developing an accuracy-prompt toolkit to reduce COVID-19 misinformation online
Bibliographic record
Abstract
Recent research suggests that shifting users’ attention to accuracy increases the quality of news they subsequently share online. Here we help develop this initial observation into a suite of deployable interventions for practitioners. We ask (i) how prior results generalize to other approaches for prompting users to consider accuracy, and (ii) for whom these prompts are more versus less effective. In a large survey experiment examining participants’ intentions to share true and false headlines about COVID-19, we identify a variety of different accuracy prompts that successfully increase sharing discernment across a wide range of demographic subgroups while maintaining user autonomy. Research questions•There is mounting evidence that inattention to accuracy plays an important role in the spread of misinformation online. Here we examine the utility of a suite of different accuracy prompts aimed at increasing the quality of news shared by social media users.•Which approaches to shifting attention towards accuracy are most effective? •Does the effectiveness of the accuracy prompts vary based on social media user characteristics? Assessing effectiveness across subgroups is practically important for examining the generalizability of the treatments, and is theoretically important for exploring the underlying mechanism.Essay summary•Using survey experiments with N=9,070 American social media users (quota-matched to the national distribution on age, gender, ethnicity, and geographic region), we compared the effect of different treatments designed to induce people to think about accuracy when deciding what news to share. Participants received one of the treatments (or were assigned to a control condition), and then indicated how likely they would be to share a series of true and false news posts about COVID-19. •We identified three lightweight, easily-implementable approaches that each increased sharing discernment (the quality of news shared, measured as the difference in sharing probability of true versus false headlines) by roughly 50%, and a slightly more lengthy approach that increased sharing discernment by close to 100%. We also found that another approach that seemed promising ex ante (descriptive norms) was ineffective. Further-more, gender, race, partisanship, and concern about COVID-19 did not moderate effectiveness, suggesting that the accuracy prompts will be effective for a wide range of demographic subgroups. Finally, helping to illuminate the mechanism behind the effect, the prompts were more effective for participants who were more attentive, reflective, engaged with COVID-related news, concerned about accuracy, college-educated, and middle-aged. •From a practical perspective, our results suggest a menu of accuracy prompts that are effective in our experimental setting and that technology companies could consider testing on their own services.
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 | 0.001 | 0.008 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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".