This piece with quotes from @profsanderlinden.bsky.social and @jayvanbavel.bsky.social does a bit of a dive into whether fact-checking works, surrounding the Meta platform policy changes. It notes a large (N=20K) meta-analysis indicates a significant positive effect. www.nature.com/articles/d41... - ThreadSky

About ThreadSky

jbakcoleman.bsky.social • 44 days ago

This piece with quotes from @profsanderlinden.bsky.social and @jayvanbavel.bsky.social does a bit of a dive into whether fact-checking works, surrounding the Meta platform policy changes. It notes a large (N=20K) meta-analysis indicates a significant positive effect.

https://www.nature.com/articles/d41586-025-00027-0

Comments

jbakcoleman.bsky.social•44 days ago

While this is the standard scientists use to determine if something works, it gets conflated in this piece and elsewhere with "gets the job done". Its a very different question to ask whether fact-checking can sufficiently/meanignfully reduce misinformation on-platform.

jbakcoleman.bsky.social•44 days ago

Jay's comments get at this a bit, and some of our work has tried to put estimates on the impact. In our model, it would falll into the "nudge".. which we find doesn't do a lot on its own. It works in the scientific sense in our model, but...

https://www.nature.com/articles/s41562-022-01388-6

jbakcoleman.bsky.social•44 days ago

...not at the scale required to make a meaningful dent in the distribution of information on a platform. I think it's important to move beyond NHST criteria for whether something works and start thinking more like applied scientists. Does a vaccine with significant 10% effectiveness "work"? Etc..

jbakcoleman.bsky.social•44 days ago

If we need a 20,000 person meta-analysis to develop confidence in a method, it probably isn't up to the task on its own. If this is what science has found so far, it suggests the platforms have more fundamental design flaws that need addressed which go beyond the scope of quick fixes.

digthatdata.bsky.social•44 days ago

For the purpose of science communication, we should probably favor the phrase "measurable" in place of "significant".

jbakcoleman.bsky.social•44 days ago

Or even “detectable”

ngamita.bsky.social•44 days ago

📌

Posting Rules

Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service

Comments

Posting Rules

Reply