Definitely agree effectiveness is subjective (and corrections do have effectiveness issues that aren't always clear in research). However, I also believe meta-analyses need to clearly define outcomes to ensure accurate interpretation, especially when outcomes used are inconsistent with convention.

Thank you so much! Glad to be able to clarify the puzzling results 😊

Huge thanks to my brilliant coauthors @brionyswire.bsky.social, @ulliecker.bsky.social, Joe DeGutis & Li Qian Tay, and to the original authors for their respectful engagement with our Matters Arising. (5/5)

So, while corrections are far from perfect, they are effective. (4/5)

However, when we analysed the two outcome variables separately–consistent with convention–we find that corrections do reduce, though do not eliminate, the effects of science-relevant misinformation. (3/5)

The paper reached this conclusion by pooling two different types of outcome variables into a single effect estimate. Because the two variables tend to have equal and opposite effects, combining them produced an average null. (2/5)