ResearchBox #4341 - 'Meaningless Means #4: Correcting Scientific Misinformation'

BOX INFORMATION

SUPPLEMENTARY FILES FOR
Joseph Simmons, 'Meaningless Means #4: Correcting Scientific Misinformation', Data Colada
https://datacolada.org/127

CITING THIS RESEARCHBOX
Simmons, J. (2025). ResearchBox 4341, 'Meaningless Means #4: Correcting Scientific Misinformation', https://ResearchBox.org/4341. Zenodo. https://doi.org/10.5281/zenodo.15686444

LICENSE FOR USE
All content posted to ResearchBox is under a CC By 4.0 License (all use is allowed as long as authorship of the content is attributed). When using content from ResearchBox please cite the original work, and provide a link to the URL for this box (https://researchbox.org/4341).

BOX PUBLIC SINCE
June 17, 2025   

BOX CREATORS
Joseph Simmons ([email protected])

ABSTRACT
In June 2023, Nature Human Behaviour published, “A meta-analysis of correction effects in science-relevant misinformation”. The paper’s abstract presents a dire conclusion: “attempts to debunk science-relevant misinformation were, on average, not successful (d = 0.11, p = .142).” But maybe things are not so dire. To obtain that result, the meta-analysis did what almost all meta-analyses do: they averaged effects from very different studies [4]. In this case, the meta-analysis averaged together what an intervention accomplished with what it didn't accomplish. When you present people with misinformation and then tell them that it’s wrong, you can measure two different outcomes: (1) A correction effect: How much people correct their beliefs (coded positively) and (2) A persistence effect: How much their misinformed beliefs persist (coded negatively). The meta-analysis includes both the correction effects and the persistence effects and then averages them, and that makes it look like people do not correct their beliefs at all. But when you analyze these effects separately, you can see that people do correct their beliefs, just not all the way. All told, this example highlights, in a particularly stark way, one of the big problems with meta-analysis: Averaging together different things can lead to the wrong conclusion.