@mathew @celesteh @-dawid_dettlaff@mastodon.social @cramerd @GeePawHill
Loath as I am to appear to agree in any way with the worst person in this thread, having read the study it definitely seems to have interpreted the data problematically.
- The study seems more likely to have attracted sexual aggressors than general population. The authors were aware that the way they phrased their question might discourage aggressors from participating and so we're careful not to push them away, but seemingly failed to consider that their phrasing might encourage participation of aggressors in ways it does not for non-aggressors.
- While all forms of sexual coercion are bad, some are definitely worse than others; and the less bad ones were much more commonly selected than the more bad ones (and for some of the categories, possibly because I'm not well versed in the language of sexual coercion, I'm not sure what the authors mean). The authors note that low down in the paper as evidence of authenticity, but they don't separate them out by numbers. Raw data is available upon request, so it is possible to review this more thoroughly, but I currently cannot do so.
- This survey was conducted over the internet, offered a monetary reward, and targeted young men. All of these are known problems in recent surveys, as bots are deployed to obtain money in this way. One study (which is probably not the best, but my time is limited) is here: https://pmc.ncbi.nlm.nih.gov/articles/PMC11646990/ YouGov recently retracted one of their surveys due to inaccurate data. The authors claim to have attempted to account for that, but with raw data currently unavailable ("Available upon request" fuck off, open or bust. At least donate to the Internet Archive $2/GB (for what is probably <5 MB, so <1¢) and link to that.) I can't say how well they did.
- Minor note: I'm not thrilled with the occasional typo here. It suggests that the peer reviewers were a little sloppy when they were reviewing their peers. On top of that, some of the polished data is clearly incorrect, with percentages that add up to far more than 100.
So overall, it's an important study for an important topic, but we need to stop repeating the results as being "95% of young men use sexual force against women" when, without further review (currently no other scholarly articles cite this one, which is understandable given how recent it is), the actual conclusion is the much less compelling but more nuanced and accurate "95% of respondants to a survey targeting male sexual aggressors, some of whom are likely bots, have used some coercive strategy to obtain sex, with the majority being at least somewhat socially acceptable. Also, some of the data is mathematically impossible." Any glaring flaws in the popular perception of this study will be used to discredit future studies, so we should address them now before an incorrect idea becomes entrenched.
And to be very clear, I do not support sexual (or otherwise) harassment, coercion, or assault in any way, nor do I consider men (or anyone else) to not be responsible for their actions, with limited exceptions for coercion. I understand that by posting this I am accepting that people will read my words without thoroughly understanding them and will send what in their minds will be justifiably angry or dismissive replies that fail to address what I actually said and meant. But please don't.