I ended up having to explain why I'm critical of this paper using words not gifs, so here it is (thread) #WomenInSTEM #WomeninScience #patriarchy #leanin https://twitter.com/megcevans/status/1331572912508190722

First, ok yes, the paper calls on organisations to look past confidence in its concluding comments. But it does this right at the end, after suggesting mentoring to "strengthen students self-assessments", i.e build the confidence of these poor widdle women

(make sure you're not being mentored by other women though!!!)

The study's background justification and hypothess all focus on women's confidence, not men's

Even when discussing "demand side" (i.e organisational hiring practices), it focusses on women

These hypotheses are so simplistic I want to blow my brains out. And yes, that's how hypothesis testing works, but.... maybe... this method isn't up to dealing with the complexities of the problem?

And yes, there's only so much space to talk about stuff in a paper - but - it's singular focus on *~confidence~* as a predictor of course means it's going to come up as...a predictor. They surveyed graduates about self confidence and related that to salary. That's it.

What the study DIDN'T do is survey the organisations or individuals hiring them, nor highlight the systemic socio-cultural issues (other than *~confidence~*) that influence pay gaps.

Finally, the title implies causality. This this would have been published in a supposedly selective and prestigious journal? #PNAS

Thanks @EzzyOD for highlighting this paper from 2012 (yes, in @PNASNews ) that did actually test the "demand side" of hiring https://twitter.com/EzzyOD/status/1331705335040466944?s=19