I briefly put this as an edited-to-add in the first post
but people already read and commented there so here's a new post.PDF of the actual paper!
WHAT? They ran an ANOVA on a discrete variable. FAIL. And they reported M-values, not p-values. What the crap is an M-value? At least the variance they found was pretty small, which justifies a small sample size but ANOVA is the wrong analysis to use when there are only seven possible values of the dependent variable! And how big can your variance get when there are only seven possible values?
The second part of the study is better, though it still has the problem of using discrete variables in an analysis that's intended for continuous variables.
Also it sounds like the students who participated in the study were involved with the psychology department somehow; they got course credit for it and psych students would be much more aware of this and more inclined to participate than students in other departments. So basically the conclusions can only be applied to undergrads who at least took a psych class, or possibly are psych majors. This is potentially a huge source of bias.
Anyway, their methods are a little screwey but I don't doubt that women who aren't inclined to be geeky would find geek stuff more off-putting than non-geeky men would, because of the widespread perception (spread by the MSM articles about this paper, for example) that geek stuff is only for boys. It's not just the MSM misinterpreting the paper, though: from the paper's discussion section on the first part of the study:
This study suggests that a student’s choice of classes or a major can be shaped by simply the appearance of classrooms, hallways, and offices—therefore, providing compelling evidence for the power of environments in signaling who belongs.
The authors of the study are claiming that it's the posters etc. themselves that are the problem. No, it's that these objects signal to people that it's more likely to be an anti-woman environment than the neutral one, due to stereotypes about geek culture - I mean, it's not as if a non-geeky workplace never discriminates against women. The underlying problem is the discrimination. It bothers me that this paper is going to be used as evidence that we can just put up some different posters and then when that doesn't result in equal representation, just blame it on our ladybrains or "opting out" or whatever other bullshit they come up with that allows the patriarchy to keep steamrolling along.
I think it's reasonable to expect some people will read the articles about this paper and institute a "no geek stuff" policy in schools and workplaces in an effort to make women feel more comfortable. That would be stupid. The only effects would be:
1. to make men resent women in the workplace more, because it would be seen as special treatment
2. to piss off the geeks
3. to really really piss off the geeks who are also female
4. to totally fail to address any of the real sexism that the geek stuff symbolizes to some women
What would be a much better solution is to educate everyone about how they might discriminate or create a hostile environment without realizing it, for example by commenting on people's appearance, giving a man credit for work a woman did, expecting women to clean up after everyone, and expecting women in technical jobs to take care of secretarial tasks. ALL of this has happened to me in molecular biology jobs in academia and industry. And it was not cool. And forbidding me from putting Star Trek crap on my desk wouldn't have changed any of it.