Article in July 2012 PS.
Based on responses to a survey, they model the likelihood of being an AP, Associate, or Full as a function of a bunch of variables (age, gender, rank of PhD program, etc.).
Since all these models are cross-sectional, is it me, or is their inference that "women are less likely than men to MOVE FROM assistant to associate" (p. 475) completely baseless? Maybe it's true, maybe it isn't, but their data don't allow them to make a determination one way or the other, right?
I mean, we learn that age is correlated with rank. Surely this doesn't mean that, conditional on being an assistant, older people are more likely to be promoted to associates, since we don't have longitudinal data. And the authors are smart enough never to make such a claim; in fact they state "the older one is, the more likely he or she will be at a higher rank" (480).
So why do the authors make that exact claim when it comes to gender (i.e., that it affects "promotion" rather than simply being correlated with rank)? They make it at the beginning, then on p.485: "significant advantage for men in the probability of BECOMING an associate professor."
They also don't know the difference between odds ratio and odds, but that's a common mistake.
Thoughts?