I’m so very sorry if this is your work environment. My sympathies to all of the women who have been poorly treated at Baylor. I hope they all get rockin’ jobs somewhere better. I would imagine that they’ve got very strong records if they were even permitted to go up for tenure there.

Small sample size or not, those numbers are significant and speak to a larger institutional problem.

If the sample size is indeed too small, as you appear to acknowledge above (but the insidehighered.com article ignores), then talking about — and comparing — things like annual “rejection rates” is a slippery slope.

Among other reasons, it allows the other side to play the same game.

Suppose that Baylor has five women up for tenure next year. Three get tenure. Suddenly, Baylor can now boast of how its tenure rate for women has nearly doubled!

The problem is that even at a biggish university like Baylor, there are still relatively few regular women faculty that any year-to-year comparison is always going to be too small a sample. And of course, institutions always claim have perfectly logical and clear reasons for rejecting anyone’s tenure application. A twenty- or thirty-year analysis of Baylor’s record of tenuring women would be nice, but kind of beside the point of this post, which was about the sudden and dramatic rate of tenure denials this year in particular (going from 10 to 40% from 2007 to 2008). My point in this post was to highlight the distinct possibility that Baylor’s process wasn’t playing fair with anyone, and that (as readers of this blog might suspect) they were playing even more unfairly with the women. Neither I nor the readers here can know the details of each tenure case–but it’s fascinating how in every instance it’s women faculty who pay a higher price. When administrators interfere in in tenure cases, it’s always more women who lose, not more men.

And I agree. It does seem improbable that it was mere random variation. But, after a closer look, the laws of statistics say: with a sample of only 9, it’s entirely possible and, alas, not very meaningful.

Suppose a baseball player has 6 hits in 9 at-bats against a certain pitcher. Does he absolutely dominate this pitcher? Many fans and announcers will reply: of course! But the truth is: we don’t know. We don’t have nearly enough data.

To reiterate, it would appear that something fishy is going on. And it would certainly not surprise me if gender discrimination were going on. But one can’t base that on a tiny data point taken from one year. I’d be much more interested in Baylor’s aggregate data over the past two decades or so.

The sample size — a sub-set of an already small sample size — is simply waaaay too small.

I get it but I think it is complicated. I like the idea of 2 years to tenure better, but does it bring its own complications?