Oh man these guys are lucky that I'm spending 100 hours a week dealing with #COVID19 right now.

https://www.tandfonline.com/doi/pdf/10.1080/03075079.2020.1723533

(retweeting b/c original tweet was deleted)

While I don't have the time to spend on this, there are so many things that could conceivably be wrong with this.

1. For starters, we know men self-cite ~60% more than women do. That helps substantially right there.

2. Citation rates vary hugely among fields and even subfields within fields. Women may be overrepresented in fields or subfields with lower citation rates. We've seen this sort of thing in previous analyses.

3. "The results show that at this point, and regardless of other potential differences in age, funding, number of children etc., female professors had, on average, lower levels of scholarly achievement than male professors. "

No, they had fewer papers and citations.

4. In the replies to the original thread, @lorakolodny mades a critical point: https://twitter.com/lorakolodny/status/1248076210233470978

Yes, absolutely! Otherwise you can't conclude that a higher mean for males implies a higher bar for males. Do the authors address this?

I'm not saying that this particular problem occurs in these data, but here is the sort of thing that can go wrong.

In this diagram, men and women have the same modal number of citations. The citation distribution for men has a longer right tail, for any number of reasons. Promoted men have a higher mean number of citations than promoted women.

But the bar is higher for women than for men!

If you look at un-promoted or those who are denied promotion, you could pick this up. Unpromoted men have a *lower* mean that unpromoted women in this diagram.

Again, not saying this has happened in the data. Just saying that a higher mean does not imply a higher bar.

I could go on, but in fairness to the authors I'm putting these out there as potential pitfalls rather than certain mistakes.

Someone else is going to do the heavy lifting on this one.

(Probably a woman, given the way academia seems to work. It's OK; I'll get the credit. /s)

But finally, FFS, if you are telling sex (rather than gender) from names and pictures, you're doing it wrong.

Bonus: The acknowledgments show a stunning lack of self-awareness. Rarely do people so blatantly out themselves as Dunning-Kruger victims.

h/t @BugsWormsNBats