I don't think code reviewers are looking at the gender of the committer and reaching for their scrutiny glasses. Other than the rank issue mentioned, it's fully possible that the code submitted may have co-incidentally been of lower quality in the window sampled. This has no bearing to the author's gender.

A similar sampling can be done on race, color, religion, sexual orientation ... but it's easy to forget that correlation isn't causation. And I say this as a person belonging to a minority.

Headline grabbing non-news

I don't work at facebook and have no idea about their process.

If code reviews were done with no knowledge of tho submitted the pull request, and the results of the study still concluded that females had higher number of comments and a higher rate of rejections.

Would it still be gender bias?