...by a very small margin. It's hard to determine how maintainer preceived the submitter's gender, so they tested for how they could have perceived instead (av

...by a very small margin. It's hard to determine how maintainer preceived the submitter's gender, so they tested for how they could have perceived instead (availability of identifying information - and even that is very hard to determine and limited the sample size). Looking at the big picture, a lot of pull requests are accepted from both genders. I'd be concerned if the difference was >20%, but 4% with all the listed limitations in the study doesn't get me excited.