This is really one instance of a much larger problem in the "soft" sciences, where results that don't replicate are commonplace, and 'null' results that don't a

This is really one instance of a much larger problem in the "soft" sciences, where results that don't replicate are commonplace, and 'null' results that don't appear to give "decisive" evidence for some effect are filed away and go unpublished. It's a mistake to focus on "gender bias" alone, it's a funny example but not really what's going on.