They do include the effect size of including non-binary students when they write “(nb. Non-binary students account for 0.3% of this total)” etc. so the impact on the actual data is shown, if you’re concerned about the statistical analysis. It also does make sense to group them together in this context as they are both minorities in STEM. However the way the article is written makes it clear that including non-binary students was an afterthought; if it was clear in all the data and headings that the data is for both non-binary and female students with the interpretation that they are looking at just “students who aren’t men” then it would have been a lot better.
They do include the effect size of including non-binary students when they write “(nb. Non-binary students account for 0.3% of this total)” etc. so the impact on the actual data is shown, if you’re concerned about the statistical analysis. It also does make sense to group them together in this context as they are both minorities in STEM. However the way the article is written makes it clear that including non-binary students was an afterthought; if it was clear in all the data and headings that the data is for both non-binary and female students with the interpretation that they are looking at just “students who aren’t men” then it would have been a lot better.