The researchers analyzed numbers from the National Center of Education Statistics that represented roughly 15,000 students across the country as well as teacher surveys in which math teachers were asked to assess individual students. Teachers were asked to express whether they felt their math class was too easy, too hard or appropriate for each student. By marrying the data, Riegle-Crumb and Humphries were able to determine whether the teacher’s attitude and opinion of each student was in line with the students’ actual scores.

The results were anything but assuring.

There was a clear divide between teachers’ positive assessment of their students’ abilities and their actual scores. (Read: teachers said they were doing well when really, not so much). But more upsetting was that the converse was true for white female students: Their math teachers consistently reported that they were doing more poorly in their classes than they really were.

It should be noted that while white female students came out on the bottom of this recent assessment, black females were actually favored by math teachers, who reported that they were performing better than their scores reflected.  This is in contrast to previous studies that pointed out racial bias in perceived IQs. The researchers summarized their findings: “Once we take into account that, on average, Black and Hispanic male and female students have lower grades and test scores than white males, teachers do not rate the math ability of minority students less favorably than students belonging to the traditionally advantaged category of white males.”

White females, on the other hand, were consistently rated as low-performers, regardless of the level of skill or aptitude reflected in their scores. Reigle-Crumb and Humphries offered one possible (but still completely unfair) explanation of the bias: that math teachers view black females (who make up a very small number of high-level high school math students) as having had to try harder to be successful, which could inspire higher levels of confidence or academic potential.

In addition, they posit that teachers may be “more sensitive to their own tendencies towards racial bias than gender bias.” Meaning they wouldn’t knowingly discredit a student based on race, but as gender bias is “so ingrained,” (and what? Innocuous?), it is “harder to notice and therefore harder to resist.”