Economist: How students judge their profs depends on gender

I found the graph below in the January 23rd 2016 issue of The Economist. (Source: This Economist.com webpage.)

 

This graph counts the number of times professors have been described as "horrible" or "brilliant" in reviews on the website RateMyProfessors.com, according to the professor's gender. Strikingly, female professors are described as "horrible" more often than men in every single discipline considered, and male professors are described as "brilliant" more often than women in every single discipline considered. An earlier study by a professor at Northeastern University found that "men were more likely than women to be described as "intelligent" or "funny" but less likely to be described as "nice" or "mean"." 

The three disciplines where men are most often described as brilliant are philosophy, english and history. (English also has the biggest gender gap in terms of the difference between the number of times a male vs female professor is called brilliant. The smallest gender gap is in biology, followed by business. Interestingly, biology and business are also fields where you have a lot of women students.) The three disciplines where women are most often described as horrible are mathematics, economics and computer science. The gender gap is smallest in mathematics (! male and female math professors are equally horrible, apparently) and largest in computer science. I really wonder what makes the students say of (some) female professors they're horrible and (some) male professors they're brilliant... Thoughts?