For the top quartile of female students... having a higher proportion of female professors in introductory math and science courses significantly increases the likelihood [of] ... choos[ing] a [science or technology] ... major.
Although objective measures suggest that when they enter college women and men are roughly equal in their aptitude and preparedness for careers in science, technology, engineering, and math (STEM), female college students are 37 percent less likely than male college students to obtain a bachelor's degree in one of those fields. Why is that the case, and could it change?
In Sex and Science: How Professor Gender Perpetuates the Gender Gap (NBER Working Paper No. 14959), co-authors Scott Carrell, Marianne Page, and James West exploit a unique dataset of 9,481 students who comprised the U.S. Air Force Academy's graduating classes of 2000 through 2008 to try to answer these questions. At the Academy, students are randomly assigned to course sections after taking placement exams. The students are all high achievers, with average SAT math and verbal scores at the 88th and 85th percentile respectively. Furthermore, the students have no ability to choose required course professors: all students take the same course, are taught from the same syllabus, and take the same exams.
The researchers find that female students' course grades in math and science are improved when they have a female professor, and that for the top quartile of female students as measured by SAT math scores, having a higher proportion of female professors in introductory math and science courses significantly increases the likelihood that women will choose a STEM major. Overall, their estimates suggest that increasing the fraction of female professors from zero to 100 percent would completely eliminate the gender gap in math and science majors.
On the whole, men in this sample with the same entering math ability perform substantially better than female students in introductory math and science courses. However, this gap is mitigated when top performing female students have female professors in math and science classes. Professor gender appears to be irrelevant in the humanities, though, and does not appear to affect male performance.
When all female students are considered, rather than only those in the highest quartile, having a higher proportion of female professors does not effect a woman's likelihood of taking higher level math courses or her probability of graduating with a STEM major.
-- Linda Gorman