Teaching Assessments Reflect—and May Perpetuate—Gender Bias in Academia |  Science

Teaching Assessments Reflect—and May Perpetuate—Gender Bias in Academia | Science

Teaching Assessments Reflect—and May Perpetuate—Gender Bias in Academia |  Science

Universities routinely use student teaching assessments to help make decisions about which faculty members gain tenure and promotions. But factors unrelated to teaching performance, such as gender, race, and even attractiveness, can skew these assessments, potentially exacerbating existing inequalities in academia.

Now, a new study suggests an additional source of bias: being in the academic department’s gender minority. For example, in higher-level courses, women who teach in male-dominated departments tend to do worse in student assessments, according to a study published this week in Annals of the National Academy of Sciences. The same principle applies to men who teach in predominantly female departments, the researchers note, but because women are more often in the gender minority, they are disproportionately affected.

“This just adds to the ongoing avalanche of information pointing to how [much] there is potential for error in using assessments as a way to determine employment,” says Jennie Sweet-Cushman, a political scientist at Chatham University who was not involved in the study.

Past research suggests that, under certain circumstances, individuals can be punished in their professional lives for challenging gender expectations. For example, in child care — a profession traditionally associated with women — men can be negatively biased in assessments, says the new study’s lead author, Oriana Aragón, a social psychologist at the University of Cincinnati. The same goes for women taking on managerial roles in male-dominated fields, she adds.

To see if this bias was true for university instructors, Aragón and his colleagues analyzed more than 100,000 reviews of 4,700 courses at Clemson University, an R1 US public university. They found that if more men were in a department, women had lower average student ratings when teaching higher-level courses, and vice versa. In departments with approximately equal numbers of men and women, this bias has disappeared. For lower-level courses, the differences were not statistically significant.

“The fact that both women and men were penalized illustrates how largely harmful stereotypes are,” says Asia Eaton, a social psychologist at Florida International University. “The studies in this article do an excellent job of studying gender bias in context.”

Next, the researchers designed an experiment, showing students a website for a theory department and varying the proportion of female and male professors displayed in photos on a faculty page. The researchers then presented students with a description of a simulated course in the department, including a photo and biography of an instructor. Finally, students completed a teaching assessment as if they had taken that simulated course. In male-dominated theoretical departments, students rated female professors worst for upper-level courses and men for lower-level courses.

Bringing departments closer to gender parity can help lessen biases in teaching assessments, the authors suggest. Until that happens, they propose placing equal emphasis on the achievements of men and women within university departments and for men and women to teach lower and higher level courses.

However, Sweet-Cushman notes that a department’s gender makeup is likely to reflect existing biases within the discipline and that these biases cause the assessment disparities. “The proportion itself is not the mechanism, I think.”

Angela Linse, associate dean of teaching at Pennsylvania State University, University Park, commends the study for its use of a large dataset and its new experimental design, but cautions against overinterpreting the results. “Gender bias certainly exists, both in the student population and in the faculty,” she says. But the differences in ratings found by the authors – fractions of a point on a five-point scale – “are not necessarily conclusive evidence of gender bias. Not all statistically significant differences are significant differences”.

Overall, Linse agrees that disparities in teaching assessments shouldn’t make or break a college professor’s professional destiny.

“Denying ownership for a small difference in rating scores would truly be a tragedy,” says Aragón.

Leave a Reply

Your email address will not be published. Required fields are marked *