Female peer mentors help retain college women in engineering

AMHERST, Mass. - A new study by social psychologist Nilanjana Dasgupta and her Ph.D. student Tara C. Dennehy at the University of Massachusetts Amherst found that early in college, young women in engineering majors felt more confident about their ability, a greater sense of belonging in engineering, more motivated and less anxious if they had a female, but not male, peer mentor.

At the end of the first college year, a remarkable 100 percent of women students mentored by advanced female peers were still in engineering majors, Dasgupta says. "That number is spectacular because the first year of college is typically the time of greatest attrition from STEM majors, but none of the women with female mentors dropped out," she adds.

This compares with an 18 percent dropout rate for women students with male mentors and 11 percent for women with no mentors, the control group. As the authors point out, women make up more than 50 percent of university students but hold only between 13 and 33 percent of bachelor's and master's degrees in engineering, computer and physical sciences. Engineering is notable, they add, "for having one of the lowest proportions of women among all fields in science, technology, engineering, and mathematics (STEM)."

Further, results of this National Science Foundation-funded study show that having a female mentor maintained young women's aspirations to pursue engineering careers by protecting their belonging and confidence, the researchers say. Both were associated with higher retention in engineering majors. The benefits of mentoring lasted for two years, well after the intervention ended, during the window of highest attrition from STEM majors. Details appear in the current early online edition of Proceedings of the National Academy of Sciences.

Dasgupta explains, "This same-gender intervention didn't increase belonging, confidence or motivation, but it stabilized these reactions and kept them from plummeting in an environment where women students are a tiny minority." Study controls - women with no peer mentor - showed sharp declines in feelings of belonging in engineering, confidence in ability, motivation and interest in pursuing advanced engineering degrees. But having a female mentor preserved all of these.

Also interesting, the researchers say, is that women students' first-year grades were not associated with retention in engineering majors. The assumption is that students who leave a major are doing poorly or lack skills, Dasgupta notes. But they found in the first year of college, women's performance in engineering and related classes was not at all correlated with retention in the major. "What was correlated with retention were their feelings of belonging and confidence," she says. "Women who felt that they fit into engineering and felt confident about their ability persisted in these majors."

The authors say results support the Stereotype Inoculation Model, which predicts that like a vaccine that protects against bacteria, exposure to successful own-group peers serves as a "social vaccine" to inoculate one against noxious stereotypes. This is especially effective during developmental transitions when individuals experience self-doubt and uncertainty.

The researchers began the study in 2011 and have recruited 150 incoming female engineering students over four consecutive years. Mentoring was not mentioned at recruitment, minimizing the chance that students chose to participate because of mentoring. The researchers randomly assigned participants to a female or male mentor who was an advanced student in the same major, or no mentor.

Trained mentors met with participants once a month for one academic year. Dasgupta and Dennehy assessed participants' experiences several times during the mentoring year and one year post-intervention. They are now following participants until one year post-graduation using the same survey to assess belonging, confidence, motivation, anxieties, retention in engineering majors and actual career pursuit.

Dennehy notes that participants rated male and female mentors as equally conscientious, supportive and available, yet women students assigned female peer mentors experienced large benefits, while those assigned male peer mentors looked about the same as controls with no mentors. "A key takeaway is that in the transition to college when young women take classes where they become aware of being a tiny numeric minority, self-doubt may take hold. It is in those critical transitions when female peer mentors are most effective," Dasgupta adds.

The authors point out that while female peer mentors had significantly more desirable effects on first-year women in engineering, "this does not mean male mentors are unimportant. We expect that female mentors' support will become less critical as women move beyond college transition, at which point male and female mentors may become equally effective." Further, "male faculty who are scientists and engineers play important roles as advisors and career sponsors," in women's careers, they note.

The researchers say these findings open the door to testing how generalizable the results are to students in other STEM fields. They also suggest similar effects may extend beyond gender to other underrepresented groups in STEM such as African-American and Latino students and first-generation college students.

Dasgupta adds, "Now that we know this own-group peer mentoring is so effective, I would like to leverage these findings to institutionalize the intervention here at UMass Amherst. I'd like to take this evidence-based best practice and make it the normal part of what we do to recruit and retain underrepresented students in STEM fields broadly. This is now a field-tested remedy that demonstrably grows the pipeline of underrepresented students in STEM."

###

---

---