Opinion guest column

The MechE qualifying exam fails women

Half of all women who failed the MIT Mechanical Engineering doctoral qualifying examination would likely have passed if they were men

9465 opn graph
Bar chart illustrating the exam pass rates of men and women in the department.
courtesy of nicholas selby

Half of all women who failed the MIT Mechanical Engineering doctoral qualifying examination would likely have passed if they were men. Although this proportion is appalling, to many of us, it is not surprising.

The MIT Mechanical Engineering qualifying exams are notorious. Students study for months to prepare for the two-day qualifier consisting of three 30-minute oral exams and one hour-long research presentation. Most students pass, but those that do not must repeat the process, forgoing a second semester of research progress to study. If they fail again, they are dropped from the program.

In Spring 2020, we surveyed students who took the qualifying exams in the previous three years based on lists distributed by the department. Our survey requested qualifying exam results and demographic information.

Of the estimated 143 men and 50 women who took the qualifying exams in the past three years, 83 men and 32 women responded to our survey. As illustrated above, 84% of surveyed men passed on their first try, whereas only 67% of surveyed women passed on their first try, yielding a gender gap of 17 percentage points. This difference is statistically significant and robust to various tests of response bias. Unfortunately, no individuals identifying as non-binary responded to our survey, which we believe further reflects the lack of gender diversity and equity in our department.

The pass rate gap is not the only horrific example of gender bias in the department. According to MIT Institutional Research, the attrition rate for graduate women in Mechanical Engineering is 27% higher than for men, far worse than the School of Engineering or MIT as a whole. Furthermore, graduate men in Mechanical Engineering outnumber women three to one, and, although that ratio was improving until 2017, it has been steadily worsening since. There are no recorded statistics for students of other genders, which is a failing in itself.

Why is this happening, and how can we change this?

Realistically, the gender gap could be caused by (1) active, explicit and implicit gender bias in evaluations or (2) an exam structure that favors traits traditionally discouraged in women. Based on student testimony, we believe it is both.

To the first cause, numerous studies [1, 2, 3, 4] have shown that there is gender bias in STEM. MIT is likely no exception. This is exacerbated by advisors’ ability to advocate for borderline students during the evaluation process. Research has shown that male managers are less comfortable managing women than managing men, and given that 85% of MIT mechanical engineering faculty are men, male students are more likely to have better rapport with faculty responsible for advocating on their behalf. To the second cause, oral exams may particularly bias against women. Research has shown that men, on average, may have more confidence in their abilities than women and/or that women are confident but are penalized for showing it. Additionally, it is common for a female graduate student taking a qualifying exam to be the only woman in the room, which also introduces biased dynamics. While understudied, these biases likely affect other gender minorities as well. Regardless of the explanation, our qualifying exams demonstrate clear gender bias.

The harms of this demonstrated gender bias in qualifying exams are exacerbated by the exam’s sink-or-swim nature. A needlessly high-stakes exam is damaging to student mental health, career trajectories, and overall well-being. Students who underperform in the qualifying exams should be given mentorship and opportunities to improve, not threatened with expulsion. There is precedent for a more humane approach: many programs, including Stanford Mechanical Engineering, work with students who fail twice to demonstrate their competency in alternative formats.

We presented our findings and possible solutions to our department head and graduate officer in March. They informed us that MIT recently approved funding for our department to hire an equity officer but seemed reluctant to consider needed systemic changes to the qualifying exam structure. Later, they conducted their own analysis of qualifying exam pass rates and also observed a substantial gender gap. Furthermore, they also found a bias in course grades awarded by professors: they reported average GPAs of 4.84 and 4.77 for men and women, respectively. They declined our request to join meetings between them and university administration and concluded, “We believe that a systematic bias cannot be asserted based on the data.”

We disagree. The results on qualifying exams are indicative of a broader gender discrimination crisis within our department. The statistics we present here are not just numbers; they represent people. Thirteen percent of surveyed women never passed the exam and had to leave the program. Their futures were permanently changed after losing years of their professional career to graduate school without earning a doctorate. This outcome was directly related to their gender. This is urgent. The MIT Mechanical Engineering Department must immediately implement creative solutions and validate them with data.

To that end, our recommendations for the department are as follows:

  1. Restructure the qualifying exams to be course-based, similar to those of AeroAstro and EECS.
  2. Arm the incoming Community Equity Officer with the capacity to hold individual advisors accountable for their demographic-specific pass rates (e.g. advising on faculty promotions/raises), develop department strategy, and set priorities to promote justice and support underrepresented groups within the department.

  3. Incorporate experts with experience in changing culture around discrimination into the department Visiting Committee to provide consistent attention to the problems existing in our department and the work being done to address them.

  4. Require department administrators to report to the department and Visiting Committee on demographic-specific acceptance, graduation, and retention rates, including outcomes on qualifying exams.

  5. Make these reports public, reviewed in regular departmental fora.

These issues are broader than just Mechanical Engineering; equity issues at the Institute are pervasive. We encourage other student advocates to examine your own academic hurdles critically to understand the bias and inequitable barriers they impose. In recognition of the need for an institute-wide response in addition to work at the department level, we also support the RISE campaign. RISE is a united effort of MIT grad students that builds on the petition to Support Black Lives at MIT, the Grads for a Healthy MIT mental health campaign, and grassroots department-level student advocacy with a comprehensive list of demands that would certainly make MIT more inclusive for all. Particularly relevant to our department, the campaign calls for departmental Diversity, Equity, and Inclusion officers who are responsible for reviewing and publicizing data, like qualifying exam attrition rates, and advising departmental leadership on modifying practices in response to this data, among other duties. Experience has shown the efficacy of these diversity officers at many other institutions and departments (read more in a recent op-ed titled “Every department deserves diversity”). We call on our readers and the department to sign the RISE petition and join us in advocating for a more equitable campus. 

Becca Kurfess, Bethany Lettiere, and Jerry Ng are PhD Candidates in Mechanical Engineering at MIT. Nicholas Selby is a PhD Candidate in Electrical Engineering and Computer Science at MIT. All four authors passed the MIT Mechanical Engineering doctoral qualifying examination.