Statement on MechE qualifying exams
The following statement was issued by the leadership team in MIT's Department of Mechanical Engineering
The faculty, students, and staff of MIT’s Department of Mechanical Engineering are committed to a diverse, equitable, and inclusive environment.
A recent guest opinion column in The Tech raised concerns regarding gender bias in our graduate program – in particular in the mechanical engineering PhD qualifying exams. The Tech article claims that a gender gap1 in passing rates in the qualifying exams is a direct result of gender bias2. This claim is not supported by the data.
In March, departmental leadership was made aware of a student-led survey that suggested a potential gender bias in our doctoral qualifying exams. We took this information very seriously, gathered data over the time frame of 2013-2019, analyzed the data in depth, and had many discussions with MIT community leaders including the MIT ICEO Officer, the Associate Provost, and MIT Institutional Research, as well as with MechE faculty, student representatives from GAME and MEGA Women, and members of the MechE DEI Task Force.
Our analysis shows that a gender gap in passing rates for that time period exists, but the data does not lead to the conclusion that it is associated with gender bias. Departmental policy typically allows two attempts to pass the qualifying exams. Below is a summary of passing rate data from January 2013 through May 2019. These numbers differ from the numbers presented in The Tech column because we used more complete data from the 478 students who took the exams between 2013 and 2019, compared with the 115 people who participated in the student-led survey:
Students who passed the exam in the first attempt**
Students who chose not to take the exam a second time
Students who failed after the second attempt
Students who passed the exam*
§ gender/sex data is based on the best information available to the department
* difference is not statistically significant to a confidence level of 95%
** difference is statistically significant to a confidence level of ~95%
The final passing rates are very high for both men and women (97% and 93%, respectively), and do not differ between men and women in a statistically significant way. When the numbers are adjusted to reflect those who chose not to take the exam a second time, the passing rates increase further to 98% for men and 97% for women. The data, however, demonstrate a gender gap in passing rates in the first attempt. Additionally, the total number of women compared to men who have taken the qualifying exams serves as a reminder that we must continue to increase gender diversity in our graduate program.
We are committed to achieving a welcoming, inclusive and diverse department. The programs and initiatives we have implemented to achieve greater gender diversity have resulted in female students making up 49.6% of mechanical engineering undergraduates at MIT, compared to the national average of 14.8%,. Over the past decade, we have also put considerable effort into recruiting, retaining and supporting women in our graduate programs. In 2010, 16.4% of our graduate students were women. As of 2019, our percentages have increased to well above the national average. In 2019, 28.8% of our master’s students were women (compared to a national average of 14.4%) and 27.2% of our doctoral students were women (compared to a national average of 15.3%)3,4. Additionally, the percentage of our degree recipients who were women rose to well above the national average. In 2019, 24% of our master’s degree recipients were women (compared to a national average of 16.4%) and 30% of our doctoral degree recipients were women (compared to a national average 16.8%)4,5. Still, we have more work to do.
As part of our commitment to continue to improve, we will be taking the following steps:
- Require all faculty to take bias training before the January 2021 qualifying exams.
- Increase data tracking during the qualifying exam, including keeping record of the order of exams that the students take and the gender identity of the faculty examiners.
- Pursue new programs and initiatives to encourage more graduate school applications from women and non-binary individuals and provide more mentorship and support for our female and non-binary students during their time in our graduate program.
We recently met with representatives from GAME, MEGA Women, and the MechE DEI Task Force to discuss these actions further. In the coming weeks, we will be having more discussions with our faculty, graduate students, and alumni as well as experts to determine effective next steps for us to continue to improve and to ensure an equitable graduate program experience for all of our students.
While we have much more work to do, we are committed to a diverse, equitable, and inclusive culture where every member of our department can flourish and achieve their fullest potential.
Evelyn Wang, Department Head
Rohit Karnik, Associate Department Head for Education
Pierre Lermusiaux, Associate Department Head for Research and Operations
Nicolas Hadjiconstantinou, Graduate Officer
1 Gender gap is defined as a numerical difference between males and females.
2 Gender bias is defined as an unfair difference in the way males and females are treated.
5 Source: MIT Institutional Research