Questioning the “Mea Culpa”: Mathematically and Administratively
Prahlad Balaji IyengarI am a first-year PhD student in the EECS department studying quantum information. I was also, along with other students, ejected from the February 14 Institute faculty meeting without reasonable cause. We wanted to learn about the Committee on Academic Freedom and Campus Expression (CAFCE) committee’s plan, hear the faculty debate perspectives on free speech, and demonstrate by our silent but attentive presence the importance of incorporating the student voice. Ironically, both CAFCE committee co-chairs (joining President Sally Kornbluth) followed their presentation emphasizing open engagement with the campus community by voting to exclude us from the room. The motion was brought forth by a tenured faculty member who claimed that videos from a previous faculty meeting had been leaked to NPR. These videos, as far as I am aware, are nowhere to be found.
Prof. Peko Hosoi was invited to a post-debacle meeting with students and faculty to discuss the reason for our frustration, and conceded that it was a mistake to (1) tally the votes for but not against, and (2) not recuse the CAFCA co-chairs from the vote. She communicated her desire to issue an apology about the way the meeting was handled.
A “Mea Culpa” was issued – not to the students, but to the faculty in the January-March Faculty Newsletter. The students received a “special message” praising us for following the rules (and still getting kicked out). To attempt to justify the vote outcome independently of her actions, Prof. Hosoi proposed a mathematical defense, using the mean average of previous votes as an estimate for the expected number of voters in this faculty meeting. This analysis is faulty on three different levels.
First, I question the use of the mean as an estimator. As a toy example, suppose when you ask a cunning child to pick a number from 0 through 9, they alternate between picking either 1 or 3. In the limit, the mean is indeed 2; however, that would be a poor prediction for the very next value. The choice of the mean as an appropriate estimator for this problem makes an assumption about the underlying distribution.
Second, I observe an abuse of the Gaussian. Upon inspection, the graph provided in the “Mea Culpa” suggests that the underlying distribution for the raw number of votes is itself Gaussian. Though Gaussians are commonly used to model statistics about the data via the central limit theorem, this does not imply that the underlying distribution of voters is Gaussian. This abuse of the Gaussian betrays at worst a deliberate manipulation of the reader, and at best a confusion of the role of Gaussian distributions in statistics.
Third, I question the data that was used. The data itself is not conducive to a rigorous analysis. If we wanted to get a distribution of voter participation, we ought to have used the percentage of voters, normalized to the attendance at each meeting at the time of the vote. The model cannot accurately predict the number of voters given the total attendance without taking into account the attendance at each session.
Viewed in isolation, these could be simple mistakes. But together, they indicate a structural issue with understanding the problem.
I extend that criticism to the MIT administration’s approach to CAA and pro-Palestinian organizing on campus. Consider the following three mistakes made by the administration:
- Revoking access to an Instagram page where student groups advertise themselves, falsely implying that the CAA’s Instagram takeover was against the rules. This decision, made without precedent or due process, was later retracted and an apology was issued on the same Instagram account.
- Issuing threats of suspension to student protestors during the November 9, 2023 sit-in in Lobby 7. President Kornbluth, who was in attendance, bypassed disciplinary procedure and review processes. The “justification” that the protest broke the rules was undermined by having changed the rules just 24 hours prior. Kornbluth later retracted this threat due to the steadfastness and solidarity of the protestors.
- Issuing a no-contact order for CAA students on behalf of the IDHR staff. In January, students engaged in symbolic protest by reading Dr. Martin Luther King, Jr.’s Letter from Birmingham Jail in front of the IDHR office. The administration responded by requiring all CAA members, even in their individual capacities, to run all IDHR incident reports by Dean David Randall. This non-confidential step discouraged reporting of Islamophobia and, ironically, violated Title IX regulations. This unconstitutional gatekeeping of the IDHR process was retracted only after lawyers contacted the Office of General Counsel.
These three mistakes could be viewed in isolation. But to members of the CAA these past six months, this pattern of negligence betrays a structural problem in the way this administration approaches certain student groups.
These structural issues also manifest in the way other groups on campus are deprioritized. Jews for Ceasefire has consistently faced administrative barriers which devalue their anti-Zionist Jewish perspective. Asian American Initiative’s SOLE-designated community space was reassigned to other groups without warning. BSU’s (the Black Students’ Union) and BGSA’s (the Black Graduate Student Association) years-long campaign to reduce policing has been ignored by increasing armed police presence in our hallways and lobbies. These are not isolated oversights, they form a pattern of negligent behavior, revealing to the student body which students are considered less equal than others.
To Professor Hosoi and others interested, I present a fuller mathematical analysis here. To the administration, I can only recommend that you turn a critical lens toward your own actions these past two semesters.