Chad Topaz Queer DEI Race Traitor

@chadtopaz.bsky.social

Data science + math for social justice. Violist, yogi, husband, dad to human & dogs. 🏳️‍🌈 Views do not represent my employers.

Author of Unlocking Justice: The Power of Data to Confront Inequity and Create Change.

Order at: https://bit.ly/4nT7qUh

@chadtopaz.bsky.social

Our paper about legacy status and college admissions and underrepresented minority (URM) students was accepted today! Key result: "In the primary sample... reported legacy consideration is associated with first-year URM shares 6 to 8 percentage points lower..." To appear in Discover Education.

Legacy preference in undergraduate admissions is a policy favoring applicants with family ties to alumni. Critics argue that it reproduces advantage across generations. The issue has taken on heightened importance after Students for Fair Admissions v. Harvard. A basic question remains: Do institutions that report considering legacy enroll different shares of underrepresented racial and ethnic minority (URM) students? No centralized dataset has linked comparsble admissions-priority reports to enrollment composition. In Common Data Set reports, institutions rate admissions criteria on a shared scale and disclose enrollment by race and ethnicity. We assemble these reports into a 2010–2022 panel of highly selective US institutions and verify enrollment against federal records. In the primary sample of 36 institutions, reported legacy consideration is associated with first-year URM shares 6 to 8 percentage points lower, net of sector and other stated priorities. Because every private institution in the sample reports considering legacy, the estimate is identified by the contrast among public institutions that do and do not consider legacy status. The association is statistically significant under HC2 standard errors but not under conservative HC3 errors, and remains negative in every specification we examine. Parallel associations appear for Asian enrollment and Pell Grant receipt. In an expanded sample using federal enrollment records, the URM association is smaller and not statistically significant, while the combined URM and Asian association remains large and significant under HC2 inference. Samples are modest and unobserved characteristics remain uncontrolled, so we interpret these patterns as descriptive associations, not causal effects.
@chadtopaz.bsky.social

I for one am shocked — SHOCKED – that the econ prize has gone to two white men. Oh also, it's my favorite day of the year in a way becuase I get to remind everyone that what we call the Nobel Prize in Economics isn't a real Nobel Prize (true story, look it up). Economists love hearing this! ;)

@chadtopaz.bsky.social

Yeah I just had a paper desk rejected for being too long. In the same desk rejection, the editor also said "it would be helpful if you added discussion of [literally a list of five major areas]."

@chadtopaz.bsky.social

Many have posted the news story below. I have three responses. (1) I am not a fan of offloading human cognition onto genAI. (2) But I think the research article in question has, *potentially*, some serious problems. (3) If we are going to critique AI, it MUST be on the basis of solid evidence.

@chadtopaz.bsky.social

This feels so familiar. See skeet below from the incomparable @ruha9.bsky.social and then see screenshotted excerpt from my book Unlocking Justice. At least I can now say that in this respect I am in good company with one of my heroes, lolcry.

My clash with a third mathematics society came in 2025, with the Society for Industrial and Applied Mathematics (SIAM). I had been asked to submit a talk to a SIAM conference session on data, equity, and inclusion. My submitted conference abstract named the political moment we were living in—a government reopening Guantánamo, rolling back civil rights, erasing queer people—and argued that mathematical scientists must use data to fight back. The talk was accepted. Months later, I received an email from a SIAM officer telling me my abstract was “too political,” lacked “scientific neutrality,” and would have to be rewritten in order to remain on the conference program. I was stunned. Here was a society whose own policies affirmed the open exchange of ideas, now demanding that I soften my words to avoid discomfort. When I refused, politely but firmly, the organization persisted. A cascade of messages said my abstract was outside “the expectations of scientific discourse,” and suggested “less political” rewrites, complete with replacement sentences about “institutional change” and “societal shifts.” Imagine being a 50-year-old professor, with more than 200 talks behind you, and being told that you don’t understand the “expectations of scientific discourse.” I refused to edit my talk and was removed from the program. What became clear to me was not only that SIAM was unwilling to host an urgent conversation, but that it was willing to censor language about justice in order to protect a false idea of neutrality.
@chadtopaz.bsky.social

Yeah yeah, it's cool that the science Nobels have gone to men and the literature one just went to a woman. As we know, men are just more sciency and women are more words-y. 👀👀👀👀👀 Cool cool cool.

@chadtopaz.bsky.social

Advice nedded from social scientists: Journal recommendations for a dataset paper in criminology/empirical legal studies? Our megajournal submission STILL HAS NO EDITOR AFTER THRE MONTHS. Assembling the data is a substantial contribution, but disciplinary journals may not want dataset papers. Help??

Older posts