Columbia University disclosed Friday that it had reported faulty data on class size and faculty credentials to a publication that produces widely known college rankings — errors that it attributed to a reliance on “outdated and/or incorrect methodologies.”
With the acknowledgment of the embarrassing missteps, the Ivy League university in New York expressed contrition over problems that appeared to have been first identified by a whistleblowing math professor skeptical of how Columbia had come to be ranked second in the nation by U.S. News & World Report.
Ultimately, the university lost that ranking as it sought to answer questions the professor raised about data on class size, faculty and other matters that Columbia sent to U.S. News. Friday’s statement capped an internal review of the matter.
“We deeply regret the deficiencies in our prior reporting and are committed to doing better,” Columbia’s provost, Mary Boyce, said in the statement.
The statement provided some detail about the errors, but did not clarify exactly what went wrong. Columbia said it had overreported the number of undergraduate classes with less than 20 students and the share of full-time faculty with terminal degrees such as a PhD. But the university did not specify how much it overreported those figures.
U.S. News published last year a chart showing 83 percent of Columbia’s classes in fall 2020 had fewer than 20 students. It was the highest such share reported that year among the top 50 universities.
That is the kind of statistic likely to lure college-bound students and parents. It also was likely far off the mark.
Columbia confirmed Friday that the share of undergraduate classes in fall 2021 with under 20 students was 57 percent. The university also confirmed that the share of full-time faculty with terminal degrees was 95.4 percent.
The faculty whistleblower, Michael Thaddeus, has written that Columbia previously claimed the share with terminal degrees was 100 percent.
Also Friday, Columbia posted online for the first time statistical profiles related to its undergraduate programs in the format known as the Common Data Set. Until now, it had been the only Ivy League school that declined to take that step toward transparency.
Thaddeus said he was flabbergasted to see the data and was reviewing the information. He questioned other figures Columbia has provided over the years about class sizes.
“It’s hard to believe that this was an inadvertent error or a minor matter of methodology,” Thaddeus said. “More likely, someone in the University knew that there was serious misrepresentation afoot. If so, who was it?”
In July, after acknowledging questions about its data reporting, Columbia was removed from the No. 2 spot on the U.S. News national university list and placed in the peculiar position of being unranked. Earlier, the university said it would not submit data for the 2023 edition of the rankings, which will be announced Monday.
Columbia to skip U.S. News rankings after professor questioned data
Columbia had been a fixture in the U.S. News top 10. The last time it fell outside that exclusive club occurred in fall 2003, according to U.S. News chief data strategist Robert Morse, when it was listed 11th among national universities for the 2004 edition of the rankings.
It rose in recent years, from No. 4 in fall 2010 to No. 3 in fall 2018 and No. 2 last fall, tied with Harvard University and the Massachusetts Institute of Technology, in the 2022 published edition of the U.S. News rankings.
Columbia’s ascent on the prestige list drew skepticism from Thaddeus, a professor of mathematics, who published in February, with an update in March, a blistering critique of the university’s data reporting and the rankings themselves.
Thaddeus, curious about the numbers behind Columbia’s position, dove into the U.S. News ranking methodology and what he could glean from publicly available sources on the university’s faculty, enrollment, class sizes and finances. He found what he believed were troubling discrepancies between what the university was claiming on certain key measures and what might actually be the case.
The critique touched off the internal scrutiny at Columbia.
“The role played by Columbia itself in this drama is troubling and strange,” Thaddeus wrote in his critique. “In some ways its conduct seems typical of an elite institution with a strong interest in crafting a positive image from the data that it collects.”
Columbia is far from the only prominent college or university to make an embarrassing revelation of faulty data sent to U.S. News and others. Claremont McKenna College acknowledged in 2012 that it had misreported various admissions statistics. Emory University disclosed similar problems that year, and so did George Washington University. From time to time, other schools have acknowledged troubles with data used to rank undergraduate and graduate programs.