In February, Michael Thaddeus, a professor of mathematics at Columbia University, published a lengthy article accusing the university of falsifying data in the U.S. News rankings. Columbia's title of "No. 2 in the nation" was shattered in an instant.
Thaddeus investigated the data submitted by Columbia and questioned the accuracy of several self-reported figures, including the number of small undergraduate classes, the percentage of faculty with PhDs, the student-to-faculty ratio, and the amount of money Columbia spends on teaching.
Initially, Columbia was defending its data. But on June 30, the day before the deadline for submitting new data to U.S. News, the university suddenly changed its tune. Provost Mary Boyce said in a statement that the data would not be submitted this year. The statement noted that the university had "begun a review of the data collection and submission process".
As a result of the data falsification controversy, Columbia University eventually withdrew from the next round of the U.S. News annual ranking of the best universities.
According to the data, approximately 60,000 students applied to Columbia in 2022, of which only 2,253, or 3.73%, were accepted. A lot of that competition has to do with its ranking. According to the 2022 U.S. News National University Rankings and the QS World University Rankings, Columbia is ranked 2nd and 19th respectively.
At present, there are four credible world university rankings, namely QS World University Rankings, THE Times Higher Education World University Rankings, U.S. News World University Rankings and Soft Science World University Academic Rankings. For US institutions, U.S. News is the most informative one, followed by QS.
The focus of these rankings also varies. U.S. News, for example, ranks U.S. colleges and universities on nine major criteria: graduation and non-transfer rates (22%), undergraduate academic reputation (20%), faculty (20%), graduation rate performance (8%), social mobility (5%), financial resources per student (10%), student selectivity (7%), alumni giving rates (3%), and graduate debt (5%).
Except for the item on undergraduate academic reputation, which is based on peer assessment, the remaining 80% of the data was collected and provided by the participating universities themselves.
One of the faculty items includes a measure of undergraduate class size (8%). In the data provided by Columbia last year, the university had 82.5% of undergraduate classes of less than 20 students and only 8.9% of classes of 50 or more students. At this point, Columbia essentially beat all of its top 100 competitors.
In his challenge, Thaddeus points out that the percentage of classes of less than 20 students in Columbia's undergraduate programme should lie between 62.7 and 66.9 per cent, far short of Columbia's stated 82.5 per cent. Similarly, the percentage of classes with more than 50 students should lie between 10.6 and 12.4 per cent. If these figures are true, then Columbia's class sizes are not small compared to other schools.
Furthermore, the 2019-2021 figures suggest that class sizes at Columbia are steadily increasing and that the university is considering a significant expansion of its undergraduate student body.