The gold standard for information about college graduation rates comes from IPEDS, the Integrated Postsecondary Education Data System. Here is the page for Williams. Unfortunately, I have trouble getting race-by-sex data from this source. (Pointers welcome!) Fortunately, the Education Trust places that data (ultimately, from IPEDS, I assume) in an easier to use format here. This is the source for the screen shot above.

The 6-year graduation rate for black males from Williams College is below 70%.

Comments and questions:

1) Holy disparate impact, Uncle Ephraim! I found this number to be shockingly low. Are you surprised? I can’t find another elite school with such a horrible record.

2) I have never been sure about the exact meaning of this statistic. These students don’t graduate from Williams College in 6 years. But don’t at least some of them graduate from somewhere? If so, where and at what rates?

3) Shouldn’t this be a front page story in The Record?

4) I believe that this is data as of school year 2011-2012. So it is referring to students who started in the fall of 2006, i.e., members of the class of 2010.

5) Fortunately, the data is previous years looks better for Williams: class of 2009 (87%), 2008 (96%), 2007 (88%), 2006 (88%). But why would things get so much worse for the class of 2010? And how are things looking for recent graduating classes?

6) I am suspicious of this data for two reasons.

a) Although misleading the Feds is a bad idea, elite colleges have been painting the best possible picture for years. I am especially suspicious of the fact that many other elite liberal arts colleges fail to report any data at all.

b) There are very few (approximately 25?) African-American males in each Williams class, so the difference between 70% and 90% is only 5 or so students. So, with luck, this is just random variation.

7) Prediction: The data next year will look better, closer to the long term average of 90%. It is unlikely that something important changed for the class of 2010 and subsequent classes. Instead, this is more likely random variation caused by a small sample size.

8) If this prediction is wrong — if extremely low graduation rates from black males continue — then it is likely that Williams changed its admissions policies in 2006, admitted weaker applicants, applicants that it used to reject. Such applicants are, obviously, much less likely to graduate from Williams in 6 years.

UPDATE: Thanks to Directory of Institutional Research Courtney Wade for explaining that 6b and 7 above are correct. This was random variation on a small number of students. Latest data shows the graduation rate back around 90%.

Print  •  Email