Diversity in the First-year Class at the Ivy Plus Institutions

Diversity in the First-year Class at the Ivy Plus Institutions

 I’m not sure where to begin on this one, so let’s veer off topic a bit.  

I’ve decided I’ll likely be phasing out Higher Ed Data Stories in the near future as I go into retirement and start my new venture, which is soft launched but not officially open for business.  When I do, I’ll be posting regularly on my blog over there, but won’t be putting everything out on the web for free, as I’ve been doing on this site.  I do appreciate the contributions people made on the Buy Me Coffee site, but the hosting, software, and labor costs never balanced with the revenue, and while there was a lot of good will that came from my work, I was still in a deficit situation (especially on the time part) and I’ll need to dedicate that to the business side of things.  Medicare Parts B and D ain’t free, you know.

But this is some unfinished business, and it might be a good place to end.  You know I’ve been personally opposed to the very idea of the SAT and ACT for some time, while being professionally neutral: If colleges find value in it, I don’t care if they use one, the other, or both.

But I do care about the truth.  On that note, two issues: The headlines suggesting that lots of colleges are returning to standardized tests for first-year admissions are just not true, of course, and everyone in the business knows this.  The testing agencies are curiously silent on the misinterpretation of this information, of course.

The larger issue of “truth” is the justification put forth by the universities that are returning to the SAT or ACT.  They are all suggesting that they need the tests to find qualified students of color, or low-income students.  Is that true?  If it is, does it mean they denied admission to other, more highly qualified students of color with test scores? You can look at the data below, and while it’s not absolutely definitive, it is interesting.

Before diving in, however, some caveats:

  • IPEDS reporting recognizes “two or more” as an ethnic category, but does not allow breakouts.  So many colleges will report some percentage of students in every category they check, and of course, there is good reason to do so.  There is no reason, however, to increase the numerator and not the denominator in the equation, as some of them do.  So you may notice that the numbers here don’t line up with what colleges have published.
  • IPEDS data on income or financial need is far less clear, as it only breaks out by Pell/Non-Pell.  Perhaps the researchers who have access to the unit record data can dive in more deeply.
  • We don’t have a lot data (at least not published as supporting evidence for the claim) that says there is a problem with performance among the students admitted without tests.  If that comes to light later, it might change your perception of this data, as it should.  What I have seen shows only minor differences, and given COVID and its disproportionate effects on students, I’m not sure the SAT would survive other testing.
  • Some of these charts show Simpson’s diversity, which is a different way of thinking about diversity.  It’s not the percentage of minority students; it is essentially the chance that two randomly selected members of a group will be different.  If your population was 100 and all 100 in the group were different, you’d have perfect diversity (a value of 1).  If all 100 were the same, you’d have a value of 0.  Higher numbers indicate greater diversity.

OK.  Got it?

There are four views in the visualization.  The first shows just Hispanic and Black/African-American enrollment in the first-year classes at the “Ivy Plus” institutions (The Ivy League institutions plus Duke, MIT, Stanford, and the University of Chicago.)  You can see the trend (in both numbers and percentage of the class) over time.  The denominator is the entire class. The blue bars show data up until 2020, and the purple bars show test optional years.

The second shows the entire ethnic composition of the domestic students in the class.  Look at them collectively to start, then look at individual institutions using the control at right. 

The last two views show the Simpson’s Index of Diversity for each institution over time.  The first is for domestic students, and the second is for everyone, including international students counted as an ethnicity.  Use the highlight control to focus on one institution.

So, what do you think?  Do highly rejectives need the SAT to find students of color?  Let me know.

Source link