Opinion: An idea for the college affirmative action problem

By MAURICE REGAN

Published: 07-08-2023 6:00 AM

Maurice Regan lives in Pembroke.

That famous, inventive psychologist B.F. Skinner once described a crackpot idea that he could train pigeons to be missile guidance systems for use in WWII. This idea led to the Pelican Project and the development of a guidance system that actually worked. There was only one problem. The government had not developed a missile for this guidance system. But an amusing paper ensued, Pigeons in a Pelican.

I have had my share of crackpot ideas. I implemented a local jail suicide prevention program that actually worked for 10 years until dismantled by inept, unthinking managers. Widely applied, Jeffrey Epstein might still be alive and many conspiracy theories would be dead.

While on the Transportation Committee of the NH House, I suggested a seat belt law that seemed to have both conservative and liberal support but never made it into the legislative process as I lost the subsequent election. As part of a classroom lecture, I proposed a program to decrease misbehavior in correctional settings that probably saves money. This should remain untested as no jail is ever going to make me warden.

So here is my crackpot idea for solving the reoccurring problems of affirmative action in college admissions.

Every college should know which students profit from their courses of study. They have predictive data, including SAT/ACT scores and high school grades from rigorous courses. Colleges also have longitudinal data from the four years of undergraduate education so they can identify their predictive successes. The problem is the setting “cut off” scores.

The very “selective” colleges have a very high cut off and few applicants are admitted. However, many of the rejected applicants, if accepted, might do very well. We know this because “legacy” admissions are a category with a wider range of predictive scores and within this category, the success rate is probably the same. There are outliers. Freshman Bill Gates was admitted to Harvard with astronomical scores, dropped out but politely returned years later to give the commencement speech.

So colleges should endeavor to categorize students who will profit from their course of study, not cut out students who will be on the wrong side of an unnecessarily high cutting score. The students in this likely-to-be-successful category go into a lottery and luck fills the admission slots. The diversity and color blindness problems are solved by the random sample and the Supreme Court has more time on their hands. A college might even use this system to fill slots vacated over the four academic years. See Mr. Gates above.

Article continues after...

Yesterday's Most Read Articles

Shamir Darjee immigrated to Concord knowing no English. Now the 20-year-old just bought his family a house.
Opinion: Let’s keep our forests as forests
‘If it’s about us, it needs to include us’: As Concord weighs solutions to homelessness, people currently unhoused want a voice at the table
‘Woefully unprepared’ hiker refused to leave Sno-Cat atop Mt. Washington
‘Friends for life’: Concord woman becomes Best Buddies champion alongside high school buddy
Capital Vintage brings new affordable fits to Concord and soon, the Seacoast

Like the seat belt law I mentioned above, this should have both liberal and conservative support. For the liberals, campuses would be diversely populated by students who are very likely to do well. No liberal wants a protected group member to be selected into a college where they will most likely fail while amassing debt. For the conservatives, this is clearly color blind, as a random selection has no biases. And admitted students and casual observers are not wandering the campus speculating that a protected group member is actually unqualified for admission or that one group squeezed out qualified students. All were randomly selected from the qualified category.

Like the pigeons in a pelican, this is the guidance system. The missile of academia may be unprepared.

]]>