04 October 2017

I come to bury the GRE, not to praise it

I’ve seen a few graduate programs announce that they are not going to require students submit GRE scores any more. These announcements are widely met with praise. The GRE has minimal predictive value in long term grad school success, and it is biased against a lot of groups. And the costs stops a lot of people from applying to grad school.

Interestingly, at the start of last year, the dean of our graduate college announced that several programs were being required to add the GRE to their admission requirements. This was imposed on at from outside the institution at the state level. I can’t remember if it was UT System or the THECB.

Full disclosure. When I became the graduate program coordinator of our master’s program, I pushed and got our department to start requiring the GRE. My rationale at the time was that this was the “industry standard.” We wanted our students to go into doctoral programs, and we reasoned that we would be helping students pave the way for doctoral work by having them do it sooner rather than later.

Also, I was reacting to students who would come in the day before classes started and say, “Can I be a grad student?” At the time, there was no application deadline. And students who did that tended not to persist in the program. So requiring the GRE forced students to plan ahead, not go to grad school because there was nothing good on television that day.

I have since come around to see the many problems with the GRE. But I don’t think our department would be allowed to get rid of it, seeing how many departments were forced to require it.

But this is something I think about.

The GRE tried to solve a couple of problems. It failed to solve them, but those problems still exist. And I don’t know how to solve them. The problems are:

  • Grading policies vary wildly across institutions. (See this blog post.)
  • People interpret the same grades in different ways depending on the institution’s perceived rigour and prestige. (See this blog post.)
  • Recommendation letters are usually uniformly glowing.
  • People tend to trust recommendations “in network” from people they know either personally or by reputation.
  • The recommendation letter requirement reinforces power dynamics that leave early career researcher at the mercy of bullies and other poor supervisors. (Added 3 December 2018)

Students from famous universities who have rubbed shoulders with famous professors and can convince them to send a form letter get deep advantages in grad school acceptance. In other words, we end up selecting for students for grad school who already have a lot of “social capital.” If we want to diversify science, this is not the way to go about it. Diverse students come from diverse institutions, as Terry McGlynn has noted.

In theory, the GRE could have acted as a leveler for the playing field. It didn’t. But the problem it could have tackled is one that we still need to tackle. What can help level the playing field for students against “prestige”?

Related posts

What grades should look like
The “Texas transcript” is a good idea, but won’t solve grade inflation

External links
Students, Rejoice — Standardized Testing May Soon Be Dead
Nine types of admissions bias

No comments: