Archive for January 27th, 2010



January 27, 2010

Andrew Revkin has this cute video up on his blog:


GRE Hate

January 27, 2010

If I may indulge myself here a bit. I’m reading our graduate applications at the moment. We have lots and lots of them… by which I mean, lots and lots. When they come in, we sort candidates by several categorization standards, including their GPA, their last name, their stated area of interest, and their GRE score.

I’m sure I’m not alone in this, but so far as I can tell, the GRE is a near meaningless measurement of a candidate’s potential. I fail to see why we (academics) place so much weight on it.

Oh, sure. Since the GRE score is the only standardized metric between all candidates, it offers a compelling measurement against which to compare. X person got 720/760/5; Y person got 680/720/5.5; Z person got 500/620/4. (These are made up numbers.)

Seems like a no-brainer. Z person doesn’t make the cut. It’s a battle between X and Y.

And yet, there’s considerable vagueness in the scores. When one looks at the statement of purpose, or at the letters of recommendation, it frequently turns out that X and Y are boring drones, where Z is an interesting and motivated prospective, maybe with a serious and well-conceived project, or maybe even in some cases with a significant history of professional accomplishment. The “bad GRE effect” is amplified in candidates who haven’t been in the educational system for a while, perhaps because they’ve been in the business world doing something that doesn’t demand that they calculate the volume of a sphere.

Moreover, it’s not entirely clear how standardized they in fact are. I’ve heard from more than one candidate with widely discrepant scores between two contiguous test sessions. To my mind, if there’s a discrepancy of several hundred points between a set of scores from the same person, this test doesn’t have much to offer in the way of measurement.

So what, exactly, are the GREs in place to measure? And why do we (academics) continue to rely on the GREs as strongly as we do? Why do we (academics) sing the praises of our departmental units when we have a high GRE average, all the while admitting to one another that the GREs are unreliable?

I’m very much persuaded that the GRE ought not to play a strong role in admissions decisions at all, and at best should serve only as a vague data-point or an extra accomplishment signalling positively in a person’s favor, but not negatively in a person’s disfavor. Others with whom I’ve spoken disagree. They think the GRE can serve as a workable cutoff.

What do you think?

(With my luck, I’ll get sued by ETS for disparaging the GRE, but I think this is a legitimate academic concern. I suspect that they disclaim the value of the GRE as the sole measure of a candidate’s fit; but I similarly suspect that they think it’s a valuable measure of something. Just what it’s a valuable measure of is beyond me.)