changes in subject test percentiles over the last 15 years

Forum for the GRE subject test in mathematics.
Post Reply
fluentmundo
Posts: 10
Joined: Tue Nov 10, 2015 2:19 pm

changes in subject test percentiles over the last 15 years

Post by fluentmundo » Mon May 16, 2016 9:06 pm

Hi all,

I thought it would be interesting to see how the subject test distribution has moved over time. The test was rescaled in October 2001 (according to this), so there's almost fifteen years of data. I've compiled what I've found here:

Image

The data comes from ETS's Guide to the Use of GRE Scores, released each year. I haven't been able to find the percentile tables for a few years, but there is a definite trend: more people are scoring higher. In some cases the percentile rank for a certain scaled score has dropped up to 10 or 11 points since 2001-2004. (Scores now in the mid-80s used to be in the mid-90s.) The mean has also increased quite significantly, and the number of tests has almost doubled. I'm sure that is because a lot more people are taking the test today than took it fifteen years ago, but I also wonder whether some of the increase in the number of tests is due to an increase in the number of people who are taking the test more than once.

Curious to hear what people think about this trend -- what might explain it, and what significance it could have for graduate admissions.

fierydemon
Posts: 39
Joined: Sat Sep 14, 2013 11:57 pm

Re: changes in subject test percentiles over the last 15 years

Post by fierydemon » Fri May 20, 2016 1:29 am

Hey! Would you like to disclose your score?

arima
Posts: 36
Joined: Thu Mar 12, 2015 9:25 pm

Re: changes in subject test percentiles over the last 15 years

Post by arima » Sat May 21, 2016 5:53 pm

The dramatic increase in test takers is primarily due to more internationals taking the test. Many internationals are strong in terms of standardized testing which has caused this re-adjustment. The correlation is far from perfect between standardized GRE math subject test taking and core math ability. I see it all the time with students having amazing GRE math subject test scores but are very suspect in the classroom with homework details and course examinations.

Though the test score is a factor, it is fortunate to see some top departments have flex and see other factors. For example, there are a few at Berkeley (past and entering) who have test scores in the 60% range. Their admittance to Berkeley are on more important factors like strength of UG math institution, math gpa, grad courses taken, research undertaken, letters, etc.

Yes, try to score well on the exam but folks need to see beyond the scores and not be so obsessed with silly comments "I got a 85% on the exam...so disappointed...what I am to do with my life?"

fluentmundo
Posts: 10
Joined: Tue Nov 10, 2015 2:19 pm

Re: changes in subject test percentiles over the last 15 years

Post by fluentmundo » Thu Jun 02, 2016 1:33 am

arima wrote:For example, there are a few at Berkeley (past and entering) who have test scores in the 60% range. Their admittance to Berkeley are on more important factors like strength of UG math institution, math gpa, grad courses taken, research undertaken, letters, etc.
Berkeley (for pure math) states explicitly that it expects at or above an 80th percentile.

arima
Posts: 36
Joined: Thu Mar 12, 2015 9:25 pm

Re: changes in subject test percentiles over the last 15 years

Post by arima » Thu Jun 02, 2016 2:40 pm

Expectations versus actuality are different. Expectations is a general statement. Actuality is a case by case assessment.

Search the forum and there is one poster of this year who states her GRE is in 50-60% range and she got accepted. I know of another person also with the same GRE score at Berkeley. These folks (pure math) had enough compensating factors.



Post Reply