GRE percentile change

[cooperstreet]

I took the GREs in 2009. I had a 750 V, 710 Q, and 5.5. In percentile terms, they were 99% V, 73% Q, and 93% AW.

However, after 2011 the percentiles were revised. A 710Q in 2011 was around 68%!!!! Will this be taken into account by adcoms? what does everyone think?

[iowaguy]

IMHO, your 710 will be compared against other peoples' current scores, so yes it will be taken into account. When schools publish the GRE scores of their incoming class, they publish the hard scores, not the percentiles.

According to ETS concordance tables, in 2012 (as of April 30), a 710Q is the same as a 155, which is 64th percentile. i.e. in 2012 (those applying now) 36% of test takers scored higher than you in Q. Those people are your current competition...

Edited December 7, 2012 by iowaguy

[cooperstreet]

That's infuriating. When I took the test my score was in the 73rd percentile. Now its in the 64th percentile, and nothing I did changed. bummer.

[ThisGuyRiteHere]

thats rough.

[cooperstreet]

Yeah, you're telling me. I feel like I'm automatically out of the running because my quant percentile was revised down nearly 10 percentage points

[TakeruK]

That's infuriating. When I took the test my score was in the 73rd percentile. Now its in the 64th percentile, and nothing I did changed. bummer.

I agreed that it's a huge bummer. My scores did the same, and it was a 6 months difference (I wrote the GRE just before the Revised GRE came out, so they published a new concordance table).

However, it's justified. Nothing YOU did change, but everyone else did. Percentiles are not about how well you scored (that's what scaled scores are for), they are about how well your score compares with everyone else's scores. ETS computes the percentiles based on the last 5 years of tests, so every year, the oldest test scores are thrown out, and percentiles are recomputed with the latest years' scores. If the scores required for a certain percentile has moved up, it means that the most recent year's test takers have scored higher than the test takes 6 years ago.

This is why I don't think percentile ranks are not really that meaningful in minute detail. They're useful to tell that a student in the 85th percentile range is better than say 60th percentile, but how can you really distinguish e.g. 73rd vs. 70th. In addition, almost every school that has published a "cutoff" publishes a scaled score (e.g out of 800 or out of 170), not a percentile. And it's important to remember that the percentile is your ranking amongst ALL GRE takers. You are not competing against everyone who wrote the GRE for a grad school spot -- you are just competing with the other people who applied to the same program! So it's likely that the department will instead compare your scaled scores with one another.

[Penelope Higgins]

Nobody in political science makes decisions based on GRE scores. They are used as a simple way to weed out files that won't be considered. And scores above 700 on the old test will be enough to prevent your file from getting weeded out at every department. After that, decisions are made based on a more careful look at the other components of your application.

In other words, take a deep breath. There's much less to worry about here than it seems.

[cooperstreet]

Josefmoore,

Could you provide specifics? Thanks.

[jacib]

I don't know about minimums, but I think they can be inferred (to an extent) from the averages. SwissChocolate made a really good list of the Looking at those, 700Q is on the low side at top programs. Considering that there are several perfect scores admitted at some of the top top programs admit, it's probably not unimaginably low to be averaged in ([155+170]/2=162.5 or about average), but it is distinctly below average at Columbia-Duke-Madison-Berkeley-NYU. However, OP did of course have an above average V score (169 by the new system, near perfect). From the wording that SwissChocolate quoted, it looks like the two scores were considered together at some places so, at least at those places, OP is probably only slightly below average and therefore definitely above any strict cut-off.