Big Change in Percentile Rank

[groverj3]

I'm applying to several schools but noticed a large change in my percetile rank scores when I looked on the ETS website.

For some background. I took the GRE back in 2010, before the format changed. After working for a year since graduating in 2011 I decided to finally apply. Since my scores are recent enough they should be acceptable to use. However, I needed to send them to more places than I originally did after I took the test and noticed that my percentile rank scores have changed signifcantly since then. My verbal stayed relatively the same, but my Quantitative dropped from 65% to 53% (raw score of 650), and my Writing rose from 84% to 92% (raw score of 5.0).

Personally, I think comparing scores across the old and new test is kind of silly to begin with. However, I'm not ETS. My Quantitative score wasn't stellar, but now I'm pretty mad that it was lowered by over 10%.

I was just going off of my original score report when filling out applications, but now that I've noticed a big change (let's be honest, if it was only a few % I probably wouldn't care. Over 10% is a big deal though) should I be concerned about the % rank scores being that much different from what will be in score reports from ETS? The grad admissions offices have generally been less than helpful in answering this question. They either don't understand what I'm asking or tell me to include both scores which doesn't make a lot of sense to me when the raw scores do not change, only the percentiles.

[Eigen]

When you say that your percentile scores have changed, what exactly are you basing that on?

From my understanding, your percentile scores were set from when you took the test, since they were relative to other people taking the test at the same rough time.

Did you get a new score report from ETS that's making you think your scores have changed, or just looking at score to percentile conversion tables on the ETS website?

[pemdas]

percentiles of the old format gre cannot be shifted anymore

new, revised gre percentiles were reviewed in July and some deflation took place with the quan percentiles

[emmm]

I think the OP is correct and the percentiles do change. It is silly, in my opinion, to compare students at different times and on different tests, but it seems to be what ETS is doing.

[groverj3]

When you say that your percentile scores have changed, what exactly are you basing that on?

From my understanding, your percentile scores were set from when you took the test, since they were relative to other people taking the test at the same rough time.

Did you get a new score report from ETS that's making you think your scores have changed, or just looking at score to percentile conversion tables on the ETS website?

I looked at the scores on the ETS website due to needing to send them to more schools. The percentile scores on there are not the same as the report I was mailed after taking the test in 2010. The raw scores are obviously the same but ETS continually updates your % Below rank based on how people do on the test after you take it.

percentiles of the old format gre cannot be shifted anymore

new, revised gre percentiles were reviewed in July and some deflation took place with the quan percentiles

Well, it's possible my scores shifted a while ago and I wouldn't have noticed until now. However, I haven't read anywhere that they can't shift percentiles anymore. If they can convert your score to a score on a test you never took (the new GRE, post 2010) I see no reason why ETS wouldn't decide to do something equally ridiculous by changing percentiles based on comparing you to people taking the new version.

Deflation also wouldn't explain why my writing percentile went up 10%.

I think the OP is correct and the percentiles do change. It is silly, in my opinion, to compare students at different times and on different tests, but it seems to be what ETS is doing.

I somewhat understand comparing people to each other as long as they take the same version of the test. However, if my scores went down simply because people are doing better now on a test I didn't take I have problem with that kind of logic.

[groverj3]

I guess my main question here is: If I already submitted some apps self reporting the percentile ranks on my score report from 2010 will admissions people think I'm falsifying my record when they receive scores directly from ETS that are different? I'd rather not have people throwing out my applications because they think I'm lying.

Edited September 21, 2012 by groverj3

[Guest Gnome Chomsky]

Wouldn't that just mean people are doing better now than they were 2 years ago (when you took it)? I understand the scores on math and verbal have to be translated from a 200-800 scale to a 130-170 scale, but the writing scale is still 1-6. I think the changes in percentile are because they are comparing you to the average scores of this year. The average scores of the year you took it are no longer relevant. But I think you might need to notify that one school and tell them that there was a change in your percentile so that they don't think you were trying to pull a fast one on them.

[groverj3]

I've already asked two admissions offices about this and they seem not to understand what I'm talking about. I suppose I'm just going to have to call them.

ETS... making lives difficult, as always .

[mtlve]

I took the old format over five years ago and the new format recently. My total score went up 100 (using their published conversion to old format), and my individual verbal and quant both went up at least 10%. This is a completely different test now, and it may be more accurate of a judge of who will do well in graduate school (analogies less of an indicator than the reading comprehension focus now in verbal for example). I too strongly doubt the consistency of the current gre though, but I really do not understand their scoring system. My writing also changed very drastically between tests.

Time will tell what happens with this new test lol

Edited October 7, 2012 by mtlve

[emmm]

I can't imagine that schools are unaware of the changeable nature of the percentiles -- it is not a new issue. It is also not just unique to you, so don't worry about it. As long as you report your raw score accurately, you should be ok.

[hopeforbest]

Original percentiles (July 2011 test date) (sent with original scores): V 93%; Q 77%; W 67%

New percentiles (from Concordance Tables & now on my ETS site): V 91%; Q 71%; W 73%

Interestingly, the Interpretive Data Chart in the ETS Guide to the Use of Scores 2010-11, for the original percentiles indicated that they were "Based on the performance of all examinees who tested between July 1, 2006, and June 30, 2009."

The Concordance Tables for the prior scale/new scale and percentile rank on which the revised percentiles are based, indicates that the percentiles are "Based on the performance of all examinees who tested between August 1, 2011 and April 30, 2012. percentile ranks will be updated in July 2013."

So - instead of being "compared" with a 3-year sample for percentile purposes, it appears that the "old" scores are now being compared to an 8-month sample of new test takers - this simply does not seem statistically sound!

[groverj3]

That's because it isn't statistically or logically sound.

We'll see how this impacts me getting interviews and hopefully being accepted somewhere. I could see it being a big negative on my application.