New GRE Percentiles

[ay761]

I have a question about the new percentiles. According to the ETS website, my math score (149) would have been in the 50th percentile just last month. But now it is in the 39th percentile because I took the exam 2 weeks too late, I guess. Math isn't that important to my program... but still, scoring in the 50th percentile at least would have been nice. I don't really understand how this is fair, considering I will be applying along with people who took the exam before July and might have a lower score but a higher percentile than me. For verbal it makes a difference too, though not as much... the 85th percentile compared to the 80th percentile. Maybe I am missing something here (hence why I did so poorly on the math section!).

So my question is... will people who got the exact same score as me, but took the exam a month or two ago, have a different percentile than me listed on their official score reports... ?!

[ANDS!]

How is it unfair? It's like taking an exam and scoring high, but having a low percentile, and then your buddy takes the same exam, on the same scale, but scores lower percentile wise. The exam is the same, it's simply who you are scored against that is different.

[pemdas]

I have a question about the new percentiles. According to the ETS website, my math score (149) would have been in the 50th percentile just last month. But now it is in the 39th percentile because I took the exam 2 weeks too late, I guess. Math isn't that important to my program... but still, scoring in the 50th percentile at least would have been nice. I don't really understand how this is fair, considering I will be applying along with people who took the exam before July and might have a lower score but a higher percentile than me. For verbal it makes a difference too, though not as much... the 85th percentile compared to the 80th percentile. Maybe I am missing something here (hence why I did so poorly on the math section!).

So my question is... will people who got the exact same score as me, but took the exam a month or two ago, have a different percentile than me listed on their official score reports... ?!

don't worry as soon as new percentiles are effected the scores of those who took exam prior to rescaling of percentiles will be ammended automatically too. That is the schools will receive new reports indicating score-percentile adjustments and the score reports online will be updated with new percentiles too. If i were you, i would not worry about my percentiles against those who tested earlier but would think that the percentile obtained by you is below 50% anyways (!) - don't be so jealous and retake the test.

Edited July 29, 2012 by pemdas

[BrianM]

How is it unfair? It's like taking an exam and scoring high, but having a low percentile, and then your buddy takes the same exam, on the same scale, but scores lower percentile wise. The exam is the same, it's simply who you are scored against that is different.

It is unfair, and the whole process is too.

My % rank show down almost 20% because of this stupid adjustment. Good job ETS.

[vonLipwig]

From the following webpage: http://www.ets.org/gre/revised_general/scores/understand/

Regardless of when the reported scores were earned, the percentile ranks for General Test and Subject Test scores are based on the scores of all examinees who tested within a recent time period.

From this, it seems that equal scores would be given equal ranks, but that rank might shift as more people take the test.

[wine in coffee cups]

don't worry as soon as new percentiles are effected the scores of those who took exam prior to rescaling of percentiles will be ammended automatically too. That is the schools will receive new reports indicating score-percentile adjustments and the score reports online will be updated with new percentiles too. If i were you, i would not worry about my percentiles against those who tested earlier but would think that the percentile obtained by you is below 50% anyways (!) - don't be so jealous and retake the test.

This is correct. The percentile you see on the original paper copy of your scores is not necessarily the percentile that you'll see if you check your scores online right now or the percentile reported when you send scores to schools. Percentile adjustments are applied broadly, not just to a group of very recent test takers. People who took the test a little while ago and had an acceptable percentile at the time might be in for a nasty surprise.

My subject test result was adjusted down by 5 percentiles in the four years since I took it, and my analytical writing was adjusted up by 5 percentiles over the course of a few months last year. In both cases, the percentile shown online (and what I assume was sent to schools) is not what was on the original paper report, though the scores remained the same.

[villina]

The resulting percentile of your test may be even higher than that was two months ago, depending on how next people will pass their tests.

[1Q84]

This is correct. The percentile you see on the original paper copy of your scores is not necessarily the percentile that you'll see if you check your scores online right now or the percentile reported when you send scores to schools. Percentile adjustments are applied broadly, not just to a group of very recent test takers. People who took the test a little while ago and had an acceptable percentile at the time might be in for a nasty surprise.

My subject test result was adjusted down by 5 percentiles in the four years since I took it, and my analytical writing was adjusted up by 5 percentiles over the course of a few months last year. In both cases, the percentile shown online (and what I assume was sent to schools) is not what was on the original paper report, though the scores remained the same.

I'm guessing the percentiles throughout the years are not shown, merely the latest rankings?

[jmbrown88]

I know that percentiles have changed quit a bit over the years, where a 680 (in the quant) was in the 68th percentile and now a 720 (in the quant) is 68th percentile, but does score count for anything? I always felt like scores in the 700's were good, but 68th percentile just seems low to me...this whole GRE business is a mind f**k if you ask me.

[smc8339]

Overall, what are admissions comittees looking at? Your score or the percentile that you're in? From what I've seen, the few schools that actually list a recommended overall GRE score don't mention percentiles. I would hope that they just take the raw score, and don't really pay attention to the percentile - I think the percentile just gives one perspective, something to quantify how they did.

[pemdas]

Overall, what are admissions comittees looking at? Your score or the percentile that you're in? From what I've seen, the few schools that actually list a recommended overall GRE score don't mention percentiles. I would hope that they just take the raw score, and don't really pay attention to the percentile - I think the percentile just gives one perspective, something to quantify how they did.

that's true about cut-offs, but when it comes to comparing two absolutely matching candidates against each other - GRE, Ibt-Toefl, Acads + research/work experience - percentiles could be well gamed. The candidates with higher percentile scales will be favored by AdComs over those with lower percentile ones.

Some more facts: In 2011 according to ETS data the GRE General test takers' pool reached 800,000 people. Say roughly 15% people score at high percentiles either in quant or verbal or the combined (85% ranking, Q 162 or V 161, or combined). It makes the competition amongst some 120,000 people intense for some 22,500 places (rated schools 1-100) for each cohort accordingly - Masters and PhDs. Of course those competing for Masters will face more intense competition, because not every one looks after PhDs. Again this will be off-set by the number of places for Masters and PhDs available, since the latter would require a university funding and would be limited. Thus, 5 people per 1 admission place of course after all other criteria will be met, i.e. GRE, Acads and research/work experience

[kaykaykay]

that's true about cut-offs, but when it comes to comparing two absolutely matching candidates against each other - GRE, Ibt-Toefl, Acads + research/work experience - percentiles could be well gamed. The candidates with higher percentile scales will be favored by AdComs over those with lower percentile ones.

No, just no. Adcoms are university professors. They are smart. They compare apples with apples (points with points). I have never heard anyone(professors) referring to percentages in my field. Moreover, they will compare the points to the points in your field (and the other applicants to the department) and not a hypothetical overall plane. I was told that my verbal score was low, even though I had over 90%. When I mentioned the percentage they shrugged their soulders. I was very offended at first but it was how they read the points..

[pemdas]

No, just no. Adcoms are university professors. They are smart. They compare apples with apples (points with points). I have never heard anyone(professors) referring to percentages in my field. Moreover, they will compare the points to the points in your field (and the other applicants to the department) and not a hypothetical overall plane. I was told that my verbal score was low, even though I had over 90%. When I mentioned the percentage they shrugged their soulders. I was very offended at first but it was how they read the points..

Which field you are major in?

Point to point comparison takes one to the percentile comparison. The latter are always adjusted to their updated ranks. Say if you scored 161 in verbal and that would mean 89% before April 01 2012, after that date you will see 86% rank in your score report. This is given all scores stay as they come initially but percentiles are changed. Now tell me what would be the sense for such a smart professor from your school to look at ostensible data available from ETS score report when there's much more relevant percentile rank data available for their meticulous review

coming back to my question again, don't take an umbrage for my question

if you are humanities major, the worse you stand at this time, as not all said is literally meant - don't you look after the word implications any, the scores may imply percentiles too and not only the points

if you are tech guy, then you will know percentiles are not percentage

Edited August 5, 2012 by pemdas

[TheFez]

LOL... These GRE discussions attract some goofy discussions.

OF COURSE what matters are percentiles - since you are trying to find the best candidates.

AND OF COURSE percentiles are directly related to the numeric score - that's why ETS has been struggling to fine tune the percentiles associated with different scores. In fact the old 800 point system as revised, in part, to provide more granularity to the high end.

And people should quit bitching about the GRE. It's a reality. And it's actually a useful tools since without some standardized test of some sort you cannot compare "apples to apples". But a score is just another expression of a percentile. Most people who complain about the GRE are not those with high GRE scores.

And - what many students don't want to face is that a good GRE score can be achieved with some effort AND THAT is part of what the score reflects to a smart Ad Comm. A student who doesn't have top ability and doesn't practice or study for the GRE will not do well. But even a limited ability student who applies themselves can score in the 80+ percentile. SO the GRE is a signal of how well a student prepares.

[alf10087]

LOL... These GRE discussions attract some goofy discussions.

OF COURSE what matters are percentiles - since you are trying to find the best candidates.

AND OF COURSE percentiles are directly related to the numeric score - that's why ETS has been struggling to fine tune the percentiles associated with different scores. In fact the old 800 point system as revised, in part, to provide more granularity to the high end.

And people should quit bitching about the GRE. It's a reality. And it's actually a useful tools since without some standardized test of some sort you cannot compare "apples to apples". But a score is just another expression of a percentile. Most people who complain about the GRE are not those with high GRE scores.

And - what many students don't want to face is that a good GRE score can be achieved with some effort AND THAT is part of what the score reflects to a smart Ad Comm. A student who doesn't have top ability and doesn't practice or study for the GRE will not do well. But even a limited ability student who applies themselves can score in the 80+ percentile. SO the GRE is a signal of how well a student prepares.

Agreed. Its not just a matter of how well you are in math or verbal reasoning, the GRE also measures the ability a person has to adapt and excel at any given thing, even if it's not his field of study, language or type of test he is good at.

[pemdas]

I am glad forum readers are not deluded with "score" mania, but then there are many tyros who come from countries in which non-standarized tests are attypical. So mine was attempt to elucidate the situation at this time

(tried to add use vocab-wise )

Edited August 6, 2012 by pemdas

[kaykaykay]

"

Which field you are major in?

if you are tech guy, then you will know percentiles are not percentage "

Which field you are major in?

if you are tech guy, then you will know percentiles are not percentage

I am not a guy thank you. And my field is actually under my user name. The reason for looking at points and not percentiles is very easy: percentiles change and they are not really relevant as the adcoms want to find the best applicants from the application pool they have, not from the entire world.

They will not compare two 161 points based on the percentiles, as the points are the same, they mean the same only the testing pool has changed. To give you an easy(?) comparison you cannot say that a train is faster than the one you saw yesterday because you see it when you are standing and not from a car as yesterday.

Anyway, don't believe me if you do not want but as you can see I am already attending a program and I learned some things about the selection process after I got in.

[1Q84]

A student who doesn't have top ability and doesn't practice or study for the GRE will not do well. But even a limited ability student who applies themselves can score in the 80+ percentile. SO the GRE is a signal of how well a student prepares.

Yes. I had never taken a standardised test before studying for the GRE. I've found it has focused me in a lot of ways and brought out a lot of positive academic attributes in me. Are there faults with it? Yes, of course, and it will be that way with any standardised test on earth.

I am glad forum readers are not deluded with "score" mania, but then there are many tyros who come from countries in which non-standarized tests are attypical. So mine was attempt to elucidate the situation at this time

(tried to add use vocab-wise )

When I saw "tyro", I knew you were studying your GRE vocab lists... and I have a feeling GRE troll is trolling.

I am not a guy thank you. And my field is actually under my user name. The reason for looking at points and not percentiles is very easy: percentiles change and they are not really relevant as the adcoms want to find the best applicants from the application pool they have, not from the entire world.

They will not compare two 161 points based on the percentiles, as the points are the same, they mean the same only the testing pool has changed. To give you an easy(?) comparison you cannot say that a train is faster than the one you saw yesterday because you see it when you are standing and not from a car as yesterday.

Anyway, don't believe me if you do not want but as you can see I am already attending a program and I learned some things about the selection process after I got in.

Yeah, I don't really see how this is up for debate, anyhow. It most definitely is going to be a major factor to some and not to others. Trying to make definitive statements about adcomms is just a fool's errand.