Scores and percentile question

Post Reply
User avatar
WontonBurritoMeals
Posts: 102
Joined: Sat Feb 23, 2008 9:43 pm

Scores and percentile question

Post by WontonBurritoMeals » Mon Nov 09, 2009 5:06 am

In the different PGRE Tests, the score doesn't necessarily correspond to a particular percentile. How the hell is that supposed to work? One year, a 90 percentile gets you a score of 990. The next year, someone has to be in the top 1 percentile. How did this come about?

May the wind be always at your back,
-WontonBurritoMeals

larry burns
Posts: 77
Joined: Wed Jul 02, 2008 1:12 pm

Re: Scores and percentile question

Post by larry burns » Sat Oct 23, 2010 3:13 am

I'm also curious to know. For example, today I got a 750 corresponding to 77% on the 8677 exam. 750 being too low for me, I was unhappy until I compared 77 percentile to the 0177, and noticed that its corresponds to 790-800, which is alot higher. Which matters more for grad schools? Your actual score or the percentile?

User avatar
quizivex
Posts: 1031
Joined: Tue Jan 09, 2007 6:13 am

Re: Scores and percentile question

Post by quizivex » Sat Oct 23, 2010 4:14 am

larry burns wrote:Which matters more for grad schools? Your actual score or the percentile?
In theory they're equivalent. Grad schools pick their favorite students among all the applicants. You're basically being compared to the rest of the pool as a whole. It doesn't matter whether they look at the percentile or the scaled score, the effect will be the same. :idea:

(Obviously there are minor exceptions if cutoffs of some sort are used, or there's a psychological tendency to view the difference between 890 and 900 as more substantial than the difference between the percentiles 87% and 88%.)

(BTW, Percentiles are "updated" annually so even if you took the test 5 years ago when percentiles were higher for a given score, the current percentile corresponding to your score will be on the report you send today.)

User avatar
HappyQuark
Posts: 762
Joined: Thu Apr 16, 2009 2:08 am

Re: Scores and percentile question

Post by HappyQuark » Sat Oct 23, 2010 4:30 am

quizivex wrote:
larry burns wrote:Which matters more for grad schools? Your actual score or the percentile?
In theory they're equivalent. Grad schools pick their favorite students among all the applicants. You're basically being compared to the rest of the pool as a whole. It doesn't matter whether they look at the percentile or the scaled score, the effect will be the same. :idea:

(Obviously there are minor exceptions if cutoffs of some sort are used, or there's a psychological tendency to view the difference between 890 and 900 as more substantial than the difference between the percentiles 87% and 88%.)

(BTW, Percentiles are "updated" annually so even if you took the test 5 years ago when percentiles were higher for a given score, the current percentile corresponding to your score will be on the report you send today.)
I was about to answer in exactly the same way but then I remembered that you can submit scores from previous years, if you've taken them that is. If I recall correctly, your scores are saved for up to 5 years after you've taken the test. With that said, there are probably very few people using older scores and the scores likely don't differ so significantly that it would ever influence there decision.

User avatar
quizivex
Posts: 1031
Joined: Tue Jan 09, 2007 6:13 am

Re: Scores and percentile question

Post by quizivex » Sat Oct 23, 2010 5:00 am

Actually, what I meant by my final comment is that the current year's percentiles are put on every score report that ETS mails... even if the test was taken years ago. The percentile the schools see today may be a little different from the one on the original report mailed to you. Only if you sent reports with the old test (the 4 "free" reports), and the schools were nice enough to keep them, could they be seeing the old percentile.

From experience, percentiles can change substantially over a short time... I took the general GRE early... summer after sophomore year. The writing came back 5.5 - 82%. I felt I wrote very well and since so many people taking the test aren't even native English speakers, I thought 82% seemed unfairly low. So I requested a rescore hoping to get a 6, and they sabotaged me to 4.5. Then, senior year when I was applying, some of the students on the profile thread who got 5.5 listed their percentile as a whopping 90%.

Also, a 990 PGRE was 97% for a while, including my year, but last year it was 95%.

User avatar
HappyQuark
Posts: 762
Joined: Thu Apr 16, 2009 2:08 am

Re: Scores and percentile question

Post by HappyQuark » Sat Oct 23, 2010 5:58 pm

quizivex wrote:Actually, what I meant by my final comment is that the current year's percentiles are put on every score report that ETS mails... even if the test was taken years ago. The percentile the schools see today may be a little different from the one on the original report mailed to you. Only if you sent reports with the old test (the 4 "free" reports), and the schools were nice enough to keep them, could they be seeing the old percentile.

From experience, percentiles can change substantially over a short time... I took the general GRE early... summer after sophomore year. The writing came back 5.5 - 82%. I felt I wrote very well and since so many people taking the test aren't even native English speakers, I thought 82% seemed unfairly low. So I requested a rescore hoping to get a 6, and they sabotaged me to 4.5. Then, senior year when I was applying, some of the students on the profile thread who got 5.5 listed their percentile as a whopping 90%.

Also, a 990 PGRE was 97% for a while, including my year, but last year it was 95%.
Ah, I see what you mean. I thought you were suggesting that because everyone is subject to the same scaling, a 700 (72%) will be lower than an 880 (91%) regardless of whether you are looking at the score or the percentage which is only true of scores taken from the same exam day.



Post Reply