Score variation on test day and doing practice exams

This topic has expert replies
Junior | Next Rank: 30 Posts
Posts: 23
Joined: Thu Feb 07, 2013 2:51 am
Thanked: 1 times
Hello again,

I want to know while I'm doing the practice exams and get a certain score, for example 650. What kind of a difference is usually normal on the test day.

Of course on the test day the anxiety and nervousness of a person can be tricky but I don't think that scores should deviate by more than 60-70 points the most from the ones on the practice exams.

What I'm asking here is what is usally the normal difference in test score on the test day compared to practice exam scores? I want to know that because I want to factor in this deviation in score while doing practice exams and get to a comfortable score.

Thank you in advance

User avatar
GMAT Instructor
Posts: 646
Joined: Mon May 21, 2012 7:08 am
Thanked: 322 times
Followed by:143 members

by Kasia@EconomistGMAT » Wed Feb 27, 2013 7:35 am
The average variation is around 30 (max 50) points. If the difference between simulation test results and the real GMAT is more significant, it must be due to some special circumstances (noise, distraction, tiredness, illness, etc.)
Kasia
Senior Instructor
Master GMAT - the #1 rated GMAT course

"¢ If you found my post helpful, please click the "thank" button and/or follow me.

"¢ Take a 7 day free trial and find out why Economist GMAT is the highest rated GMAT course - https://gmat.economist.com/

"¢ Read GMAT Economist reviews - https://reviews.beatthegmat.com/economis ... mat-course

GMAT/MBA Expert

User avatar
GMAT Instructor
Posts: 2621
Joined: Mon Jun 02, 2008 3:17 am
Location: Montreal
Thanked: 1090 times
Followed by:355 members
GMAT Score:780

by Ian Stewart » Wed Feb 27, 2013 7:50 am
Kasia@EconomistGMAT wrote:The average variation is around 30 (max 50) points. If the difference between simulation test results and the real GMAT is more significant, it must be due to some special circumstances (noise, distraction, tiredness, illness, etc.)
That's not quite true. The 'standard error' of GMAT scores is about 30 points. If you know statistics terminology, that means your test scores will be normally distributed around your 'true level' with a standard deviation of 30 points. In concrete terms, if you perform normally on one test, you'd expect the score to be within 30 points of your 'true level' about 2/3 of the time, and within 60 points of your 'true level' 95% of the time. So for one thing, it is not completely improbable that your test day score is 60 points above or below your level, even under normal circumstances. But for another, if you've only taken one practice test, on that test you might have been lucky and overperformed by 30 points. Then on test day, you might be unlucky and underperform by 30 points - neither of those events are at all unlikely. Then you'd see a 60 point gap between the two test scores. It's only if you've taken a few practice tests that you can really get an accurate baseline of your ability, and only then is the standard error meaningful.

I'm fairly sure, though my advanced stats is a bit rusty, that the average difference you'd expect between two test scores knowing the standard error is √2 times the standard error, so you'd expect, on average, about a 40 point difference between the scores you get if you take two consecutive tests.
For online GMAT math tutoring, or to buy my higher-level Quant books and problem sets, contact me at ianstewartgmat at gmail.com

ianstewartgmat.com

User avatar
GMAT Instructor
Posts: 646
Joined: Mon May 21, 2012 7:08 am
Thanked: 322 times
Followed by:143 members

by Kasia@EconomistGMAT » Wed Feb 27, 2013 7:59 am
The test-taker obviously should not estimate his exam results based on only one simulation test. I would suggest taking at least several tests and then counting the average from them. This would give us the base score that we could use while estimating the real GMAT score.
Kasia
Senior Instructor
Master GMAT - the #1 rated GMAT course

"¢ If you found my post helpful, please click the "thank" button and/or follow me.

"¢ Take a 7 day free trial and find out why Economist GMAT is the highest rated GMAT course - https://gmat.economist.com/

"¢ Read GMAT Economist reviews - https://reviews.beatthegmat.com/economis ... mat-course

Junior | Next Rank: 30 Posts
Posts: 23
Joined: Thu Feb 07, 2013 2:51 am
Thanked: 1 times

by ynatchev » Wed Feb 27, 2013 11:01 pm
Thank you guys for you answers. I wanted to know that because I was wondering if the practice exams are a good estimate of the score that you might recieve on the actual GMAT.

As Kasia suggested and I strongly agree with her that it is nesessary to do many practice exams, not just 1. If the variation is only 30-40 point than I'll aim for a score by 50-60 point higher than my targeted score of above 600.

Again thank you guys for the help.

Newbie | Next Rank: 10 Posts
Posts: 4
Joined: Tue Jan 08, 2013 4:21 pm

by harshbajaj09 » Thu Feb 28, 2013 7:45 am
Does GMAT penalize more for getting consecutive wrong answers than getting same number of answers wrong but in gaps?
Normally how many questions one can afford to get wrong to get a 650 both in verbal and quant(rough figures)?
I am posting these questions because I got less number of questions wrong but still my score was quite low.I think the reason is that they were wrong in a stretch. Am I correct?

GMAT/MBA Expert

User avatar
GMAT Instructor
Posts: 16207
Joined: Mon Dec 08, 2008 6:26 pm
Location: Vancouver, BC
Thanked: 5254 times
Followed by:1268 members
GMAT Score:770

by Brent@GMATPrepNow » Thu Feb 28, 2013 8:35 am
harshbajaj09 wrote: Normally how many questions one can afford to get wrong to get a 650 both in verbal and quant(rough figures)?
I am posting these questions because I got less number of questions wrong but still my score was quite low.I think the reason is that they were wrong in a stretch. Am I correct?
The number correct answers has little to do with one's score. Consider this experiment that I performed:

I took GMATPrep Practice Test #1 four times, and each time I answered every second question correctly (I did this for the quant section only). Given that I correctly answered exactly half of the questions each time, you'd expect my quant scores to be roughly the same for all 4 tests.

My 4 scaled scores were: 19, 23, 26 and 42

This represents a percentile range from approximately 7th percentile to the 57th percentile.

Takeaway: Your scores are not based on the number of questions answer correctly; they're based on the level of difficulty of the questions you answer correctly.

If you're interested, we have a free video explaining the GMAT scoring algorithm: https://www.gmatprepnow.com/module/gener ... es?id=1251

Cheers,
Brent
Brent Hanneson - Creator of GMATPrepNow.com
Image

Newbie | Next Rank: 10 Posts
Posts: 4
Joined: Tue Jan 08, 2013 4:21 pm

by harshbajaj09 » Thu Feb 28, 2013 9:42 am
Thanks Brent. This was something I was totally unaware. Can I get solutions for GMATPREP questions. I got few questions wrong and I dont know how to solve them. Where can I get solutions for GMATPREP questions?

GMAT/MBA Expert

User avatar
GMAT Instructor
Posts: 16207
Joined: Mon Dec 08, 2008 6:26 pm
Location: Vancouver, BC
Thanked: 5254 times
Followed by:1268 members
GMAT Score:770

by Brent@GMATPrepNow » Thu Feb 28, 2013 10:49 am
It's safe to say that each GMATPrep question has been answered several times (over ten times in many instances).
Just use BTG's search feature (and a few keywords) to find the solutions.
Or, if you'd like a fresh-off-the-grill solution, you can post the question in the forums.

Cheers,
Brent
Brent Hanneson - Creator of GMATPrepNow.com
Image