GMATPREP reliability

It is very strange indeed. I would have said no until today when there was a post about a 60 point jump day over day.
gmat-prep-variance-78812.html

Another aspect could be luck and relaxation - if you were taking a test relaxed and not stressed, that may have increased your performance or if you were guessing several questions, you may have been very lucky that day....

I would say - take another test, possibly by a different provider (Such as a free Manhattan GMAT Test). And that will tell you if you should head to the test center or a book store. Or maybe you can wait another half a year and get a 780 if the pattern works?

Hmmm, it seems there ia a number of GMATprep test takers experiencing the weird twists recently... I don't know what is going on, Gprep used to be realy good indicator and the scores between two tests (especially math tests) used to be very similar. I think I had Q49 on both tests and questions were of totally equal weight.

The only dicrepancy I've noticed is between Verbal scores (or should I say Verbal tests?). Verbal test on GMATprep 2 is significantly harder than that in Gprep 1, solely due to Sentence Correction. While CR and RC are on par in two tests, SC questions in 2nd test are real hell. There were some real mind twisters overthere, especially in the end of the 2nd test, so my scores in the Verbal portion of Gprep 2 were always sligthly lower.

On the real test, verbal questions are more similar to those in 1st test, so one shouldn't worry if experience lower score in the 2nd. While in 2nd test there is 5-6 mind blowing SC questions, on the real deal you'll probably get 1 or 2 such a difficult.

yes I really can't claim that I increased 50 points in ability over the past year, since I haven't been studying at all. I mean it is possible that my mind grew stronger after the 32 credits of classes I took, but that is a stretch. Yeah I guess the only way to really know is to do GPREP2 or an MGMAT.
I always thought that the software made the tests identical in difficulty, except that the questions used in GPREP1 were held out of the pool for GPREP2...

Well, they had intention to make two equally difficult test, but it seems they've failed to do so, at least when it comes to SC.

Try another test, but don't use MGMAT tests as any measure of your GMAT abilities. While some test takers find it good for practice, MGMAT is totally useless as a score predictor on the real test.

so gmatprep is the only real reliable score predictor? I mean I heard MGMAT is supposed to be just as accurate as gmatprep, I know MGMAT is supposed to be like the most professional of the prep programs out there, from their website it seems like they've spent a lot of money on getting it right.

It is good as a practice tool, but not as an estimator of the future test performance. I dont like MGMAT tests and haven't realy used them, I think their questions are not representative of what you are about to see on the real deal.

However, many peole find it good for practice. On the other hand, very very few people are going to tell you that there is any correlation between MGMAT scores and ones in the real test.

That does bring up a good point tho, i mean, who wants to take a bunch of CATs if its not for practice, i mean having a reliable reference score is key to deciding whether or not taking the test again is a good idea, but I mean knowing exactly the range you can expect to get wont change your scores any, so practice is more important than an accurate prediction. Most of the people I have talked to in the last day or so say that if you add 20 to MGMAT scores its usually close, but GMATPREP is the only really reliable one.
I did take a GMATPREP test just after taking the GMAT my first time (a year ago), and got a 700 (46Q 39V), which is almost the exact score I got on the test (this is because almost all of my test prep the first time was with Kaplan Premier, where my scores were in the 550-600 range).
I guess that brings up the point that waiting a year increased my score 50 points, being that there is no other explanation for an increase of that size. After reviewing my answers, I didn't find any of the ones I got right to be "lucky," where I suppose a fifth of the ones you get wrong should be, and statistical variability could not substantiate 50 points, maybe 10-20 but not 50. I was a sophomore in undergrad when I took it the first time, and now I just finished my junior year, so I guess if you are young and trying to get it out of the way early, just going to classes and getting older can increase your score. I hadn't studied at all, I decided just before I took this practice test that if it wasn't over 700 i would consider not even taking the GMAT again. I expected to do much worse because of no preparation at all; I had posted a few times on these forums but that was it really. I guess my mind just developed more through the academic rigors of the past semester.

1. Keep in mind that 44V is a great score, the 97th percentile. 50Q is the 94th percentile.

2. The overriding factor when calculating your score is the difficulty of the questions you faced, not the number of mistakes.

3. There is some natural variability to the scores: someone doing the GMAT on two consecutive Mondays would probably see the same scores.

4. You may have done the GMATPrep at a time of day that works better for you, or under less pressure. Did you do the AWA portion as well?

5. You may have learnt something important in the last two days: for example, how weighted averages work or the rule about the pronoun which.

I completed the test in full, giving 100% on the AWAs, as I wanted to get the most accurate estimation of my potential score.
I did start at one or two in the afternoon, instead of 8am like most of the scheduled tests start, and that could key into it, but I don't think this could be the only reason, as I had had a rough day and done a lot of things. I remember thinking that being tired may effect my score negatively, but I guess it didn't...
I seriously just picked around in my studying, I think all i really did was post some things on GMATclub, I mean I may have looked at Kaplan 800 for a bit, but nothing really enlightening came to me.
I think the only big factor is that a year passed, and that i had accumulated reasoning capabilities in that time. I increased both verbal and quantitative scores by 3 or more points each, I mean I have never heard of a 3-month study program producing that kind of results, unless it was from beginning to end.
I do remember that when I took the test the first time I had to work until 11pm the night before, which may have resulted in a negatively distorted actual GMAT result and this prep test shows the upside of what I could actually score.
Either way, I am not going to complain, I am taking the other prep test this sunday then deleting all the answers so I can try to forget them. I am taking the GMAT again on August 28th, try to get a 750+ I suppose, it would help solidify the school I want to go to.

Kevincan, I see you recieved a 790, congratulations.

Do you have any advice, I think I am going to embark on a 3-month quest to try and get an 800. If I am as smart as I think I am, I may be able to crack the test, but I need whatever help I can get. Do you have a debrief posted or any immediate advice you could give?

thanks.

I tried to emulate actual test conditions as much as possible.
Drove over 30 miles to a neutral test site, my father's house.
took exact time limits instead of estimates
took at 8 am

and the result was negative
720 44Q 44V

I saw many very simple mistakes on the Q, mostly due to nervousness I believe.
In any event, I think my preparation should take place in these type of conditions instead of just whenever I want in my own house, because I might end up with skewed results.

Very good idea.
I actually took the GMAT in a different city (because the test center in my city was booked out for 2 months).
I ended up taking a trip and staying at a hotel for a few days (had to take TOEFL first and then GMAT).

This worked out quite well (obviously).

I took TOEFL on Saturday morning and then GMAT on Monday morning.
I wanted to make sure of a few things:
1. I had enough of a break between the two tests to rest, though I could not stay for too long
2. I took TOEFL first and since it is a much easier exam, it enabled me to visit the test center and figure out the location, procedures, computers, environment, etc without the stress that GMAT puts.
3. TOEFL also ended up being a confidence booster. I got perfect score on the multiple choice, so that reduced my stress on the GMAT day.
4. I was able to relax that weekend since I was in a different city, I was away from all my contacts, trouble, and just had one friend I hung out with. She was also into GMAT and MBA, so this was very helpful to have her support as well.

I would say I probably got 20-30 points because of going out of town and taking TOEFL first, but that's my very subjective opinion.


PS: DK - you have 198 posts at this moment. 2 more posts and you will have access to all GMAT Club's tests - take them. They will boost your score - see comments here if you have any doubts: https://gmatclub.com/blog/testimonials/

Yes I now feel that relaxation and familiarity with everything around you are key to focusing on the more difficult problems. Most of the mistakes I made were on the Q section and they were very simple errors in the tricks of the more difficult problems, most of the problems concerned around time constraints. I think I was just nervous at the new surroundings and could not concentrate effectively, a key element of the actual test day. I would suggest practicing under these conditions to everyone who is preparing to take the GMAT, as it is a definite aspect.

In light of bb's experience, I am now throwing the idea of making two test appointments, and rejecting the first score regardless of how I feel I did. I know its like throwing $250 down the drain, but I think if its like 2-3 days before the other appointment it could seriously improve my score. Any thoughts? do colleges actually look at attempted GMATs where the score was refused?

Thank you btw for the 200 posts comment, i have taken the two free ones and they were very helpful.

I don't think you can schedule two appointments within 30 days of each other.

even so, I mean a month before would probably be sufficient I think to become adapted to the surroundings. The question is does that leave a bad mark on my applications?

I also have another question... what is the average drop due to nervousness/anxiety that most people can expect during the actual test? For example if one tests 770 consistently, is he expected to get around a 720 on the actual due to nerves? On the LSAT most people attribute a 3 point drop, so a 175 range correlates to an approximate 172 on the actual. Of course there are deviations, but this is what is empirically observed

You can see this very well demonstrated here: https://gmatclub.com/gmat-score-estimate

Just use the Performance Under Pressure Slider all the way to the left and you will how much stress impacts performance - can be quite a bit but on average about 10-20 points (based on GMAT Prep - very different for other tests), though it varies as different people perform differently under stress. I got 10 points more than on a practice test the day before.

In your particular example.
If you consistently get 770 on GMAT prep, then here are your scores for different levels of stress:

Average: 760
Below Average: 710
Stress Wreck: 670