Doubts on Scores of GMATPrep and MGMAT

[Suresh]

Hi,

I am really getting confused with the scores that i am getting in the GMATPrep tests. I have got only 10 questions incorrect in both VA and QA and got just a meagre 620 (QA 44, VA 31). I dont understand why it is so and I am taking GMAT on Jun 17.

I had taken MGMAT and got 670 (QA 47, VA 34) though I had got 20 VA questions incorrect and 12 QA questions incorrect.

I know that each question has different difficulty levels and scored differently. But what I dont understand is that, even scoring 27 questions and 31 questions correct in the GMATPrep s/w, I am just getting 620. So how many questions should I be getting right to get a score of 700+?

What do you think my ideal score would be in the actual test with this kind of performance?

Please help me get these doubts cleared, you wonderful MGMATters.

Thanks


Suresh

[dave1618]

I wouldnt pay much attention to the number of correct questions per section as a determinant of your final score. From my experience it was more so about the level of difficulty of the questions than anything else. IE if you are consistently getting questions in the range of 710-740 level difficulty, you could probably miss half of the questions per section bc the test might be going from a 740 level question that you got wrong, to a 720 level question you got right, to a 730 level question you got wrong to a 720 level questions you got wrong to a 710 level question you got right, back up to a 720 level question you got wrong again, back to a 710 level question you got right. From what I understand of the algorithm it is more concerned with the range of questions you are fluctuating between. I'd say the number correct/incorrect isnt even a factor in the score calculation. I'm no expert by any means but this was my impression. Im sure one of the MGMAt people can answer you more accurately, as they do this for a living (pause...).

Based on my experience (see my post above yours on this board) I think that prepping with the actual GMAT tests provided on MBA.com is a much wiser move. While MAnhatten GMAT is fantastic for many things, their tests failed me in a way in that I was given a false sense of confidence by their CAT tests. Note that I can only speak for myself and could very well be an outlying case from the norm. I'm curious to hear others' experiences.

With that said, the manhattan gmat prep CAT tests are amazing in their answer explanations and the analytical and reporting tools they provide to hone in on certain types of questions you get wrong and gain a more holistic picture of your performance. Based on my experience, if I were advising you I would say to put little to no weight on the scores you get on the MGMAT tests and do not use these tests as an indicator of where you will actually score but more so as a great resource to go over test problems and understand their explanations.

[rfernandez]

The way the GMAT is scored is nothing like the way your teachers scored your tests in school. It's not so much about how many you get correct vs. incorrect, but rather WHICH questions you get correct and WHICH questions you get incorrect. dave1618's example illustrates this idea well.

The GMAT Prep practice CAT should be part of every student's preparation, as its scoring algorithm is identical to the real thing. The big drawback, however, is that analyzing your results is a chore and the problem solutions are inconsistent in their quality.

While our practice CAT certainly aims to be predictive of a student's score on the official exam, this by itself is not its sole purpose. I stress to my students to use the exam as a diagnostic: what content areas are difficult for you and require more attention? How's your timing on each question type? Are there any trends you need to nip in the bud? What were the specific areas of improvement from one test to the next? I also stress using the exam as a planning tool: what should I prioritize in my studying? Are there obvious strengths that I can lay aside for now? Should I do lots of timing drills? Etc. All of these questions are important and are tangential to the final score.

All that said, we are very committed to making our exam as predictive as possible. Our exam's standard error is about 50 points. Not a small number, but neither is the official exam's standard error of about 30 points. So please continue to give us feedback on the exams and we'll continue to refine our algorithm. Thanks, and good luck with your studying.

[Suresh]

thanks Rey,

[StaceyKoprince]

Great advice from above! Glad we were able to help, Suresh.