MGMAT Artifically Lower Our First Test Result???

[juke]

I just finished an MGMAT test and got a 610, despite testing a 650 on GMAT Prep two days ago.

Analyzing my results, I found that the average question level I received was around a 695 (MANY more 700-800's than 500-600's) and answered correctly was around a 670..

I suspect that MGMAT might have artificially deflated my score to:
1. To convince me I sucked to make me buy their products and services.
2. Make me think I did poorly, so when I use them, I'll see my score artificially increase because of their practice materials.

I am pretty sure Knewton did this by giving me a 560 on a test when I scored a 630 on a GMAT Prep the week before. It makes you think you need them and makes the conditions of their "guarantee" easier to achieve.

Does anybody have any statements about this? Did you receive drastically lower 1st exam results, before purchasing anything, or
receive many questions far outside the bounds of what you should have been getting?

Can I get an official statement from MGMAT stating that they do not tamper with the test results to make students buy their products?

[StaceyKoprince]

No, none of our tests artificially lower your test results. Period. We do not offer a guarantee, or promise any kind of score increase, for two reasons. First, we have no stake in your starting score; you either want to use our products and services or you don't. Second, you are the single biggest factor in your own score improvement - not us. We can't guarantee how hard or how effectively you will work.

Standardized tests in general are not anywhere near as precise as you might think. The real test has a standard deviation of about 30 points and our tests have an SD of about 50 points. Your test results (on our test) are not outside the normal standard deviation.

Next, adaptive tests are not scored based upon an average of your performance or difficulty levels, nor are they based on the percentage that you answer correctly. You'll have a much better chance of achieving your goal score if you learn more about how the GMAT is scored. Take a look at the Scoring section of our free e-book The GMAT Uncovered to learn why what you described is not at all unusual for an adaptive test.

The "average" difficulty levels that you calculated are not necessarily reflective of what you were given on our test. Although you see a rating of, say, 600-700 for a particular problem, we have actually assigned difficulty levels* in an increment of 10 points (610, 620, etc). The questions that you answered correctly, for example, may have averaged much closer to 600; the 600-700 "range" does not mean that those questions average to 650.

*Those difficulty levels are assigned based upon the data that we gather as our students take the test; we collate all of the data and determine how difficult the question is across the population of test-takers.

I'm guessing that you had timing issues, and that your scoring level was higher in the middle of the section than at the end of the section (or sections - perhaps this happened on both quant and verbal). Your final score for a section is not an average of your performance across the section; it is simply where you're at when the section ends. You could be scoring at the 99th percentile for part of the section, but if your score drops to 70th percentile by the last question, then your score is 70th percentile.

As I said, do go read more about how the scoring algorithm works; until you understand it, you will be likely to try to take the test the way you took tests in school - and that will lead to a lower score than you would otherwise be capable of getting.

You can use this article to analyze your test results:
http://www.manhattangmat.com/blog/index ... ice-tests/

And these articles will help with any timing issues:
http://www.manhattangmat.com/blog/index ... to-win-it/
http://www.manhattangmat.com/blog/index ... nt-part-1/

Finally, I have to add this. We take a great deal of pride in what we do. Our entire goal is to help our students get the scores that they want to get and we work hard every single day to make our materials the best in the business. Students seem to agree, as our books are the #1 selling books, and many of those students have not taken one of our tests before buying the books - so we're not "fooling" them. I find the suggestion that we would engage in dishonest practices in order to make money extremely offensive.

Good luck with your studies. I hope you get the score that you want.

[juke]

Thanks Stacey; that was a great response.
In my case, I did not know the questions were measured to the closest 10, I didn't know about the higher standard deviation, and I also did do better at the beginning than at the end, so I rest assured :)

[StaceyKoprince]

Thank you for your response; I appreciate it. :)

Good luck with your studies; let us know if you have any other questions!