Credit for the GMAT CATs

[Night reader]

I am just copying and pasting my reply from a topic placed into the BTG's Strategy sub-forum here

>>>... various GMAT preparation companies develop and introduce their CATs. Now, I am asking you and the BTG community: 'How many times, before going into a deal, or taking some one's word for counting, or believing something and acting thereon, had we asked ourselves - why any credit should be given at all?' I guess the vast majority of us, even easy believers, would respond 'always'.

Let's question - why we had believed in MGMAT, Princeton Review, Knewton, Kaplan, Grockit and others? Is there any statistical evidence suggesting that actual scores reported by the certain number of recent (within 5-year period) GMAT test takers and their practice CAT results are correlated? Is there any precise test deviation data used in place, except for that taken from the internet forums, web blogs? What is the sampling error, population statistics... No, there is nothing or almost nothing indicating an objective approach to the CAT algorithms distributed with the mentioned company software products. In fact, these companies could not come close to the real GMAT scoring algorithm patented after the GMAC and ACT Inc. even within 100 score deviation. Yet, some would argue and believe that MGMAT has 60+ deviation from the actual GMAT score, or Kaplan has 80+, whatever.

I have certain doubts about the GMAT Prep software too. It is unlikely that the exam scoring algorithm would be made available to the population of potential examinees. Some elements - yes, general description and imitation of the process - yes, but the mechanism of test scoring itself, - I just have strong doubts about this.
>>

[Ttronn]

I agree its difficult to determine with any certainty how credible any assistance provider is going to be. But in all honesty if you use tried and tested resources (Veritas, PowerScore, MGMAT etc.) and the official guide to make sure you understand the concepts being tested, and use the GMAT PREP software and any other CAT programs (GMAT specific of course) to ensure that you're comfortable with the timing you can be confident that you've given the test your best shot with the most complete preparation. Try not to worry so much about who is (more) credible and just focus on learning the concepts and becoming familiar and confident with the test's timing constraints.

[kola_member]

I have my share of doubts about these CATS too but i would say the GmatPrep CATS are the closest to the real thing. I say closest because there might still be some variations, however if you look at the raw scores of your latest GmatPrep, you can say with some certainty that they will be freaky close to the real thing. And how does this happen ? i really have no definite answer. For example, check out point #3 from a debrief posted recently by a fellow member :
https://www.beatthegmat.com/from-640-to- ... tml#317492

[aslan]

@Night reader:I think what you are trying to get at is whether there is any 'standard bar' by which we can measure Or are measuring any 2nd party CATs?.

I think there isn't any.Primarily because these test makers do not really have any official feedback from their customers once they give the tests, and the few that actually return to give, are only some 0.01% of the lot.This surely cannot make a company get closer to how IT itself is measuring against the 'real' thing.

The other thing is that we also get totally involved in how we are doing on a particular CAT, irrespective of how we should measure ourselves with respect to what officially is provided (GMATprep).GMAT does not not care how you are doing on other tests, it follows its own line.The original has updated percentile, and thus the difficulty of increase or decrease of questions depending on 'who did how on the previous test' measures up; in addition, research questions also impact somewhat as probably GMATprep has some as well.

In sum all the testing companies are marketing 'themselves' without any KPI's to conform to....and its just like branding of shampoos or any other product, all do the same thing and nobody has certified results in its offering.So take all with a grain of salt with whatever aspect you like about them.

[fitzgerald23]

I think you have to take those tests for what they are. They give you some type of approximation of your skill level based on their own research. You also have to have some faith in the companies as they are well established and well established companies usually do hire people to consult with them that would be familiar with the actual test and what they are trying to duplicate with their own tests. I dont think any company will tell you to take a free test and if you score 700+ that you are set. The whole purpose of the CATS is to get used to the testing format and to measure your improvement after your studying of the concepts.

[Night reader]

I think you have to take those tests for what they are. They give you some type of approximation of your skill level based on their own research. You also have to have some faith in the companies as they are well established and well established companies usually do hire people to consult with them that would be familiar with the actual test and what they are trying to duplicate with their own tests. I dont think any company will tell you to take a free test and if you score 700+ that you are set. The whole purpose of the CATS is to get used to the testing format and to measure your improvement after your studying of the concepts.

Fitzgerald, we may not walk on air... CAT is not a format element of the GMAT exam; it has the content implications. The format of GMAT is well described for the population of test takers. CAT is giving you attempts based on your prior attempts, and this rests more on content. The problem with GMAC's exam is that its content is ever changing. None may approximate which score level of questions follow what. When tested for GMAT, one is put under dynamically changing conditions. Depending on one's physical, emotional, and intellectual properties, an examinee may obtain wildly varying results.

Who can prove that with the purchase and use of GMAT preparation tool kit(s) and CATs one will ensure three properties mentioned above are set adequately?

GMAC does testing of about 1/3 of all real exam questions purely for experimental purpose. It may objectively assess a test taker's strengths and weaknesses based on the pool of prior exam results and statistical data. Now see, what the GMAT preparation companies do - they offer free diagnostic and commercial CATs which can be taken under or not under the real test conditions. Moreover the commercial CATs will offer their own algorithms qualifying our test taking capacities.

I am questioning you, whether there is any correlation between the existing CATs and actual GMAT scores? If there is one, I have not comprehended it yet. I am not saying that commercial CATs are useless; we can do timed tests from the GMAT preparation company sources, and we may try to develop some of our test taking capacities. But agree, there is not any significance in one's CAT scores for approximation of his/her actual exam scores.

[Ttronn]

I think you have to take those tests for what they are. They give you some type of approximation of your skill level based on their own research. You also have to have some faith in the companies as they are well established and well established companies usually do hire people to consult with them that would be familiar with the actual test and what they are trying to duplicate with their own tests. I dont think any company will tell you to take a free test and if you score 700+ that you are set. The whole purpose of the CATS is to get used to the testing format and to measure your improvement after your studying of the concepts.

Fitzgerald, we may not walk on air... CAT is not a format element of the GMAT exam; it has the content implications. The format of GMAT is well described for the population of test takers. CAT is giving you attempts based on your prior attempts, and this rests more on content. The problem with GMAC's exam is that its content is ever changing. None may approximate which score level of questions follow what. When tested for GMAT, one is put under dynamically changing conditions. Depending on one's physical, emotional, and intellectual properties, an examinee may obtain wildly varying results.

Who can prove that with the purchase and use of GMAT preparation tool kit(s) and CATs one will ensure three properties mentioned above are set adequately?

GMAC does testing of about 1/3 of all real exam questions purely for experimental purpose. It may objectively assess a test taker's strengths and weaknesses based on the pool of prior exam results and statistical data. Now see, what the GMAT preparation companies do - they offer free diagnostic and commercial CATs which can be taken under or not under the real test conditions. Moreover the commercial CATs will offer their own algorithms qualifying our test taking capacities.

I am questioning you, whether there is any correlation between the existing CATs and actual GMAT scores? If there is one, I have not comprehended it yet. I am not saying that commercial CATs are useless; we can do timed tests from the GMAT preparation company sources, and we may try to develop some of our test taking capacities. But agree, there is not any significance in one's CAT scores for approximation of his/her actual exam scores.

And this is exactly why these CAT tests should be used only to judge your ability to deal with the time constraints and not the content itself in my opinion. The best way to ensure you're comfortable with the content and scoring at your specific target is to study the material (with a study aid if you wish) so that you are sure to understand the concepts. At that point move on to the CAT tests. But like you said, they are imperfect because they "may not" accurately represent how the actual GMAT will adjust your test content based upon previous answers.

Basically I caution anyone that uses these CAT practice tests to help determine whether they are "ready" or not based on the scores they're getting because they may or may not be accurate representations of the way the actual GMAT will react to your test administration

[gmatmachoman]

Folk,

As a part of this commuinty, we have done extensive study /R&D on GMAT.
You may ask me whether GMAT Prep & Real GMAT are same.

Here lies the answer:

https://www.mba.com/mba/thegmat/download ... onsoftware

The GMAC society itself says in its website they use same technology for both GMAT Prep & Real Test.


SO DONT WORRY ABOUT THAT!!

If u want to have fine insights, i suggest you to take a peep in to our study on R&D on GMAT prep below undersigned link.

Partially yes, Knewton seems to give a better shot at scaling of scores.
I had continuous 4 720's in practice test and real one turned out to be 710..

MGMAT: They are almost good. u can find the mean of the scores and find the SD..That can tell u the "range"( almost scaled one).

My suggestion is GMAT prep is a "better indicator" provided u have not solved those questions prior to it. Assuming u r gonna solve then for the first time, GMAT prep + Knewton are probable indicators of real test.
Again with a pinch of salt, u need to simulate the exact real conditions.

[uwhusky]

Study hard, sleep well, and take the test.

Why complicate things?

[Night reader]

Study hard, sleep well, and take the test.

Why complicate things?

yes, that's how it all goes eventually.

[kola_member]

Study hard, sleep well, and take the test.

Why complicate things?

When you are NOT satisfied with your score, want some more, dig to the core ? err sorry...

[uwhusky]

Study hard, sleep well, and take the test.

Why complicate things?

When you are NOT satisfied with your score, want some more, dig to the core ? err sorry...

Correct, but I do not feel that the direction of the approach is on the right path. You guys are discussing aspects that are beyond your control and have no direct relationship to your GMAT performance.

[Night reader]

Study hard, sleep well, and take the test.

Why complicate things?

When you are NOT satisfied with your score, want some more, dig to the core ? err sorry...

Correct, but I do not feel that the direction of the approach is on the right path. You guys are discussing aspects that are beyond your control and have no direct relationship to your GMAT performance.

And the topic was started with all its criticism put onto CATs of the GMAT preparation companies. I believe the efforts employed for the CATs solely to crack the thing are futile.

I started this topic because of the endless talks going at various GMAT forums about the usability of CATs.

Fell free to introduce your pros and cons about CATs.

Uwhusky: when you drive a car you must observe the traffic congestion, right? It's beyond your control, still an obstacle on your destination...

CATs are viewed in the same manner. We want to discuss them now. We want to clear the situation about CATs.