Hello, Welcome to this post! In this post we shall discuss about the GMAT testing engine.

Quick question: Is GMAT exam a computer based test or a computer adaptive test [CAT] in any way?

Ans: it’s adaptive and it is question adaptive.  That means, if you do the current question correct, the next question will be harder than the present one. If you get the current question wrong, the next question would be easier than the one you answered wrong.

Don’t think that it is better to get the present questions wrong so that we may get easier question later; the score is directly proportional to the level of difficulty. Hence, the easier questions are, the lower score you get for the particular question.

For understanding the factors affecting the GMAT score, we shall consider two students Anil and Neha taking the GMAT. We will study their performance on a graph with the x-axis being the number of questions and the y-axis being the level of difficulty which will correspond to easy (E), medium (M) and hard (H) questions as shown.

Graph 1







Anil starts taking the test. Because the GMAT test does not know whether Anil’s standard is, it starts off with a medium level question. Suppose he gets the first question correctly, he moves up by getting a difficult question. If he gets the second question also correctly, he moves further up in difficulty. Now if he gets the third question wrong then the fourth question he gets is not of the same level of difficulty as the third question but moves down slightly. As he got the second question is correct, the fourth question is in-between the second and the third question. If he answers the fourth question correctly then he will go up and get fifth question in-between the third and the fourth. Now if he gets the fifth question wrong, then he will move down and get a sixth question in-between the fourth and the fifth question. Similarly if he gets the sixth question correctly he will move up and get a question in-between the fifth and the sixth question as shown.

Graph 2







Now consider Neha, she also starts taking this GMAT test. She bombs the first question, so she goes down and gets a second question which is easier than the first question. Suppose she gets even the second question wrong, she goes further down and gets a third question which is easier than the second question. Now if she gets the third question correct then she moves up and gets a fourth question which is between the second and the third question. And if she gets the fourth question correct then she moves up and gets the fifth question which is in-between the second and the fourth level. Again if she gets the fifth question correct then the server may give a question which is of the same difficulty as the second question or even better. And if her sixth question is answered correctly then she gets a question which is between the first and the second as shown.

Graph 3







Now observe the graph above and count the number of correct and wrong questions which Anil and Neha got. You might have observed that both of them have attempted 6 questions and out of which both have got 4 correct and 2 questions wrong. As score is directly proportional to the level of difficulty, we can say that easy, medium and hard corresponds to 0, 30 and 60 respectively (as the score range is from 0 to 60 raw points). If both of them proceed in a similar manner, Anil ends up getting around a 45 and Neha ends up getting around 30 raw points.

Graph 4






So the first inference about the Computer Adaptive System on GMAT is:

Initial questions are very crucial”

Now if you are the person, who read the GMAT Official Guide, you would oppose the above statement saying it is a ‘Myth’. It is a myth indeed as it is not the ONLY factor that affects the score in GMAT.

We are going to see the second factor now. Again consider two more students, let’s say Sunil and Sneha. Suppose both are at the tail end of the examination and both have six questions to do in only two minutes time. Sunil acknowledges that he cannot do all the six questions in a span of two minutes, so he guesses all the six questions. He gets two questions correctly and four questions wrong. Sneha on the other hand tries to get as many questions correctly as possible so attempts them one by one. She does three questions correctly in a span of two minutes and bam! “time up”, so she leaves the other three questions unanswered. Now who get a better score? Sunil gets a much better score than Sneha, as there is a huge penalty of leaving the questions unanswered.

So the second inference about the Computer Adaptive System on GMAT is that:

Completing the given questions in the given time is extremely important

Seeing the above two factors, can we come up with the following strategy:

“As there are 37 & 41 questions in quant & verbal respectively, which should be answered in a span of 75 minutes each, can we do the first 20-25 questions in 70 minutes and guess the remaining in last 5 minutes?”

No! That would not be the best strategy because of the third factor

No continuous mistakes

There will be high probability of getting continuous mistakes in guessing 17 questions in a span of 5 minutes. So even though the graph may go up in the first 20 questions, because of continuous mistakes, it will rapidly shoot down as demonstrated in the first scenario and it may even come down below than where it started.

So to score higher on a GMAT, the following are the three most important factors:

  • Initial questions being critically important
  • Completing the given questions in the given time
  • No continuous mistakes

No matter who you are, at some point of time, you will fall into two categories,

1) You will not know how to do a question (since the level increases) or

2) You will not have time to do a question.

So forget worrying about how to deal with the hard topics in GMAT, GMAT itself needs a strategy to crack. At Kaplan – LogIQuest we teach you this strategy by giving you a pacing chart explaining how to pace yourself through the questions. We will also provide you “the unique Kaplan methodology” on how to approach initial, middle and final questions.  For more details on GMAT the test and Kaplan methodology, walk into your nearest LogIQuest centre and speak with our faculty. Or if you are a fan of reading more, try our GMAT strategy posts about How to prepare for a GMAT 720+ score.

Anirudh is from teaching background. He is a senior Math faculty and has several certifications as a quant trainer to teach for GRE, Gmat, SAT, PSAT & ACT. He taught more than 100 GRE, 100 SAT and 50 GMAT batches. He has also taught in many prestigious institutions such as IIT Hyderabad, Oakridge international school etc., He received many awards as a Quant Trainer.