I have taken 4 Mgmat cats, and my quant experience is as follows;
The first one was a close fight, at the end I got 50, after the exam I saw that after the first couple of questions all questions were of difficulty 700-800. This makes sense
The second one again was a close fight, I got 51, and this time time again nearly all of the questions were of difficulty 700-800. This makes sense too.
The third one was a different story, this time I started a bit better at the beginning and at the end got 51 however the questions seemed easier. After the exam I saw that after the first half, half of the remaining questions were of difficulty 600-700 (the rest 700-800). However I was definetly doing better than the first two attempts at the half of the exam. So what is going here?
The fourth was even more interesting. This time I was doing even better than my third attempt and I got 51, This time I saw after first 10 or 15 questions I was rarely asked a question of difficulty 700-800. The rest was between 500-600 and 600-700. So what is going on here?
The summary of this is, after a certain threshold the better I do the easier the rest of the questions. Assuming this is not coincide, my question is does the same applies to real GMAT?
One possible explanation: The 6 CATS draw questions from a common pool with insufficient number of hard questions. After some time the algorithm needs to use them a bit economically. This is OK.
One another explanation: This is completely intentional. It is said that the scoring of the GMAT algorithm takes into account two main parameters. The first is how hard were the questions answered correctly, the second is how easy were the questions answered incorrectly. So is there a point for CAT that it really makes no sense to continue asking difficult questions since even if the test-taker answers them incorrectly he will still be at the top-scoring segment. Therefore it would make more sense to ask easier questions that though he is still likely to answer correctly, if he does not it will lower his score?
The second seems reasonable to me, although I do not like as a test-taker that CAT can be so cunning, however there still is a catch. Assume I was doing well in terms of correctly answered questions but at the expense of the time. Then, easier questions at the end would be exactly the thing that I would need. On the other hand if the cat continued to ask harder questions after some time probably my wrong answers would be sufficient in number to lower the score. So here comes a related question: does the CAT takes into account the remaining time while determining the difficulty of questions.
The first one was a close fight, at the end I got 50, after the exam I saw that after the first couple of questions all questions were of difficulty 700-800. This makes sense
The second one again was a close fight, I got 51, and this time time again nearly all of the questions were of difficulty 700-800. This makes sense too.
The third one was a different story, this time I started a bit better at the beginning and at the end got 51 however the questions seemed easier. After the exam I saw that after the first half, half of the remaining questions were of difficulty 600-700 (the rest 700-800). However I was definetly doing better than the first two attempts at the half of the exam. So what is going here?
The fourth was even more interesting. This time I was doing even better than my third attempt and I got 51, This time I saw after first 10 or 15 questions I was rarely asked a question of difficulty 700-800. The rest was between 500-600 and 600-700. So what is going on here?
The summary of this is, after a certain threshold the better I do the easier the rest of the questions. Assuming this is not coincide, my question is does the same applies to real GMAT?
One possible explanation: The 6 CATS draw questions from a common pool with insufficient number of hard questions. After some time the algorithm needs to use them a bit economically. This is OK.
One another explanation: This is completely intentional. It is said that the scoring of the GMAT algorithm takes into account two main parameters. The first is how hard were the questions answered correctly, the second is how easy were the questions answered incorrectly. So is there a point for CAT that it really makes no sense to continue asking difficult questions since even if the test-taker answers them incorrectly he will still be at the top-scoring segment. Therefore it would make more sense to ask easier questions that though he is still likely to answer correctly, if he does not it will lower his score?
The second seems reasonable to me, although I do not like as a test-taker that CAT can be so cunning, however there still is a catch. Assume I was doing well in terms of correctly answered questions but at the expense of the time. Then, easier questions at the end would be exactly the thing that I would need. On the other hand if the cat continued to ask harder questions after some time probably my wrong answers would be sufficient in number to lower the score. So here comes a related question: does the CAT takes into account the remaining time while determining the difficulty of questions.
