GMAT Scoring demystified ..... atleast partly

Let us deal the top 3 questions on GMAT scoring one by one.

(1) How does the GMAT CAT work ? How does it figure which questions to ask ?

If you have not already done so, please do go through this page:

http://www.pagalguy.com/forum/gmat-and-related-discussions/16716-manhattangmat-how-gmat-finds-your.html

It is amazing to see the Master from Manhattan take the entire scoring process apart. A must read for anyone wanting to peek into the intracasies.

Now, that you know how the questions pop up, we go to the next question:

(2) For given RAW scores what would be the corresponding scores ?

Here is a nice little table to compute your scaled scores, given that you have Q and V raw scores:

Q+V Raw Score ScaledScore
48-50 400
49-51 420
51-53 430
52-54 440
54-56 450
55-57 460
56-58 470
59-60 480
61-62 490
62-63 500
62-63 510
62-64 520
63-64 530
65-66 540
66-67 550
67-69 560
68-71 570
69-72 580

70-72 590

71-73 600

72-75 610

73-76 620

74-77 630

75-78 640

76-79 650

78-81 660

79-81 670

81-83 680

81-83 690

82-86 700

84-86 710

85-87 720

87-90 730

89-90 740
90-91 750

92-93 760

94-96 770

96-98 780

97-100 790

100-101 800

Remember that GMAT looks for your expertise across areas. Which means if you make similar mistakes in the same subject area your overall scaled score will be lower than an other person who uniformaly distributes his mistakes in all the subject area. For example, my Q+V raw score is 87, but scaled score is not 720 (according to above table). So use the above table more to estimate with a fair degree of accuracy where your scaled scores would be.

Now, let us get to the third problem:

(3) How many mistakes can I afford in each section ?

Now, this is a slightly tricky question as it is not just the number of mistakes per se, but also 2 other factors which need to be considered:

1. Frequency of the mistakes: If you get questions wrong one after the other, you are in greater risk than if you distribute your mistakes over a range. For example, lets say, from questions 21 to 30 there are 2 candidates X and Y and their frequency of mistakes is: X marks the wrong answers for questions 22, 26 and 29, while Y marks the wrong ones for 23,24,25. Then Y would be penalized heavier than X.
2. Position of the mistakes: Has been said a gazillion times before, but the first 10 odd questions have a greater variation/fluctuation while assessing your score than the questions towards the end. Think about it. If you want to guage a person in an interview you start out asking the "average" questions if he cannot answer those, you would immediately drop down to the "basic" ones and it would be a long time before you could throw anything challenging at him. Whereas if he answered all the average ones you would pose him the "difficult" ones and it would be a long, long series of wrongs answers before you get fed up and start asking him "basic" questions.

However, it is possible for very high scores (read as "very few mistakes") to show with a fair degree of accuracy what the association between mistakes and RAW scores. The following data I found from the net to be fairly representative:

Verbal
# mistakes score
0-1 51
2-4 50-45
5-7 40-44
8-13 39-35

Quant
# mistakes score
0-2 51
3-5 50
6-1 49
11-14 48

Now that you know how GMAT scoring works, go ahead and hit those GMATPreps and PowerPreps !! All the best !