The scoring pattern in GMAT is slightly confusing me. I am trying various ways to find this out, but couldn't reach any plausible explanation. Ok. My questions are...
1. If i want to score 49-50 in quant, how many mistakes can i afford to make?
2. If i want a score of above 40 (or even above 38) how many mistakes (in terms of number of questions) can i afford to make?
I know that the score depends on the level of difficulty of the question and it is hard to tell the weightage. I heard that in verbal, SC carries more weightage than other questions. Is it true?
Quant:
In my practice tests i am making 8 mistakes consistently. Most of them are from DS. But again this is not because i don't the concepts. It is because of my sheer negligence and carelessness. With more practice i believe i can reduce this number to 3. I want to know how much i can score in quant with 3 mistakes. If anyone has a pattern or a mechanism to explain this, please go ahead and help me. I am targeting a score of 50 in quant. (49 in the worst case scenario).
Verbal:
Can you believe that i made 15 mistakes in my last practice test. Sounds demotivating, but that is reality for me. 7 mistakes in SC, 4 in RC and 3 in CR. As i am targeting a score of 40+ in verbal, i need to improve. Gotta practice...
Lets see if i can reduce the number of mistakes in my next practice test. Aloha!!!
Subscribe to:
Post Comments (Atom)
3 comments:
the gmat algorithm has been discussed often, but i doubt anyone knows the exact details.
the key to score well is to make as few mistakes (ideally none) in the first 10-15 questions. if that takes a little more time, so be it. a good first leg ensures that you will be in the top spectrum of scores. if you make 8 mistakes in the first 15 questions, you will be penalised heavily against what will happen if you make the same no of mistakes during the later questions(the later the better).
FG is OTM on this. It is really really difficult to try and quantify the target score (out of 51) in the number of right/wrong answers.
This being because some questions are test questions that don't count, and we don't know how much certain questions are weighted against others.
FG's point that the first questions are more important stands to reason, as the algorithm most likely works like a probability tree, with the wider range in difficulty being early on, when the algorithm has less historical data about the test taker.
@FG,t-test
Thanks for your response. Points taken. I am not gonna break my head anymore on this. It is better to concentrate on minimizing mistakes rather than thinking how many i can afford to make.
Post a Comment