first 10 question GMAT myth or not?

[dentobizz]

Is the importance of first 10 questions on GMAT really overrated?
In quant section 3 scenarios were tested :-
A] first ten questions wrong last 27 correct --score was Q38
B] last ten questions wrong first 27 correct--score was Q50
C]first 14 questions correct,10 questions in the middle wrong, and last 13 correct--score was Q49

Please refer the thread for more details-
http://gmatclub.com/forum/gmat-prep-sof ... l#p1171779

So clearly if you get the first ten wrong you hurt your score pretty badly that cannot be improved later on but the GMAC says that the first 8/10 questions don't matter that much.
Why is there a contradiction? Please comment. Thanks

[Levent-g]

Hi,

I just asked a similiar question. Possibly worthy to read.

time-management-first-10-questions-t17471.html

[StaceyKoprince]

Thanks, levent.

This is actually what I was getting at in my response to your other post. I'm going to start by saying: people have been looking for ways to "game" the scoring algorithm for decades. It doesn't work. The people who make the GMAT are VERY smart. There isn't some magic-bullet out there.

These "theoretical" test scenarios that people use to try to "prove" that the early questions are worth more... they are, by definition, not the way the test would actually work in practice. And that invalidates the results.

For instance, anyone who really can get the first 27 in a row correct is very unlikely to get the last 10 wrong (unless they simply run out of time). It's next to impossible to get 27 in a row correct, period. Not going to happen. This is an adaptive test. You're hitting the stratosphere after 5+ correct questions in a row.

Next, anyone who cannot answer a single one of the first 10 questions correctly is not then going to answer 27 in a row correctly. In other words, this "theoretical" scenario testing does not match real-test conditions.

When someone tries to "manipulate" the algorithm to extremes in an unrealistic manner... don't expect the results to reflect the reality of how things will happen when a real person takes the real test.

If you want to dive more into the nitty-gritty on how the algorithm actually does work (and why the above extreme scenarios are pretty much useless), read this:
http://www.manhattangmat.com/blog/index ... ions-myth/