quantium17
Is it possible the algorithm is taking into account time spent answering question? i.e. spending more time on a question than average and getting incorrect lowers difficulty even more and vice versa?
I am trying to find patterns with these ridiculous edge case quant scores and am wondering if anyone else sees any kind of pattern or correlation with time spent and overall score.
Hi. Good thinking and good idea about very very very unlikely. I can tell you that I have not noticed any correlation in the GMAT prep exams. Nobody knows the reality of the actual gem Matic exam but I don’t see many reasons why GMAC would hide a critical piece of scoring information.
my leading theory, still unproven of course, so merely hypothesis is that the question bank was lacking hard questions. This may be partially coincidental. On the GMAT exam, you have to get a certain number of certain types of questions, for example on the verbal section, you need to get four reading comprehension passages, sometimes it may be three but it’s never less than three and it’s never more than four. simmer on the quant, you probably need to get some arithmetic questions and some algebra and some probability and combinations if you are getting up there in difficulty but if you answered probability and combinations in the exam and the exam had to serve you algebra questions and it did not have any hard algebra questions, it could only serve medium level difficulty. Similarly, if the question bank was not loaded with enough of difficult questions of all, something with this may happen.
However, this theory may prove complete bollocks as we are hoping to speak to a team within the GMAC that knows how scoring works to help us explain some of these very unusual results. However the timeline for this is keep getting pushed out and I think the earliest is going to be January 👀