Assessing question difficulty accuracy?? (700 scorers, plz)

I got a pretty long passage towards the end, it was about African American rights I believe... I was feeling super exhausted by the end of the exam and most of all very depressed. But it turned out that I did good!

Anyway, what I am trying to say is that first and foremost you should not try to guess the difficulty level of the questions you see on the GMAT. There are a few reasons for this:

1. You should focus on solving the question, not wondering if it's a 600-level or a 700-level. Whatever the test throws at you, your duty is to work through it!

2. Difficulty is a subjective matter. What seems difficult to me might be a breeze for you and the other way around.

3. You may already know this, but here goes: the difficulty of a question does not come from it being long or convoluted, but rather from the number of people that answer it correctly. The experimental questions in the test are presented to you for exactly that reason: they first measure the number of people that get it right/wrong and then they assign a difficulty level and push the question in the "real" test pool.

About canceling your score: you should know that this will still appear on your score report, the one that gets sent to school, so the B school will know that you've taken the test before and canceled the score. I'm not sure if you've made a mistake or not, no one can tell you that. However, I for one would have a hard time getting out of the test center without a score in my hand (maybe because I'm such a dork), but that's just my personal taste!

Best of luck on your retake!

odannyboi -

Let me echo what Dana said so well above...

There is also the variable of the questions that do not count. The questions that are being tested to be used on future tests. These "experimentals" are questions that do not count toward your score. You are likely to have several of these on the test. The consensus is that very often one whole reading passage - and the questions associated with it - is experimental. So that may be the very passage that you are referencing.

I had a student once who thought about canceling his score because on the Quant side he faced several questions in the last half of the section that he thought we too simple. He figured that he must have really messed up since he had heard some things about the test adapting and he thought that this adaptation was a perfect sort of linear progression - as you get questions right each is harder than the previous. The reality is much more complicated as Dana mentions, and questions that do not count complicate it further. Anyway he held his breath and asked the computer for his score. He actually scored over the 90th percentile on the Quant and those questions had no predictive relationship to his score at all.

I generally tell students that if it is your first time taking the test there are very few circumstances when you would really want to cancel and these are do to your own tangible assessment of your performance - you left 10 questions completely unanswered or you were on cold medication and lost track of what you were doing during the test, that type of thing.

The good news is that you no doubt were doing much better than you thought and can now look forward to really scoring well on the retake!

Thanks for the replies from the both of you! I know, I really made a mistake by not reporting my score (I had a caffeine/energy crash halfway through verbal).

I will assume that I got a 490 and study as if I got a 490. I've been doing more DS problems since I am pretty bad at those and have getting a higher percentage right.

My problem was that I would take a practice test, then barely study over what I did wrong. My form of studying was kinda stubborn. I just wanted higher practice scores without even looking at what I did wrong. I'm changing my mentality. Thanks for the expert advice!