i suppose the right answer is A, but actually, the right answer is 5/37, which isn't up there.
This question is based on the known percentages of a normal distribution, and standard deviation. In group A, 16% of the test takers scored below 440, and in group b, 2.5% scored below 440. If we assume 100 students in each class, means 18.5 total scored below 440, so the answer is 2.5/18.5 = 5/37.
In the Princeton Review version of Standard Deviation, they ignore that .5% possibility for convenience sake, so in their world, Class B would have only 2% below 440, making the fraction 2 out of 18, or 1/9.
I don't know why the simplify things like that when it's wrong, but I guess it's a good way to make this stuff taste better to a large portion of their students.
At the same time, I haven't seen a GMAT question that actually required knowing those percentages, and I don't even teach them to my students, so I'm not sure this is even relevant.
Does that sound right?