by noah Mon Dec 05, 2011 11:21 pm
You're just about there! All you needed to do was tweak your numbers. Remember, you're trying to make the numbers work to explain the discrepancy. So, play with them until they do. Take a look:
The discrepancy in this stimulus is that half the respondents to an alumni survey reported they were in the top quartile of the class. How can half fit into a quartile?
If you're not seeing it, compare what I just wrote with this: The discrepancy in this stimulus is that half the alumni reported they were in the top quartile of the class.
See the difference? The real stimulus (the top one) is about the folks that responded to a survey. Perhaps only the smarties responded. Let's imagine there were 100 people in the class, and 50 responded to the survey. If all the top 25 students responded, that would make up half the respondents, even though they're a quarter of the whole class. This is what (A) hinges on.
(B) is tempting if you thought that the solution had to be that folks were lying or delusional about their ranking. But (B) tells us that most people were correct!
(C) is in some ways the reverse of (A). We need to see a higher proportion of smarties responding, not fewer of them.
(D) makes the discrepancy harder to explain. If almost everyone responded, than how did a half fit into a quarter?!
(E) is out of scope. Who cares how they calculated the grades?
I hope that clears it up.