Next: M&M's Up: CHANCE Previous: Regression to the

# GPA's

## Discussion

For your journal, you have been asked to use the data available on the CHANCE gopher to investigate whether women's SAT scores underpredict their achievement in college, as measured either by cumulative GPA or by performance in individual courses. In order to help you with this assignment, we would like you to discuss specific strategies for investigating this question.

1. Let's take the particular question of how well the total SAT score (SATV + SATM) predicts cumulative GPA. We have data listing for one graduating class of Dartmouth psychology majors their SAT scores, together with their final GPA. What can we do with this data? What tests can we run? What confounding factors will we have to worry about in interpreting the results of the tests?

2. Suppose we wanted to come up with some way of translating between men's and women's SAT scores. How could we do this? In particular, how could we determine what total SAT score for a man would be equivalent to a women's 1290?

3. Suppose we find that women's scores systematically underpredict their achievement in college. Should this difference be taken into account in the UC admissions process? If so, then how?

NOTICE. Since the CHANCE fair will be held on Thursday 3 June, the last due date for journals has been changed to Tuesday 1 June.

snell@dartmouth.edu
Tue Jun 28 15:24:59 EDT 1994