Suppose you were employed as a consultant by a school district and charged with determining how much suspending a bad-behaving student reduces or increases his or her academic performance.

(Throughout these questions when I say "suspension" I mean "out-of-school suspensions.")

(Is it possible that suspending a misbehaving student will increase his average test scores? I'd say, maybe. What if suspension deters future bad behavior?)

What cross-sectional dataset could we use to study this relationship?

a) The high school class of 2012's final GPA for each student, along with the total number of times each student was suspended throughout his or her high school career.

b) Weekly student quiz scores for all students in the school district, along with a weekly measure of the number of days they missed school because of suspension, for a period of 30 weeks.

c)The district's annual dropout rate and suspension rate (unduplicated headcount) for the years 1950 through 2012.

d)All of the above.