New teaching method in one class: diff-in-diff

Question

A teacher tries a new method in Class A and keeps Class B on the usual method as a control. Average test scores before and after:

Class A (treated): 60 before, 75 after.
Class B (control): 62 before, 70 after.

Compute the difference-in-differences estimate of the new method's effect, and explain why it is smaller than the raw 15-point gain.

Accepted Answer

The raw gain overstates the method. Class A rose 15 points (from 60 to 75), and it is tempting to credit the new method with all 15. But Class B, on the old method, rose 8 points (from 62 to 70) over the same period. That 8-point rise reflects ordinary learning over the term, easier second test, or general maturation, forces that also lifted Class A. Crediting the full 15 to the method double-counts the part that would have happened anyway. Difference-in-differences. Subtract the control's change from the treated group's change: 	ext{DiD} = (75 - 60) - (70 - 62) = 15 - 8 = 7. The estimated e…

New teaching method in one class: diff-in-diff

Your answer

More Statistics questions