Test scores rose after tutoring. Did tutoring do it?

Question

A school launches an after-school tutoring program in September. It tests the participating students in the fall and again in the spring, finds their average math score rose by 12 points, and reports this as the effect of tutoring.

Why can the school not attribute the 12-point gain to tutoring, and what would fix the design?

Accepted Answer

Students improve over a school year anyway. Between fall and spring, children get older, sit through a full year of regular classes, and simply mature. This natural growth, called maturation, would raise scores even with no tutoring at all. The fall-to-spring gain bundles the tutoring effect together with a full year of ordinary learning. Who joined also matters. If students or parents chose to join, participants may differ from non-participants in motivation, which further muddies a simple before-and-after comparison. The fix: a comparison group. Take similar students and assign them at rand…

Test scores rose after tutoring. Did tutoring do it?

Your answer

More Statistics questions