Publisher: Educational Theory, 55 (3)
Page Numbers: 245-162

This article argues that just as both hammers and saws are needed to build a good house, both variance and process approaches are needed to build a good understanding of causation in education. This essay discusses four research designs for pursuing questions about causation in education. Two of them take a variance approach; the other two take a process approach. The point of this discussion is to illustrate, first, their respective strengths and, second, their necessary interdependence.