By training teachers to vary systematically the way in which they performed recitation strategy, it was hoped to determine (1) whether teachers can be trained to control precisely the way they teach when using that strategy and (2) whether different versions have measurably different effects on student learning. (MM)

This paper reports on an emerging technique that yields the kind of knowledge about cause-and-effect connections between teaching practices and student achievement on which teacher education should be based. That knowledge comes from the completion in recent studies of what has been termed the "descriptive-correlational-experimental loop."…