Abstract
This paper develops a simple model of student choice to explain why some teaching innovations have only a negligible effect on mean student performance. Teaching innovations are defined as small changes in pedagogy that enable students to more quickly convert time to knowledge. In modeling student behavior it is assumed that some students are more interested in their level of performance or in minimizing their effort than in mastering a subject. The model demonstrates that if students set grade targets, some students choose to learn less and the change in mean student achievement will tend to be small.