Whatever the innovation, the adoption-rejection decision must be related to precise evaluation data; and in cases where positive benefits are the clear result of an innovative program, the implementation of an effective diffusion program can magnify and multiply the benefits for a larger population of students. The Peterborough Project demonstrates that educational theory and practice can be successfully and productively fused in a school setting. It gives a significant role to a rigorous scientific methodology in the improvement of educational practice through its elaborate evaluation procedure. This procedure involves student performance objectives and criterion-referenced testing, pilot trial and field testing, experimental design considerations, and hypothesis testing. The Project also assumes that linear strategies of educational development and innovation installation are inadequate. It, therefore, conceptualizes a working model for school change that depends on a high degree of school autonomy in curriculum decisions as well as on classroom teachers' acceptance of central responsibility for the school curriculum. (Author/WM)