Project Follow Through - Dissemination of Results

Dissemination of Results

In 1972, the OE created the Joint Dissemination Review Panel (JDRP) and the National Diffusion Network (NDN) to disseminate information about effective models to schools and districts nationwide (Watkins, 1997, p. 47; Rhine, 1981, p. 307). JDRP reviewed programs for effectiveness according to a mixture of empirical and holistic criteria. NDN was responsible for disseminating the results based on the recommendations of JDRP. Watkins (1997) criticizes the dissemination criteria for two reasons. First, the organizations identified programs for dissemination that were not part of the Follow Through experiment with no empirical validation. Second, JDRP and NDN endorsed programs that showed improvement in areas “such as self-concept, attitude, and mental or physical health (of students) …(or) if it has a positive impact on individuals other than students, for example if it results in improved instructional behavior of teachers” (p. 47), but did not raise students’ academic achievement. Thus, programs “that had been incapable of demonstrating improved academic performance in the Follow Through evaluation” were recommended for adoption by schools and districts. Watkins cites the former Commissioner of Education, Ernest Boyer, who wrote with dismay that “Since only one of the sponsors (Direct Instruction) was found to produce positive results more consistently than any of the others, it would be inappropriate and irresponsible to disseminate information on all of the models” (Watkins, 1997, p. 48).

Of course, it would have been ideal to have the kind of conclusiveness associated with laboratory experiments when we conduct social experiments in communities and schools. Andy B. Anderson (1975) wrote that “the idea of a controlled experiment has long been recognized as a goal worth pursuing in the social and behavioral sciences for the same obvious reason that made this mode of inquiry the predominant research strategy of the natural and physical sciences: the controlled experiment permits the most unequivocal assessment of a variable’s influence on another variable” (p. 13). Particularly when experimentation is used as a tool for informing policy decisions (e.g., in recommending the efficacy of some instructional approaches with disadvantaged students over other, less effective interventions), the design should be of the highest degree of rigor possible. For a variety of reasons, Follow Through did not have the classic characteristics of a true experiment.