Harris Cooper Jeffrey C. Valentine Kelly Charlton Univernity of Missouri, Columbia
Few disciplines present as much challenge to the research synthesist as does special education. The challenge arises primarily from the wide variety of research designs, outcome measures, and diagnostic labels that characterize research in special education. For example, it is not unusual to find the same special educational intervention tested using one-group pretest--posttest designs, nonequivalent control group designs, experimental designs, single-subject designs, and case studies. Outcomes measured for this one intervention might include academic achievement, social and behavioral adjustment, attitudes, and self-concept.
As difficult as the synthesis task may be, there are important reasons for special education researchers to undertake the challenge. First, accumulation is a cornerstone of the scientific enterprise and without syntheses special education research will be severely limited in its use by policymakers. If researchers easily succumb to the conclusion that "these studies, ostensibly evaluating the same intervention, are too dissimilar to warrant aggregation," it is a short leap to the conclusion that "this study cannot be generalized to settings, people, and programs not contained in them." Obviously, adopting this stance can be overly cautious and can greatly undermine the value of research. With this in mind, those who undertake the task of gathering, summarizing, and integrating special education literatures deserve added acclaim.
Second, special education researchers need to keep in mind the fact that the process of aggregation, and the glossing over of differences in set-