The performance of children (56 boys and 36 girls) with different motivational and cognitive characteristics in three "open" and three "traditional" fourthgrade classrooms was assessed. Cluster analysis of factor scores representing child orientations, motives, and prior achievement produced six "types." Three-way analyses of variance investigated the effects of child type, classroom type (open vs. traditional), and sex of child (and various interactions) on several outcome measures, including academic achievement, creativity, inquiry skill, social-educational attitudes, and teacher ratings of children's classroom behavior. Main effects appeared for each of the three independent variables, along with several Child-type x Class-type interactions. An approach using child types or clusters rather than abstracted dimensions may facilitate future Attribute x Treatment interaction research and applications.
Although a great deal has been writtenabout "open education" in the past several years (e.g., Blackie, 1971;Featherstone, 1971;Kohl, 1969;Silberman, 1970), little evaluational research has been reported until quite recently. The most relevant research which has been done to date (e.g.,