The research combines longitudinal observations with experimental interventions to investigate a cognitive stage model for predicting learning sequences and probabilities of behavior acquisition in normal, at-risk, and retarded children. The research has major implications for sequencing instructional materials in intervention packages. The major experimental hypothesis is that diverse social and object-oriented behaviors, typically acquired in the first two years, are related in their development by stages of underlying structural organization. This organization suggests the design of two types of training studies. For each type, subjects varying in the level of behaviors they display on related dimensions are matched and trained on a criterion dimension. The first type of training study tests the hypothesis that individuals displaying behaviors at higher levels on related dimensions will benefit more from training on the targeted behavior. The second type of training study tests the hypothesis that certain cognitive bahaviors at the same or lower level of organization as a target are prerequisites to its subsequent acquisition. The basis for both predictions is that, if structurally-related cognitive behaviors are already in the individual's repertoire, the underlying organization needed to acquire the new behavior must be present. Two major longitudinal observational hypotheses, generated by the stage model, will be tested. The first hypothesis is that sequences of social and cognitive behavioral acquisition will be invariant across three samples of normal, at-risk and handicapped infants, despite variations in acquisition rates. The second hypothesis is that individual progress in behavioral acquisition across all samples will be characterized by parallel alternating periods of stage transition and consolidation for both domains. Transitions will occur at specific mental rather than chronological ages. The coordinated use of both experimental and observational methodologies will produce results of greater generalizability than either design alone, since a subset of the multiple hypothesized functional relationships among behaviors, observed longitudinally, can be verified in the context of experimental training studies. Such complementary information, generated within one consistent theoretical framework, has major implications for the design of interventions used by psychology and education with the handicapped.
