Abstract
There are qualitative differences in the modal hypothesis-sampling systems of preoperational, concrete operational, and formal operational children and young adults. These are congenial to predictions derived from Piaget. Preoperational kindergarten children show mostly stereotypes, while their concrete operational peers exhibit use of dimension and hypothesis-checking in more than half their problems. After training formal operational eighth graders and college students show perfect processing in most problems; those who score as concrete operational appear unable to acquire this strategy. Cognitive level is not a good predictor of performance among middle-aged (40 to 50 yr.) and older adults (60 to 80 yr.).
