| Abstract: | Traditional means of evaluating ability in class inclusion do not take account of the effects of guessing on performance. A new analysis of response patterns rectifies this problem. Strategic components of performance are estimated on the basis of the differential consequences of inclusion logic, comparison of the subclasses, and guessing. The importance of this analysis is demonstrated by a study with two language conditions. The 32 children in each of four age groups (4, 5, 6, and 8 years) were randomly assigned to either the standard Piagetian condition or to a reduced-language condition in which visual symbols were used to represent the classes. The 4-year-olds gave fewer correct answers in the standard than in the reduced-language condition (p? |
