ABSTRACT
An empirical account of intellectual development is expected to be sensitive to performance on assessment tasks. A common assumption is that the characteristic features of human thinking are revealed by specified assessment tasks which share a common structure. The original versions of Piagetian tasks are reliable. Yet when new tasks with new procedures and new methods of assessment are used, there is a differential pattern of failure and success. Piaget's 'top-down' interests lay in the identification and description of intellectual structures by means of which performances on tasks are viewed. In short, there is a general problem which faces all developmental accounts where inferences about unobserved processes are made from observed performances on assessment tasks and conversely. Thus, there is ample scope for mismatching of intellectual and task structures due to the heterogeneity of developmentalists' practice. The point to make about this heterogeneity is that, due to its presence, confounding of theoretical and empirical issues tends to occur.
