Abstract
The weaknesses and merits of a Piagetean and an information‐processing point of view are briefly discussed. It is concluded that a detailed diagnosis of the actual psychological processes is necessary before one can decide upon the usefulness of alternative formal models. The high inter‐ and intra‐individual variability makes it necessary to base the analyses on the asymptotic performance level of individual subjects. Response latency may be a very sensitive measure of complexity.