Abstract
Problem-solving performance of 24 Ss who solved the disctransfer problem by moving the discs was compared to the performance of 24 Ss who anticipated moves by naming the discs by number and pointing to where they should go. When the problem was solved by moving the discs, and the moves and the successive collections of discs were perceptually present, Ss were more likely to state principles for solving the problems than when it was solved by anticipating moves and when Ss had to imagine the moves and the successive collections of discs. With respect to the errors made, however, Ss who solved the problem by anticipating moves benefitted the most from being able to state a principle.