Piagetian Psychology in Intelligent Tutoring Systems

In this paper, we describe the use of Piaget’s notion of cognitive development in the building of pre-tests that would allow improving a tutor’s reasoning ability. We are interested in finding individual differences that not only predict a student’s overall performance, but that can also be easily applied to actual tutoring decisions. Our hypothesis was that students with different levels of cognitive development should behave differently in the tutor, and that this is the reason why they need to be taught with different strategies. We thought it was very likely that our population of elementary school students would have different cognitive levels, so that this feature would be an essential aspect to take into account to adapt the response of a tutoring system. We have adapted classic Piagetian tasks used to measure level of cognitive development for use on computer. We found that this measure predicts student performance at a variety of grain sizes: understanding of hints, amount of time to solve problems, failure rate and also the number of problems students need to attempt to master a topic. We also describe how these measures of cognitive development can be usefully applied to improve the behavior of the tutor for students at different cognitive levels.