Making learning easier using examples in problem solving tasks: Application of the AMBRE project
Speaker: Sandra Nogri, Université de Paris 8
Date: 26 January 2006
A classic learning situation consists of presenting solved problems followed by isomorphic problems to be solved. This way of teaching and learning raises a variety of questions: what is the nature of the knowledge that is built? In what conditions does this knowledge get built and used? In the current research, we aimed to deepen the study of the conditions that make learning easier by using examples. Research exists on learning by examples and has shown different knowledge generalisation processes (detecting similarities, explanatory processes, adaptation) that learners can use when faced with examples and problems to solve. This research also identifies sources of progress factors. However, according to previous research, a single factor seems to have quite variable effects.
From our point of view, a better understanding of the conditions that favour learning from examples will put to use work on the process of generalisation and studies that address sources of progress. We conducted different experiments in order to show certain modalities for presenting problems that can lead to the use of a generalisation process. We specifically tested the effect of one factor: the salience of the element necessary for solving the problem. These experiments show that decreasing the salience of this element leads students to use the process of knowledge generalisation.
This topic not only presents a theoretical perspective, but also a practical interest. A better understanding of the conditions that facilitate learning should provide some direction for the conception of computer learning environments. Parallel to our experimental studies, we have participated in the creation of a computer learning environment from the AMBRE project. Within this project, we made use of example-based learning research to identify the best conditions of this kind of learning and we suggest recommendations that use example-based learning guided by the system. We then evaluated the impact of a learning environment on addition problems in primary school based on the principles of the AMBRE project and using the proposed recommendations.
This research led us to reflect on the application of these theories and data from cognitive psychology for the construction of Computer Learning Environments as well as ways to assess those environments.
Click here to view the presentation.
(If you are having trouble viewing, see technical advice)