In a recent model of cognitive development, Karmiloff-Smith (1992) alerts us to levels of cognitive development at which children's understanding of a problem can be implicit, or can be particularly resistant to input and disconfirmatory evidence. At the same time, numerous studies have illustrated the beneficial effects of working in small groups on children's learning. This study asks whether the benefits of group work are available to children who have different levels of representation about balancing an object on a fulcrum, and in particular whether children who use a "center" strategy for balancing derive less benefit from social processes than children at other levels. A balance beam task was used to explore 5 to 7 year old children's behavior and to categorise them according to levels of representation derived from Karmiloff-Smith's model. The 103 children then worked in groups either with others classified at the same or at a different level of representation. Subsequently, children were post-tested individually and classified again. More children were found to benefit from group discussion when they were with children at different levels to themselves. However, this effect was not found for children with a "center theory" who were predicted to be resistant to input or disconfirmatory evidence. This supports Karmiloff-Smith's claim for a level of knowledge which is resistant to input, and adds a cautionary note to the literature advocating group collaboration, since the benefits may not be the same for all children.
- PEER COLLABORATION