Comparison of Natural and Model-Trained Conservers

Abstract

This study examined whether natural conservers differ from trained ones. In Exp. 1, four groups were tested: low partial conservers, high partial conservers, younger full conservers, and older full conservers. The first three were examined again in Exp. 2. In both experiments, high and low partial conserver groups observed conservation modeling, thereby creating a subset of trained conservers. In the first experiment, extinction modeling was introduced, and in the second, generalization and delayed testing were studied. The results of Exp. 1 showed that the stability for conservation by trained conservers, especially low partial conservers, was less than that offered by natural conservers, although even low partial conservers maintained a high average score. In Exp. 2, there were no significant differences among the three groups on the delayed posttest and the delayed generalization tests. These findings suggest that trained conservers can acquire stable conservation concepts through modeling, although low partial conservers are inferior in stability for conservation, and that modeling is more effective when it is developmentally appropriate.