Abstract
The purpose of the present experiment was to obtain further evidence of the importance of initial imitation of modeled conservation (quantity) for the attainment of mastery using an authentic observational learning situation. Nonconservers were selected during a pretest; half succeeded and half failed an anticipation task. In tasks that did not involve conservation, half the anticipators and nonanticipators were pretrained in generalized imitation of modeled responses; the other half underwent sensitization. One-third of the subjects were then exposed to a model demonstrating correct conservation judgments with a compensation argument, one-third were exposed to a model who provided incorrect, nonconservation judgments with a justification referring to width, and one-third saw no demonstration at all. Posttesting was conducted immediately after demonstration and again four weeks later. Progress was restricted to those anticipator groups exposed to the correct model; the learning of these competent children was higher after pretraining in imitation than after sensitization, and their achievement improved from immediate to delayed posttesting. These data were accounted for by a sequence of imitation and comprehension processes, prompted by a degree of compliant rehearsal of modeled conservation, which was positively correlated with initial competence. They appear to be incompatible with a claim that imitation does not intervene in the cognitive processing that fosters the observational acquisition of conservation.