Conservation Accidents Revisited

Abstract

McGarrigle and Donaldson's (1975) claim that the Genevan conservation test procedure underestimates the child's cognitive competence is examined. Experiment 1 reports a successful replication of their original work. We then consider several objections to McGarrigle and Donaldson's claim that the tasks they used did in fact test the child's ability to conserve number. Firstly, as only four counters were used in their study the child's judgment of numerosity could have been perceptually rather than conceptually based (Winer 1974). Secondly, teddy may have acted as a distracting agent leading the child to ignore the post-transformational state of the arrays and to answer the experimenter's post-transformational question (correctly) by mere repetition of his earlier response. In experiment 2 a procedure which is free from these objections, but otherwise similar to McGarrigle and Donaldson's procedure, is employed, together with appropriate comparison conditions. This procedure results in a decrease in the overall incidence of conservation responses, but fails to eradicate the difference between accidental and intentional transformation of the arrays (p < 0.001). In the final part of the paper we consider the implications of these and McGarrigle and Donaldson's results for our understanding of the social psychology of the conservation task.