Causes and associations - a developmental investigation
This thesis explores preschoolers' use of causal-associations (associations between specific instrument and effects) and causal relations incorporating knowledge of transformation from initial to final states. Gelman, Bullock and Meek (1980) claimed that preschoolers use causal relations. However, in their study it was possible to make correct choices using causal-associations. Therefore, in Experiments 1 and 2 tasks were designed to distinguish between use of the two relations. Preschoolers made inferences about instruments which could produce the transformations depicted within event sequences (based on those used by Gelman et al.). Performance on tasks requiring selection of causal relations was significantly worse than performance on tasks where causal-associations could be used. Two methods of reasoning, Relational (causal) and Associative- Causal Matching, were identified. Modified versions of the tasks in Experiment 1 were used in Experiments 3, 4 and 5. Four-year-olds were more proficient at using Relational Methods than 3-year-olds although both ages demonstrated the ability to use this method. In Experiment 6 children had to construct sequences. Three-year-olds preferred associative constructions to causal ones. This preference may have influenced their performance in the previous experiments. Experiments 1, 2, 3 and 5 incorporated sequences with compound end-states (e.g. wet and broken cup). Three-year-olds clearly preferred to focus on just one of these attributes, chosen on the basis of salience. This preference evidently contributed to their lower scores throughout these experiments. However, even when relative salience was controlled (Experiment 7) or when single attributes were used (Experiment 4) 3-year-olds' performance was worse than 4-year-olds' in terms of choices based on causal relations. These experiments indicate that preschoolers use both Associative-Causal and Relational Methods. There is evidence for a shift from a preference for judgements based on associative relations to a preference for causal relations between 3 and 5 years. The ability to deal with compound features also appears to develop over the preschool years.