An investigation of cognitive factors involved in the development of problem-solving strategies by young children
There is much current interest in children's problem-solving, both within education, and within psychology. The present study explores the development of young children's problem-solving abilities, and the cognitive factors which might be related to this. Such development is conceptualised in terms of the emergence of increasingly sophisticated and powerful cognitive strategies. In a previous study (Whitebread, 1983), which involved 20 children aged 5 and 6 years, a strong interaction was revealed between underlying cognitive factors, strategy use and performance on a reclassification task. The present work is an extension of that study with a more complex task, and with a wider age-range of children. On this occasion, children's performance on an inductive reasoning task (the multidimensional discrimination learning task) was examined. The sample consisted of 72 Leicestershire Primary school children, comprising three equal groups of 24 children aged 6, 8 and 10 years. The children were tested on a number of cognitive factors theoretically predicted to influence performance on reasoning and problem-solving tasks. These predictors included working memory capacity, metacognitive awareness and control, style of attribution, and two measures of cognitive style (cognitive tempo and field dependence-independence). Cluster analysis of strategic components revealed a pattern of 7 clusters of increasingly complex strategic behaviours used by the children on the MDL task. These Strategy Clusters appeared to be principally differentiated by an increasing ability to integrate information gained from different trials. Two stylistic variations were also identified which were related to the number of hypotheses verbalised on each trial. Further investigation involving multiple regression analyses revealed that the major factor which predicted strategic behaviour and performance on the MDL task was metacognitive awareness and control. However, correlational analyses of subgroups revealed interactions between predictors, and between predictors and strategies, in relation to performance. No significant effects were revealed relating to gender, but age effects in relation to predictors, strategies and performance were indicated. The implications for future research and for the development of children's thinking and problem-solving skills within educational contexts is discussed.