Effects of training on four-year olds' performance of geometric analogy problems

Typescript (photocopy).The purpose of this investigation was to examine effects of training on four-year olds' ability to solve geometric analogy problems. Sixty four-year olds were screened using the Children's Analogical Reasoning Task (CART). Of these sixty children, 30 who performed below chance level were randomly assigned to either the trained or nontrained group. The nontrained group consisted of 10 subjects. The trained group consisted of 20 subjects who received training administered in three consecutive 20 minute sessions. The content of the training sessions was based on Sternberg's component processes underlying analogical reasoning. Instructional procedures followed the Direct Instruction model which emphasized explicit teacher explanation, student participation in problem solution, continual teacher feedback, and independent student practice followed by teacher review of the processes which pertained to problem solution. All problems presented during training represented analogies in the A:B::C:D format. The materials employed in the training sessions, which were sequenced from concrete to abstract and simple to complex, included concrete toy objects, abstract tree figures, parquetry blocks, and attribute blocks. Quantitative and qualitative analyses were conducted to analyze differences in subjects' performance on a pretest and on immediate and delayed posttests. Quantitative analyses included analysis of covariance, analysis of variance, and Bonferroni-adjusted t tests. Qualitative analyses consisted of comparisons of oral explanations on the pretest and posttests. Results of quantitative analysis demonstrated a significant difference between groups on the immediate and delayed posttests, F = 41.74, df = 1,27, p < .0001. For simple main effect of training there were significant differences between trained and nontrained groups on the immediate, F = 49.00, df = 1,56, p < .0001, and delayed posttests, F = 78.99, df = 1,56, p < .001. Trained subjects consistently outperformed nontrained subjects on the geometric analogy task. The increases between the pretest and delayed posttest indicated that the effects of training were maintained for a one month period. Qualitative analyses provided evidence that the verbal responses of the trained group reflected greater attention to the higher order relations within and between pairs of terms in the analogy problems following training