Transfer effects of teaching conceptual versus perceptual analogies

The study's main objective was to investigate the transfer effects of training in one type of analogies (conceptual versus perceptual) on performance of the second type. A sample of 396 kindergarten children (mean age = 5:8 years) were randomly assigned to three experimental (E1, E2, E3) and one control group. Groups E1 and E2 received training in either conceptual (n = 118) or perceptual analogies (n = 121), respectively. Training included encoding, inference, mapping, application, and response. Group E3 (n = 93) was trained only in the relation between the analogy terms A : B and the control group had no training. All groups were administered the Children's Conceptual and Perceptual Analogical Modifiability (CCPAM) test before and after the training. Groups E1 and E2 improved their test performance significantly from pre- to posttraining as compared with group E3, which showed a significant decrease in conceptual analogies and no improvement in perceptual analogies. Perceptual training (group E2) was limited to gains in perceptual analogies whereas conceptual training (group E1) was effective in improving conceptual and perceptual analogies. Group E3 showed a pre- to posttraining performance decrease. The findings are discussed in relation to mediation processes, task characteristics, conditions for transfer, and practical implications for teaching.