Causal reasoning about pairs of relations and analogical reasoning in young children

Two experiments were conducted to investigate how children aged 3-6 years performed in classical analogy problems based on single and double relations. The number of relations in an analogy is seen as an important competence factor in Halford's theory of analogical development (1987, 1992, 1993). In contrast, other authors have proposed that the critical factor in analogical success is relational familiarity (e.g. Brown, 1989; Chen & Daehler, 1989; Gentner, 1989; Goswami, 1992, 1996; Inagaki & Hatano, 1987; Vosniadou, 1989). In our experiments, we gave children analogies to solve based on pairs of causal relations (such as cutting and wetting), and also measured their performance when similar analogies were based on single relations (such as cutting or wetting). The ability to process the relations or relation used in each analogy was also measured in control conditions. The results suggested that the number of relations in an analogy does not overload available capacity. Clear 'learning-to-learn' effects were found in both studies, showing that children's analogical performance can improve significantly over the course of an experiment, even when relational familiarity is present.