A process theory of inductive reasoning tested by the teaching of domain-specific thinking strategies

Abstract

Inductive reasoning can be conceived as the process of discovering regularities by finding out identity and difference with respect to attributes of and relations between objects. This assumption gives rise to the definition of a set of inductive tasks consisting of six subsets, all of which can be solved by variants of a basic strategy. Such a theory can be tested by teaching subjects the strategy, i.e. by training experiments: Inducing the postulated processes should lead to predictable improvements in certain tasks and to no improvements in other tasks. The article provides an account of the main results of about 30 experiments. The theory seems to be sound and the training approach proved to be a powerful research method: A great number of differential effects, partly considerable in amount, have been predicted and empirically tested, most of them even by several replications. Transfer effect of an inductive thinking training on intelligence test performance is about twice as great as an average test coaching effect.