Analogical Reasoning and Schema Formation.

Abstract

The spontaneous noticing and application of analogies between semantically distant problems was examined in five experiments. Duncker's radiation problem of destroying a tumor by coincident weak rays rather than one strong ray was the major problem used, and analogies were written in the form of stories, each describing a problem and its solution. The basic paradigm involved presenting the first problem in the guise of a story comprehension task, where the story contained the first problem and solution. The radiation problem was then presented without mention of its analogous relationship to the previous story. Following the problem-solving task, a questionnaire was distributed asking subjects to apply the solution used in the first problem to the second one, even if they had done so already. Spontaneous noticing was measured by production of the analogous solution (used in the first story) to the radiation problem, on subjects' first pass through the radiation problem. Application with a hint was indicated if subjects produced this analogous solution in response to the final questionnaire, provided they had not previously produced it. Three experiments dealt with spontaneous transfer from one analogy only. The results of the one-story experiments indicated that subjects often failed to notice the correspondence between the problems, even when attempts were made to have subjects increase the generality and abstractness of the encoding of the story by summarization, as opposed to recall (Experiment I), or by including a diagram with the story that schematically illustrates the major problem constraint and solution (Experiment III). In Experiment II, spontaneous noticing increased when Maier's cord problem and a semantically close analogy to it were used. The effect of surface similarity on the noticing of analogies was discussed in relation to this result. When subjects were given two story analogies prior to presentation of the radiation problem, spontaneous noticing significantly increased (Experiment IV), especially when the same diagram that had been used in Experiment III was presented with the two stories (Experiment V). In the case of two story analogies, the increase in transfer was attributed to the induction of a problem schema describing, at an abstract level, the general constraints and solution employed in the two stories. The diagram facilitated the induction of this abstract schema in Experiment V by focussing attention on the elements common to both story analogies, despite the contextual differences. The relationship between analogical reasoning and schema formation was discussed. It was argued that analogical reasoning between two problems may always involve some kind of abstraction that may be termed schema formation, but may not always involve the application of a well-formed schema at the time the novel problem is solved.