An Experimental Analysis of Analogical Reasoning

Abstract

Previous studies of analogical reasoning leave two important questions unanswered: (1) What does analogical reasoning consist of as a process? (2) What do analogy tests test? The purpose of the studies described in this paper was to investigate the process of analogical reasoning under conditions which control for hypothesized confounding effects of word association. The results suggest that approximately half the analogy items on current tests can be solved by word association; a new analogy test minimally soluble by word association was therefore constructed. Several other extraneous variables which may also be confounded with analogical reasoning were described.

Cue learning, associative responding, and school performance in children.

“…Since the associations that a stimulus element has or acquires are likely to determine whether it will function as a cue, elements already having a common associa-tion structure for subjects of a given age might be the best potential source of cues. Willner's (1964) finding that many analogy items on standard tests can be solved by free association suggested a means for pitting correct abstract reasoning against the tendency to respond to cues on the basis of strong associations. Willner showed that, given only the third word of analogy items as a stimulus, people frequently respond with the correct answer to the analogy as their first association.…”

Reasoning in the chimpanzee: I. Analogical reasoning.

“…One seldom-controlled feature of tests of human analogical reasoning is that the correct answer is often a closer associate of the B term than is the incorrect answer (see Willner, 1964). The present experiment controlled for associations between the B term of the analogy problems and the answer alternatives with a procedure similar to that of Experiment IB.…”

The effect of analogy instruction on young children's metaphor comprehension

“…The puppet uses an arrow as a prop when looking at the analogy cards. I n order to avoid constructing analogy items that could be solved by simple word association -an issue noted by Willner (1964) -possible items for the C term were listed and presented to sixth grade students who were asked to write the first thing they thought of when they read each word. Any pair generated by word association was not used as the final pair of an analogy.…”