A comparison of reversal shifts and nonreversal shifts in human concept formation behavior.

Publisher Website
Google Scholar
Abstract
Three experiments were designed to evaluate an analysis of card-sorting behavior which assumed that such behavior on any one trial consisted of a sequence of two successive S-R associations. According to this formulation, the stimulus component of the first association would represent the test cards while the response would refer to the symbolic response made to them and the stimulus of the second association would represent the cue produced by the preceding symbolic response, while the response would be the overt card-sorting behavior. The specific hypothesis tested was that the presence of the appropriate symbolic cues, even though they might be connected to the wrong sorting response, would facilitate concept formation. According to this hypothesis a reversal shift should occur at a more rapid rate than a non-reversal shift because at the completion of the learning of the first concept, the symbolic cues appropriate to the second concept would be present for the subjects in the reversal group; they would merely be connected to the "wrong" sorting response. The experimental results obtained in a situation involving the learning of successive concepts in a multiple solution card-sorting problem were consistent with our expectations. The superiority of a reversal shift over a non-reversal shift was discovered to be independent of (1) the presence or absence of any partial reinforcement effects from previously learned concepts, and (2) whether the concept shifted to was a reverse or a "direct" concept. The results also indicated that a reversal shift produces positive transfer effects.
Keywords

This publication has 1 reference indexed in Scilit: