Partial Reinforcement and Noticeability of an Irrelevant Dimension and Their Effects on Subsequent Performance in a Concept-Shift Task

Abstract

The literature suggested that, whereas prior partial reinforcement of a relevant dimension in a concept-shift task has a detrimental effect on learning an extradimensional shift, prior partial reinforcement of an irrelevant dimension has no effect on subsequent transfer performance. The experiment involved increasing the salience of a partially reinforced irrelevant dimension by varying the number of values on this dimension during original learning and determining the effects of this on extra-dimensional shift performance. Subjects were 120 male and female students. Negative transfer occurred. The effect interacted with the number of values present on irrelevant dimensions; results were interpreted in terms of the relative noticeability of the different dimensions within the paradigm.