Concept learning with differing sequences of instances.

Publisher Website
Google Scholar
Abstract
The present study was an exploratory investigation of the rate of concept attainment under two conditions of presentation of concept instances. The hypothesis studied was that learning would proceed more rapidly under a condition in which the instances of a given concept were presented one after another without interpolation of instances of other concepts as compared with a condition in which the instances of several concepts were presented in an intermixed order. The concept materials consisted of geometrical designs varying in color, size, shape and position. Each concept was defined by a combination of two properties, e.g., large square, or small black object. The Ss were presented one at a time with eight instances (each identified by the same distinctive nonsense name) of each of the four concepts. In Cond. I, all eight instances of a given concept were presented in succession before presenting instances of a second concept, etc. In Cond. n, instances of all 4 concepts were presented in an intermixed order so that no two instances of a given concept were presented in succession without the interpolation of an instance of at least one other concept. At the end of this training Ss were asked to give a verbal description of each concept and to identify several instances of each concept by the appropriate nonsense name. Following the unmixed order of presentation Ss gave both more correct identifications and more correct verbal descriptions of the concepts than following the mixed order of presentation. Only the latter difference was statistically significant (P=.03 for Trial 1 and .01 for Trial 2).
Keywords

This publication has 1 reference indexed in Scilit: