Stimulus Control by Relevant, Irrelevant, and Redundant Components of Complex Stimuli as Assessed by Two Testing Methods

Abstract

Six pigeons learned a successive discrimination where S+ was a vertical line on a green background, and S− was either (a) a green background, (b) a red background, or (c) a vertical line on a red background. Generataon of responding to the angular orientation of line on green, dark, and red backgrounds was evaluated. Also, a component stimulus test was conducted in which green, line on green, red, line on red, dark, and line on dark, appeared separately. Steep generalization gradients were obtained from all Ss when angular orientation was varied on the green background, and similar gradients were obtained from all but one S when the background was dark. These results indicated stimulus control by line for Ss in all groups. However, results of the component stimulus test indicated clear stimulus control by line only in the case of the two Ss trained under condition (a) above, and one of the Ss in condition (b). Component stimulus tests for the other three Ss produced no clear evidence of stimulus control by line. Of the two testing methods used to determine stimulus control, stimulus generalization testing appeared to be the most sensitive and easiest to interpret.