Differential reinforcement using matched examples and nonexamples expedites the practice needed to learn concepts

Abstract

Conceptual learning refers to correctly identifying novel examples and nonexamples of a concept. Concepts can be taught by presenting stimuli and differentially reinforcing responses to examples but not nonexamples. However, it is unclear which stimuli should be used and whether specifically designed stimuli can result in correct responding even in the absence of differential reinforcement. Stimulus selection can be informed by identifying features that a stimulus must have to be an example and can have without altering whether the stimulus is an example or a nonexample. Previous research suggests that using matched examples and nonexamples (the same can‐have features present) may expedite this learning. Experiment 1 compared acquisition and conceptual learning following practice with matched and unmatched examples and nonexamples lacking exactly one must‐have feature. Participants made fewer errors and completed practice faster with matched examples and nonexamples. Both practice conditions produced similar conceptual learning. Some participants also responded accurately during baseline. To evaluate whether the testing stimuli would produce conceptual learning in the absence of differential reinforcement, Experiment 2 compared conceptual learning without any practice. No conceptual learning was observed. The results support using matched examples and nonexamples for additional research on conceptual learning in classroom and laboratory settings.