The effects of single instance, multiple instance, and general case training on generalized vending machine use by moderately and severely handicapped students.

This report provides an experimental analysis of generalized vending machine use by six moderately or severely retarded high school students. Dependent variables were training trials to criterion and performance on 10 nontrained "generalization" vending machines. A multiple-baseline design across subjects was used to compare three strategies for teaching generalized vending machine use. Training occurred with (a) a single vending machine, (b) three similar machines, or (c) three machines that sampled the range of stimulus and response variation in a defined class of vending machines. Results indicated that the third approach was the most effective method of obtaining generalized responding. Methodological implications for the experimental analysis of generalization and programmatic implications for teaching generalized behaviors are discussed.