A study by Bensberg, Colwell, and Cassel has suggested that the profoundly retarded can be taught self-help skills by using a combination of operant and classical conditioning in a one-to-one technician-patient relationship. In this study, the tasks to be taught were divided into smaller, incremental steps for easier acquisition, and correct responses were reinforced systematically. While this study demonstrated the utility of a conditioning approach in the teaching of self-help skills to profoundly retarded patients, it apparently suffered from an important methodological limitation. Bensberg relied on changes in ratings to measure improvement. While their modified Vineland Social Maturity Scale did show such improvement, it represents at best an indirect measure of the development of skills, and one that can be influenced greatly by experimenter or teacher biases. It was felt that more direct and objective measures of profoundly retarded patients' behavior would demonstrate more clearly their ability to learn self-help skills.