Management of chronic psychiatric patients by a token reinforcement system.

A token reinforcement system for the reduction of institutionalized behavior in a chronic psychiatric ward is described. Quantitative assessments were made of the effects of the system on five types of positively reinforced behavior, two types of unreinforced behavior, and two types of fined behavior. To assess whether these effects were a function of the token procedures, three experiments were carried out, the first removing tokens for a brief period, the second making tokens non-contingent on behavior, and the third removing fines in one area while maintaining them in another. The token procedures were found to be the source of the observed improvements.