Analysis and Modification of Unprogrammed Reinforcement Contingencies in a Hospital for the Developmentally Disabled

Abstract

The problem: To design a reliable observation system for monitoring of hospital attendants' on-ward interactions with hospital residents and to increase the rate of positive consequences delivered by the attendants for appropriate resident behaviors. Procedure: An on-ward interaction recording system was developed and tested for reliability by camparing observation data with data obtained by trained independent observers. Following baseline observations, six hospital attendants were instructed to record their delivery of positive consequences to residents. If an attendant's rate of delivery of positive consequences increased during this self-monitoring ccndition, that attendant was subsequently returned to baseline observation conditions as a reversal procedure. If an attendant's rate of delivery of positive consequences had not increased during self-monitoring, that attendant was then given feedback and praised for increases in the rate of delivery of positive consequences delivered. All attendants in the self-monitoring-plus-praise condition were then returned to baseline conditions as a reversal procedure. Findings: Results showed an inter-observer agreement of 80% for the on-ward observation system. Five of the six attendants increased their rates of delivery of positive consequences for appropriate resident behaviors when the attendants recorded their delivery of these positive consequences. The other attendant increased her rate when feedback and praise were added. Two of the five self-monitored attendants further increased their rates of positive consequences delivers when feedback and praise were added. Three of the six attendants increased their rates of successful attempts to change resident behaviors which paralleled their changes in positive consequences delivers. Conclusion: It was concluded that a reliable system for monitoring on-ward interactions between attendants and residents can be developed and that self-monitoring procedures can be used effectively with attendants to increase the rates of positive consequences delivered to residents for appropriate behavior. Recommendations: Further research would include the self-monitoring of other interaction behaviors specified in the current interaction analysis system. Longer periods of implementation may increase the effects of the self-monitoring procedure. More extensive training and feedback might increase inter-observer agreement with the interaction analysis system.