Performance feedback has been shown to be an effective behavior change technique in a variety of settings, across a broad spectrum of behaviors, and with subjects differing on a number of important dimensions. This study investigated the effects of publicly posted process behavior feedback with goal setting on process behaviors (i.e., behaviors intended to produce a specific outcome) and on the outcomes intended. Conversely, the effects of outcome feedback with goal setting on the outcomes intended and the related process behaviors were examined. Daily patient orientation scores (i.e., outcomes intended) were obtained for a 20% sample of 235 patients on five separate wards in a psychogeriatric hospital over a 22 week period. The number of patient contacts with staff in group reality orientation sessions (i.e., process behaviors) were collected over the same period of time. Process behavior feedback was provided daily to staff on two of the wards while staff on two different wards received feedback on the outcomes intended. Staff of the fifth ward received no feedback for either process behaviors or outcomes intended, thus serving as a comparison group. The results indicated that process behavior feedback with goal setting was effective in increasing the number of patient contacts in reality orientation groups, but had no effect on the outcomes intended (i.e., patient orientation scores). Outcome feedback with goal setting provided to staff members concerning daily patient orientation produced no effect on either the outcomes intended or process behaviors. Process and outcome measures for the control ward remained stable throughout the duration of the study. The results of a post experimental questionnaire assessing staff and patient satisfaction with the intervention are presented. The results of the study are discussed from the perspective of Gilbert's (1978) performance engineering model. Implications concerning the design of feedback programs within human service and production oriented agencies are presented. Limitations of the study are described and recommendations are made for future research in the area. / Ph. D.