BEHAVIOR THERAPY (1974) 5, 381-390 Situational and Historical Determinants of Self-Reinforcement I PAUL I(AROLY2 AND FREDERICK H. KANFER3 University of Cincinnati Magnitude of self-reinforcement (SR+) and self-punishment (SR-) was investigated in a pseudoperceptual task as a function of both the pattern (variable or consistent) of previous external reinforcement and the direc- tion of current, nonspecifie feedback (e.g., "You're getting better" or "You're getting worse" ). Working from a model which assumes SR to be based on an individual's discrimination of the discrepancy between current perform- anee and a pre-set standard, it was hypothesized that a standard (50% ac- curacy) established under consistent prior performances would produce greater magnitude SR+ (with current increments in performance) and SR- (with current performance decrements) than would the identical standard derived from a variable prior pattern of reward. A trials effect was also expected. Results confirmed the hypotheses. Implications for behavior change via self-evaluative and self-reinforcement techniques are discussed. Self-regulation refers to a process by which an individual maintains an ongoing behavioral pattern in the relative absence of immediate dis- criminative or reinforcing stimuli, and is of central concern to investiga- tors of human motivation and socialization (Kanfer, 1970). Personologists of various theoretical persuasions have speculated about the means where- by individuals learn to supplant social regulation; while clinicians have had to deal with the practical problems of training or re-training those with self-regulatory deficits--i.e., delinquents, criminals, "psychopaths," passive (dependent or depressed) individuals, retardates, etc. A behavioral model which has generated considerable research divides the regulatory process into three phases: self-monitoring, self-evalua- tion, and self-reinforcement. According to this model, the self-administra- tion of reinforcing or punishing events follows a discrimination concern- 1This research was supported, in part, by Research Grant MH-17902-03 from the National Institute of Mental Health, U. S. Public Health Service to F. H. K. Tile authors wish to express their appreciation to Robin Welch, who served as experimenter in the present study. 2 Requests for reprints should be sent to Paul Karo]y, Department of Psychology, University of Cincinnati, Cincinnati, Ohio 45221. 3 Now at the University of Illinois. 381 Copyright © 1974 by Academic Press, Inc. All rights of reproduction in any form reserved.