Journal of Personality and Social Psychology 1969, Vol. 11 No. 1 25 - 33 MAINTENA N CE OF SELF-ATTRIBUTED AND . DRUG-ATTRIBUTED BEHAVIOR CHANGE* GERALD G. DAVISON An d START VALINS State University of New York at Stony Brook It is proposed that behavior changes which are believed to be brought about by oneself will be maintained to a greater degree than behavior changes which are believed to be due to external forces or agents. Within the framework of psychoactive drug therapy, a change in overt behavior which is attributed to cue's own efforts should be more persistent than a change in overt behavior which Is attributed to a drug. An experimental analogue of drug therapy is reported in which subjects (a) underwent a pain threshold and shock tolerance test, (b) ingested a drug (really a placebo), and (c) repeated the test with the shock intensities surreptitiously halved. All subjects thus believed that a drug had changed their threshold performance. Half of the subjects were then told that they had actually received a placebo, whereas the other beef con- tinued to believe that they had received a true drug. It was found that sub- jects who attributed their behavior change to themselves (i.e., who believed they had ingested a placebo) subsequently perceived the shocks as lees painful and tolerated significantly more than subjects who attributed their behavior change to the drug. This research is concerned with a person's attributions about the causes of a change in his behavior. That is, given that a person is behaving in a par ticular manner, will the cognitive and behavioral consequences be dif- ferent if the cause of it is attributed to him- self than if it is attr i buted to an external force or agent? It will be hypothesized that behavior change which is attributed to one- - self will persist or be maintained to a greater degree than behavior change which is at- tributed to an external agent such as a drug. Although this research is restricted to overt behavior, much of it was stimulated by con- ' siderations of covert visceral behavior. If we define behavior as a perceptible physiological reaction, there is good evidence that different attributions about the source of this "internal behavior" will have different effects. Schachter and his associates (Nisbett & Schachter, 1966; Schachter & Singer, 1962) have clearly shown that if this internal behavior is at- tributed to oneself, the consequences will be different than if the behavior is attributed to an external agent such as a dreg. When it is This research was supported by Research Grants MH 12715 and MH 14557 from the National In- stitute of Mental Health. We are deeply indebted to William Samuel and Michael Weiner for their per- sistent and expert performance as experimenters. Thanks are also due to James H. Geer for his come meets on the manuscript. attributed to oneself, the person has some in- formation which rem lead to inferences about himself or about stimuli in the external en- vironment. Thus, if. I have reacted in this anger-inducing situation, then I must be angry. If I have reacted in this euphoric situa- tion, then I must be happy.. If I have reacted to these electric shocks, then they must be painful. The situation is quite different, how- ever, if the physiological reactions or internal behaviors are attributed to an external agent: The person is less able . to make inferences about the external situation or about himself if his internal behavior is drug-induced. Thus, subjects do not get angry or happy (Schachter & Singer , W62) and do not consider shocks painful (Nisbett & Schachter, 1966) if they believe that their internal behavior is drug induced. In effect, a self-attributed visceral reaction makes us stop, look, and listen, and allows us to make certain inferences about ourselves and the stimulus situation. A drug- attributed visceral reaction provides us with little or no information about ourselves or the world. These speculations about attribution are relevant to a psychiatric problem which is particularly serious but which unfortunately is far too often ignored. That is, how can psychiatric patients be weaned off tranquil- izers? The remarkable effects that the intro- 23