Frustration and hostility: the effects of differential reduction in power over decision making

Thesis (Ph.D.)--Boston UniversityThe Frustration-Aggression hypothesis of Dollard, Doob, Miller, Mowrer, and Sears assumes a universal causal relation between frustration and aggression. Since the publication of Frustration and Aggression a number of investigators have indicated that aggression (or hostility as the need system which underlies overt aggressive behavior) is not a simple function of frustration alone, and have cited the non-occurrence of aggressive reactions to frustration in specific situations. Field theorists have argued that the use of "frustration" and "aggression" as unitary concepts may obscure important aspects of the psychological situation in which a given goal blockage occurs. French, following this line, showed that whether or not hostile tensions will arise depends on various properties of the barrier and the frustrating agent. Pastore showed that if the frustrating act is perceived as "arbitrary," greater hostility appears than if the same act is perceived as "non-arbitrary." We here take the position that the arousal of hostility and aggression is dependent on the meaning that the frustrating act has for the person concerned, in terms of the power relationship between himself and the frustrating agent. Results of a series of studies on decision making in groups led us to believe that a major factor governing the arousal of hostility in a situation of social frustration is the degree to which the frustrating act implies a reduction in the ability of the frustrated person to make decisions affecting his and the group's locomotion toward goals. The present research has attempted to verify experimentally this deduction. In the framework of the field theoretical approach, we used the constructs of the Lewinian theory of tension systems in stating our basic postulate: I. Given a situation in which the locomotion of a person, P, toward a goal, G, is blocked by a force induced by another person, Q, to the degree that P perceives Q's action as implying an illegitimate reduction in P's power relative to Q, P will mobilize an inner-personal tension system coordinated to a class of goals involving the reduction of Q's ability to induce forces on him (P). The existence of such an inner-personal tension system would have certain specific effects on the behavior of P, especially in the area of his interpersonal relationships with Q. In a frustration situation defined in these terms, the conflict is between an own force and an induced force. Essentially the only way P can weaken the induced force and resolve the conflict is by going out of the field or by reducing the power of the frustrater to induce such a force on him. Since the experimental situation was so designed that the first of these alternatives was virtually eliminated, any expressions of such an inner-personal tension system would be "channelled" into the area of P's interpersonal relationships with Q. We predicted the following main effects: 1. Hostility of P toward Q will vary directly with the degree to which P perceives Q as illegitimately reducing his power relative to Q. 2. Aggression by P against Q will vary directly with the degree to which P perceives Q as illegitimately reducing his power relative to Q. 3. Amount of shift in P's focus of attention from the task in which he is engaged to personal characteristics and actions of Q will vary directly with the degree to which P perceives Q as illegitimately reducing his power relative to Q. The subjects were 153 volunteer, male, freshman cadets in the Army R.O.T.C. program at the University of Illinois. They were recruited on the basis of an appeal to assist us in evaluating different kinds of teaching methods for Armed Forces training. The Ss were formed into 30 groups of four to six persons each, and each group set up as a miniature classroom in which a teacher, the experimenter, instructed the Ss in how to make a number of paper objects. Ss were told that midway on each of the paper objects, a decision would be made about whether the teacher should go back and repeat his instructions or go on without repeating. The decisions were to be made as the resultant of three factors: the "group's desire," as expressed by the average of Ss' individual votes; the "Armed Forces' desire," as expressed by votes purporting to represent institutional training demands; and the desire to repeat or not expressed by the instructor. We established the expectation of what may be termed a pupil-centered classroom, by assigning a weight of only 1/4 to the instructor's desire, and a weight of 1 each to the desires of students and Armed Forces. At each decision point, these desires, in their different strengths, were posted as vectors on a board in view of all Ss, representing forces for or against repeating. And the decision to "repeat" or "not repeat" was to be arrived at by algebraically summing the forces appearing on the board. In each of the three treatments, the Ss were subjected to the same objective frustration by having the experimental teacher consistently act in opposition to their desires with respect to preferred teaching practice. Although "frustration" was objectively the same in the three treatments used, the degree of legitimacy of the teacher's frustrating action was experimentally varied. The relative strengths of the weighted desires were presented to the Ss so that the teacher was made to appear to be acting legitimately in frustrating them in one treatment. In the other two treatments, the teacher was presented as acting with two different degrees of illegitimacy, i.e., increasing the weight assigned to his own desire, thereby reducing the weight assigned to the students' desire and correspondingly reducing their power over the decision making process. In each treatment, five groups of Ss were run under each of two teachers. Measures on the dependent variables were taken during what subjects thought was a "break" between two halves of the instructional period, and included several questions designed to determine whether in fact the desired perceptions of the situation had been induced. Subjects' perceptions of power reduction agreed very closely with those we had attempted to induce. The variable of hostility was measured by simply having the subjects indicate on a seven point scale how much, if any, hostility they felt toward the teacher for anything he had done up to that point. Our index of aggression we took as the difference in favorability of ratings of the teacher on an adjective check list administered before and after the instructional period. Shift in focus of attention was measured by comparing subjects' reports of non-task-centered items of the teachers behavior which had been identically role-played in each treatment. The hypotheses dealing with main effects were confirmed. For each of these measures, differences among treatments were in the order predicted and significant by analysis of variance at better than the .01 level. There were no differences between teachers. In addition to the above, we tested a number of derivative hypotheses. To determine whether the hypothesized preoccupation with personal characteristics of the teacher in fact interfered with effective learning, we employed two measures. One was a recognition test of sub-assemblies of the various paper objects, the second was a performance test requiring subjects to remake the objects at the end of the experimental period. The mean recognition scores for treatments were in the predicted order and the differences significant at between the .05 and .10 level. With respect to the performance measure, no significant differences were found among the three treatments. There were significant differences among treatments in "rigidity" in problem solving, as measured by an Einstellung test, and these differences were at variance with predictions. An interpretation of these discrepancies was made in terms of a "vigilance" phenomenon. In general, the results were consistent with the existence of the postulated tension system, at levels commensurate with the amount of power reduction perceived by Ss. Under the same objective frustration, the more did Ss perceive the teacher's acts as implying a reduction in their power, the more did they respond with hostility and aggression. We suggest, then, that the Frustration-Aggression hypothesis, useful as a first approximation, must be further specified in order to apply in the richness and variety of social situations. The use of the term "frustration" as a unitary concept, as simple goal blockage, fails to take into account the fact that such a goal blockage will have differential consequences according to the meaning it has for the person concerned, and this meaning in large measure derives from the field or context in which it occurs. The prime component of any situation of frustration, that which governs the arousal of interpersonal hostility and aggression is, we suggest, degree of interference with the power of the frustrated person over decision making processes governing his locomotion in the situation. In any social situation, whether it be a classroom, a work group, a military unit or whatever, a certain amount of frustration of members seems inevitable. To the degree that such frustrations are presented to and perceived by group members as deriving from impersonal situational requirements and implying no arbitrary reduction in decision making ability, we may expect to minimize the extent to which our social organizations will produce feelings of hostility (and possibly related emotional difficulties) in group members