r', -s _.‘ .,u ‘v . - . : x J’.'Ih‘a’ 1"J'. H lk‘a mfgsnréu m ra- H 4w , ,p-h. .0. . S 32 .,.. Es ‘. .» ' ‘ ..E.... .. ' 'y... 1... v..— ¢ " v- u 1‘5 x - 'r‘.v - .~ .c a r-..- ._._. .,...vv'__.‘“vv__ THE ROLE OF ENCODING AND. DECODING ~ ~ AGGRESSIVE MESSAGES on SUBSEQUENT Hosnuw ' » : Thesis far the Degree. of Ph. D. MmmGAN STATE UNIVERSITY ROGER D. HANEY 1971 ' LIBRA R y Michigan 3 up Univcmis , This is to certify that the thesis entitled The Role of Encoding and Decoding Aggressive Messages on Subsequent Hostility presented by Roger D. Haney has been accepted towards fulfillment of the requirements for PhoDo degree in Communication 1 m. mil: motes-y l Date February 25, 192] 0-7639 y amoma av "’ HUM} & SDNS' ABSTRACT THE ROLE OF ENCODING AND DECODING AGGRESSIVE MESSAGES ON SUBSEQUENT HOSTILITY BY \ Roger D. Haney This study investigated the role of decoding and encoding aggressive messages on subsequent feelings of hostility. Previous research suggested that reading aggressive messages could increase hostility, while writing such messages could reduce the aggressive drive instigated by an act of frustration. Hypotheses were based on those premises. A frustration manipulation consisting of insults given by an experimenter while administering a bogus proficiency test was pretested and found to be effective. It was used in six experimental conditions with 111 subjects. In two decoding conditions, subjects were either asked to read a message negatively evaluating the source of frustration (Decode—Specific) or Vice-President Agnew (Decode-Nonspecific). In these encoding conditions, subjects were either asked to write a negative evaluation of the source of frustration (Encode-Specific), of Vice-President Agnew (Encode-Nonspecific) or write a Roger D. Haney message concerning the failure of a recent student strike on campus (Encode-Nonaggressive). In the sixth condition, there was no encoding or decoding activity after the frustration induction (Control). The main dependent variable was a five-scale index of attitudes toward the frustrator and the experiment (Evaluation Index). Two secondary indices consisting of items selected from the Buss-Durkee measure of general hostility were also administered. In general, both decoding and encoding led to significantly greater hostility than no activity. Only the Decode-Specific condition did not show more hostility. These results support a facilitation of aggression hypo— thesis and do not support a catharsis hypothesis. Predictions made about various types of encoding and decoding were not confirmed. Accepted by the faculty of the Department of Communications, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Doctor of Philosophy degree. 3% Director of hesis 0 V n l , Guidance Committee: 'WK.J‘ ~ "Vw ‘a”“1 ,Chairman F z/efl/i/ THE ROLE OF ENCODING AND DECODING AGGRESSIVE MESSAGES ON SUBSEQUENT HOSTILITY BY 0 02 (\ Roger D? Haney A THESIS Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1971 The research upon which this dissertation is based was performed pursuant to Contract No. HSM 42-70-32 with the National Institute of Mental Health, Health Services and Mental Health Administration, Department of Health, Education and Welfare. ii ACKNOWLEDGMENTS The author wishes to express his sincere indebted- ness to those several people who helped him throughout his graduate career. First and foremost is his wife Joyce who by her constancy and patience showed that love is indeed a behavior. Special thanks go to Dr. Bradley S. Greenberg for his considerable expertise and direction, but most importantly for his time when there were considerable demands on it. Dr. Erwin P. Bettinghaus receives the writer's thanks for his appreciation of a linguistic project more important in its pragmatics than its semantics. Thanks go to Dr. Verling C. Troldahl and Dr. Joseph F. Hanna for their respective competence in methodology and philosophy of science. Ed Wotring is commended for his able assistance as an experimenter in the administration of a frustrating study. Finally, the author wishes to express his sincere gratitude to Drs. William J. McEwen and Joseph R. Dominick, fellow graduate students, for their meaningfulness and friendship, and to Dr. David K. Berlo for his significant contribution to the communication program and his expressed interest and understanding toward a candidate in need of both. iii TABLE OF CONTENT LIST OF TABLES O O O O O 0 ~ 0 O O O I LIST OF APPENDICES . . . . . . . . . CHAPTER S I. THE RESEARCH PROBLEM.. . . . . . . . . Introduction . .w. . . . . The Frustration-Aggression- The Berkowitz Paradigm . . Catharsis . . . . . . . Catharsis Research . . . Hypotheses . . . . . . . II. RESEARCH METHODS . ... . . . Overview . . . . . . . Subjects . . . . . . . Procedures . . . . . . Antecedent Condition: Frustration cocoa}. YP oth General.Hostility Inventory . Experimental.Conditions.-. Dependent Measures . . . . III 0 RESULTS 0 O O C O O O O O O 0 IV. DI Instigated Aggression Hypotheses Catharsis.Hypotheses . . . Intervening Task Hypothese Frustration_0nly Comparisons . General Hostility Comparisons. SCUSSION . . . . . . . . . Methodological Issues. . . Sampling. . . . . . . . Measurement . . . . . . Theoretical Implications . BIBLIOGMPHY O O . O O O O O O O O C . APPENDICES iv 8888. Page vi LIST OF TABLES Pretest Results: Mean Experimenter Ratings . . . . . Pretest Results: Hostility Inventory . . . . . . . . Instigated Aggression HYPOtheses: Evaluation Index. . Catharsis Hypotheses-eEvaluation Index. . . . . . . . Intervening Task Hypotheses--Evaluation Index . . . . Frustration Only--Eva1uation Index. . . . . . . . . . Instigated Aggression Hypotheses--Hostility Inventory Catharsis Hypotheses--Hostility Inventory . . . . . . Intervening Task Hypotheses—-Hostility Inventory. . . Summary of Results--Evaluation Ratings. . . . . . . . Page 24 24 34 35 36 37 39 40 42 43 LIST OF APPENDICES Page APPENDIX A. Experimenter Evaluation . . . . . . . . 57 B. Hostility Inventory . . . . . . . . . . 58 C. Insult Technique. . . . . . . . . . . . 60 D. Encoding Manipulations. . . . . . . . . 62 E. Decoding Manipulations. . . . . . . . . 63 vi CHAPTER I THE RESEARCH PROBLEM Introduction The purpose of this study is to determine the relative effects of message encoding versus message decoding on "feelings of aggression," or hostility. Previous research has tended to investigate the role of the mass media within a frustration--aggression paradigm, arguing that communication behavior (usually in the form of "viewing" behavior) can serve as a model for aggressive acts (Bandura, Ross, and Ross, 1961; 1963 ab) can facili- tate aggressive responses, given appropriate cues (Berko- witz and Rawlings, 1963; Berkowitz, 1964, 1965; Berkowitz and Geen, 1967; Geen and Berkowitz, 1966), or can reduce feelings of aggression (Thibaut and Coules, 1952; Feshbach, 1955, 1961; Berkowitz, 1960). The question, then, is under what conditions aggression is most likely to occur and the role certain message processes play in either facilitating that aggression or reducing it. The Frustration-Aggression Hypothesis Many psychologists studying the phenomenon of aggression tend to regard it as a response to some 1 frustrating act, the classic formulation being provided by Dollard et_31, (1939). They argued that a "frustration" is "an interference with the occurrence of an instigated goal-response at its prOper time in the behavior sequence" (p. 7). Thus frustration is regarded as interference with or blockage of some drive state and can occur in a variety of ways. Brown and Farber (1951, p. 481) argued that frustration can occur when there are (1) physical barriers, (2) delays between the initiation and completion of the response sequence, (3) omission or reduction of a customary reward, or (4) the eliciting of a response tendency that is incompatible with the ongoing one." Given such a formulation, certain implications are paramount, many of which have been confirmed in the liter- ature. The strength of the frustration should be greater when it occurs closer to the goal (Buss, 1961). Frustra- tions should be additive and whether or not they are "expected" should make a difference. Thus Pastore (1952) found that the "arbitrariness" of the frustration intervenes in the frustration-aggression hypothesis. Unjustified attacks are more frustrating (not to be confused with justified aggression) than justified attacks.1 McClelland 1Buss (1961) separates physical and verbal attacks (insult) from frustration proper. Berkowitz (1962) argues that attack does fit within a frustration framework as explicated by Brown and Farber. The present paper follows this latter position in that a drive toward some goal (e.g. homeostasis) is interfered with. . . (I) t I. O (I, II! II. (I) to: H. . IP- #5:. . In. - ~Sui I .‘Hl ‘. Us 5“... d1 and Apicella (1945) found that more intense (derogatory) insults led to more intense subsequent aggression. This led Buss (1963) to investigate the proposition that some types of frustration are more effective than others in leading to aggression. "Aggression" was defined by Dollard, gt_gl, as a "sequence of behavior, the goal-response to which is the injury of the person toward whom it is directed" (p. 9). The behavior need not be overt and may be direct or indirect. To avoid the teleological problems that result if "intent to harm" is the implication of such a definition, Buss (1961), emphasized the reinforcing aspect of aggressive behavior. This can occur either in the form of the stim- ulation provided by the victim suffering injury or being in pain, or from extrinsic rewards such as money gained from a mugging. Buss defined physical aggression in terms of its consequences: either (1) pain or injury or (2) the overcoming of a barrier or the source of noxious stimuli. Verbal aggression is similarly defined as a "vocal response that delivers noxious stimuli to another organism" (p. 6). Noxious stimuli in this latter case are such things as rejection and threat, rather than pain or injury. Berkowitz, in his formulation (1962) avoided an emphasis on intent by arguing that aggression following a frustrating experience may not so much be "pushed out" by strong emotions as "pulled out" by apprOpriate cues in the aqfl‘ ”AI. 1“ ‘ «on .1- IARV l I'V‘ p. I it n D 0'“: 00‘! van h. ' - bu. n‘- u“: I. \ E h" environment. Much of Berkowitz's research has been concerned with determining the nature and efficacy of the "appropriate cues." Dollard and his associates argued that "the occurrence of aggression always presupposes the existence of frustration and, contrariwise, that the existence of frustration always leads to some form of aggression" (1939, p. 1). In other words, frustration, in some form was taken to be the necessary and sufficient condition for the occurrence of aggression. Miller (1941) qualified this formulation by recognizing that responses other than aggression can occur after a frustrating instance (fear of flight, for example).1 Miller suggested that a better phrasing of the statement is "frustration produces instigation to a number of different types of responses, one of which is an instigation to some form of aggression" (1941, p. 338). Thus frustration is no longer regarded as a sufficient condition for the instigation to aggression. Nor is frustration a necessary condition for the instigation to aggression. Berkowitz (1962) pointed out that past reinforced aggression can be sufficient, i.e., 1Carmichael (1965) even found that attitude change can be affected by frustrating experiences. Subjects were more persuaded by speeches attacking their own course and another course when they had been frustrated by an aspect of the course. Interestingly, their opinion of the source of the speech went down, even though they were persuaded by his advocated position. . «up “bl .0; a ”a“ “a.“ ‘u u c L I“: O (ll that aggression can be learned. As an example, the behavior of soldiers during wartime is cited. Bandura, Ross, and Ross (1961) found that children exposed to adult models behaving aggressively tend to imitate that behavior. This could occur whether the adult was on film, dressed as a cartoon character, or actually in the room (1963a), and was more likely to occur when the model was rewarded for his aggressive action. Buss (1961) pointed out that attack (insult) is used most often in the laboratory situation to instigate aggression. While this can fit within a frustration framework, the notion does lead to an interesting variable that is said to intervene in the frustration-aggression hypothesis: anger. Brown and Farber first proposed that the emotional state produced by the frustration.can be regarded as a motivation. Anger is said to be a drive which heightens the likelihood of aggressive behavior. In the Berkowitz formulation, "drives such as anger do not lead to the drive-specific behaviors (aggression in this case) unless there are apprOpriate cues or releasers" (1962, pp. 32-33). Thus for Berkowitz, a frustration creates a predisposition toward hostility by arousing anger. Whether or not aggressive responses occur, however, is dependent in part upon the presence of suitable aggression-evoking cues. The strength of the aggression is considered dependent upon "the intensity of the result— ing anger and the degree of association between the instigator and the releasing cue" (p. 33). Berkowitz feels that there must be some elicitory cues present-- stimuli which are associated with the previous frustration. Thus while frustration can lead to anger, the anger may not be released in the form of aggression unless other cues or stimuli are present. The observation of aggression, after frustration, is a one that has been investigated in this regard. The Berkowitz Paradigm In the typical Berkowitz experiment, subjects are first separated into two groups: Insult and Non-Insult. This serves as the frustration manipulation. The emotional state resulting from the frustration is termed "anger" and is said to create a proclivity for an aggressive act.l Subjects then view either a violent film segment in which one of the combatants is severely beaten or a non-violent film. The efficacy of various cues relating the film to the frustration act are then tested. The dependent variable is some measure of aggression and most often consists of the number of electric shocks the subject believes he is 1This is similar to the definition of hostility offered by Buss (1961, p. 12): an "implicit verbal response involving negative feelings and negative evalu- ations of people and events." administering to the frustrating agent. In general, Berko- witz has found that observing an act of aggression that has cues relevant to the source of frustration leads to an increase in aggression. Berkowitz and Rawlings (1963) provided a synopsis of the movie which either depicted the protagonist, who suffered the beating, as a villain in a more sympathetic manner--someone who was not really bad. Tharreasoned (p. 411): If the villain is defeated or punished aggressively--if he obtains the beating he merited, as is typical in most melodramas--we clearly have a case of justified aggression, and this type of fantasy violence may actually increase the likelihood that.some recently angered member of a movie or TV audience will attack his own frustrator, or perhaps even some innocent people he happens to associate with the anger instigator. Seeing the fantasy vil- lain 'get what he deserved? may make the angered individual more inclined to hurt the villain in his life, the person who angered him. They found that subjects in the Insult-Justified fantasy aggression condition showed significantly more unfriendliness to the experimenter (frustrator) as indicated by relatively greater agreement with the statement, "My attitude toward this task might have been better if there had been another experimenter instead of Mr. ' ." Berkowitz, Corwin, and Heironimus (1963) replicated the study, adding a third condition in which subjects viewed a "neutral" film about canal boats. They found similar results in the relevant conditions. However, angered sub- jects who were exposed to the nonvillain summary did i dun. ‘ 9. ngt_show more hostility than those who saw the neutral film. Thus it seems that exposure to filmed aggression may not by itself produce more aggression (in college students) than exposure to neutral films. In further experimentation, Berkowitz reasoned that cues which link the victim of a beating to the frustrating agent in some way can serve to lessen inhibi- tions against subsequent aggression. Berkowitz (1964) found that subjects gave a greater number of shocks to the frustrating agent when they were told he was a "boxer" (like the film's protagonist) than when they were told he was a "speech major." Berkowitz and Geen (1966) found more subsequent aggression when the frustrator was intro- duced as having the same name as the boxing victim in the film. Geen and Berkowitz (1966) found that the highest level of aggression occurred if the frustrator was associated by name to the boxing victim rather than the boxing victor. Berkowitz and Geen (1967) found that this occurred even if the name-mediated association was formed a££g£_the film was viewed. In summary, if subjects are angered through an act of frustration (insult), they will exhibit more aggression toward the source of frustration when they view an aggres- sive act that has been justified or linked to the frustra— tion source. This link can be formed by associating the frustrator and the victim of the observed aggression by occupation or name. 'Catharsis The concept of catharsis has its origins in the analysis of Greek drama. It was argued, principally by Aristotle, that viewing great tragedy serves to discharge one's own emotions. Freud and Breuer (1893) introduced the term in their study of hysteria to refer to the tension-reducing consequences of emotional expression. They found that symptoms of hysteria would disappear if under hypnosis (and later in the free-association tech- nique) the patient recovered the traumatic memory of a past difficulty and described the disturbing event in detail. "The injured person's reaction to the trauma only exercises a completely 'cathartic' effect if it is an adequate reaction--as, for instance, revenge. But language serves as a substitute for action; by its help, an effect can be 'abreacted' almost as effectively" (Breuer and Freud, 1893, p. 8). Dollard et_213 (1939) emphasized the "adequate reaction" in their theoretic formulation of aggression. They felt that "the expression of an act of aggression is a catharsis that reduces the instigation to all other acts of aggression" (1939, p. 53). This differs from the Aristotelian notion of catharsis in that the aggression act is not vicarious. For catharsis to occur, the aggres— sive act must be expressed by the individual rather than viewed by him. 10 This difference can be explicated in terms of encoding versus decoding. One can express an aggressive act physically or verbally. If it is done verbally, either in oral or written form, this is an act of encoding. Decoding, on the other hand, describes the situation where a person might view, or hear, or read, about someone else's aggression. An act of aggression is not performed by the individual himself. Thus there are two basic formulations of the catharsis hypothesis. The first argues that the wit- nessing of aggression (an act of decoding) can serve to vicariously reduce the instigated drive for aggression. The Berkowitz research most directly tests and refutes this form of the hypothesis. The second formulation argues that the expression of aggression (an act of encoding if aggressive messages are considered) can serve to mitigate the drive for further aggression. This expression of aggression can take place either in linguistic (Breuer and Freud) or physical (Dollard, gt_gl.) form and supposedly may be direct or indirect (Buss, 1961). It is important to note that Berkowitz has not generally tested this form of the hypo- thesis in that he has not tested for aggression after the subject's expression of aggression. In one study that did have post-aggression measures, however, Berkowitz and Holmes (1960) found that angered subjects who gave the 11 most shocks later expressed significantly more favorable judgments of the frustrator's fairness. This finding supported the earlier research of Pepitone and Reichling (1955). Catharsis Research Much of the research purporting to find catharsis has tested for it in the second sense of the formulation. There it is argued that the expression of aggression can be cathartic. Thibaut and Coules (1952) tested the hypo- thesis that "the communication of hostility through overt aggressive behavior directed toward a (personal) instigator will tend to reduce the residual hostility toward that instigator" (p. 770). They had subjects exchange a series of notes with a confederate and on the last exchange the confederate insulted the subject. Those subjects allowed to respond to this last message expressed a significantly greater number of friendly items in a subsequent person- ality sketch of the confederate than those who were not allowed to communicate back. They could not tell from this study, however, if tJIere was a catharsis effect in the communication group (fir an increase in hostility in the no-communication group. TR: test for these differing possibilities, Thibaut and chaules conducted a second experiment which differed from 'Ehe first in that the no-communication group was allowed 12 to respond to the insult message after a three-minute delay. Thoaewho had to wait_before responding expressed more hostility toward the frustrator than those who.could respond immediately. This indicated that "the thwarting of communication back to therinstigator immediately after instigation increases the'level-of hostility" (p. 773.) The authors concluded that catharsis does not occur when a source of frustration (communication delay) remains. In another study, Rosenbaum and de Charms (1960) found that college men who were low in self-esteem expressed less resentment toward a peer-frustrator after they heard someone else attack him, especially when they themselves had been given an opportunity to communicate back to the instigator. The authors qualified their catharsis interpretation, however, in that the statements made in rebuttal to the instigator were invariably mild. The felt that inhibition to aggression may have been a factor for low self-esteem people. In perhaps the most cited study showing cathartic effects, Feshbach (1955) first angered subjects and found that those who had the opportunity to express their hostility in a fantasy task using TAT cards expressed less subsequent hostility on a final questionnaire. Thus the eXpression of hostility served to reduce hostility toward the frustrator. 13 In a study preceding the Berkowitz research, Feshbach (1961) first angered subjects and then showed them a film of either a violent prize—fight or a more neu- tral film. Those who saw the fight had less hostility as measured by subject evaluation of the frustrator. Feshbach hypothesized (p. 381) that participation in a vicarious aggressive act results in a reduction in subsequent aggres- sive behavior if aggressive drive has been aroused at the time of such participation; if aggressive drive has not been aroused at the time of participation in a vicarious aggressive act, such participation results in an increase in subsequent aggressive behavior. He felt that the aggressive film served to reduce the aggressive drive instigated by the frustration. This is consistent with the argument that "the most important determiner of the cathartic effect is the presence or absence of anger" (Buss, 1961, p. 89). Such a formulation does not seem adequate, however. Berkowitz (1962, p. 220n) argued that anxiety or guilt was instilled in the subjects by watching the film. Subsequent aggression may not have occurred because subjects felt guilty rather than because the aggressive drive was reduced. Another plausible explanation is that the aggressive drive in the subjects was inhibited by the fact that they thought their evaluations of the frustrator were to be shown to the department chairman in order to evaluate the experimenter’s competence. .yan no... . ivu‘\ 'AQ \b Iii I ‘V pl ":vu "~u. 'l' h V i “I -‘~. 9. 14 Moreover, Hartman's 1969 study contradicted these results. Adolescent delinquents who saw either of two versions of a fight scene were more aggressive (as measured by duration and intensity of administered shocks) if they previously had been angered. This effect was greater when the film focused on the pain of the victim rather than the actions of the aggressor. Apparently pain served as a secondary reinforcer for subsequent aggression. It seems, then, that it is not simply the absence or presence of anger that determines the cathartic effect. It may be that the subject has to actively participate, e.g. encode, in the aggressive act, either physically or verbally, for catharsis to occur. Passive participation, e.g. decoding, seems insufficient. Thibaut and Coules (1952), Feshbach (1955) and Rosenbaum and de Charms (1960) support this more complex catharsis rationale. Hypotheses Early investigators of aggression theories proposed that aggression is the response to some frustratingact. It was soon realized that frustration is not sufficient for aggression. Other factors are necessary. Berkowitz (1962) argued that there must be other appropriate cues or stimuli in the environment which "pull out" the aggressive act. He and his associates tested the efficacy of various cues in filmed aggression. They argued 15 that if a subject is frustrated (usually through an insult technique) and then observes some aggressive act, this will lead to greater subsequent aggression. This effect will be greater to the extent that the observed aggression is relevant to the source of frustration. This argument, however, is contrary to a catharsis formulation. Other investigators, notably Feshbach, have argued that if a subject is frustrated (angered), observed aggression will vicariously cathart the drive for subsequent aggression. Past research gives little support to this formulation, however. The drive for aggression seems only to be reduced through some personal act of aggression by the subject. These possibilities can be tested for by comparing the processes of decoding versus encoding of aggressive messages. Decoding an aggressive message does not allow for the expression of aggression on the part of the subject. Thus if a person decodes an aggressive message, one would expect greater subsequent aggression. This effect should be accentuated to the extent that the aggressive message is relevant to the source of frustration (as in the Berkowitz formulation). This leads to the statement of the first three hypotheses. Hla: For angered subjects, decoding an. aggre881ve message that IS spec1f1- cally directed toward the source of anger results in a higher level of hostility than no activity. 16 Hlb: For angered subjects, decoding an aggressive message that is not specifically directed toward the source of anger results in a higher level of hostility than no activity. 1c: For angered subjects, decoding an aggressive message that is specifi- cally directed toward the source of anger results in a higher level of hostility than decoding an aggressive message that is not specifically directed toward the source of anger. It is predicted that reading an aggressive message will elicit cues which lower inhibitions against expres- sions of hostility. A message that is directed specifi- cally toward the source of anger will be most effective in this regard. Research has been cited which supports the notion that catharsis can take place. Feshbach (1961, p. 381) argued that "for an activity to have drive reducing properties, components of the drive must be present or evoked during performance of the activity; that is there must be some functional connection between the vicarious act and the original drive instigating conditions." While it is hypothesized here that decoding an aggressive message will increase feelings of hostility rather than reduce the drive for aggression, it seems that an act of aggression committed by the frustrated subject would be drive-reducing. This would be so to the extent that it was directed toward the source of anger. Such an act can take place as an act of encoding. To the extent that . ”A" H _.II' If F .. ., .unr. \ .uv» b. ‘1‘ neg: “in EVA“ '0 it i h on”, '5‘. l ‘.U.‘ 17 actively constructing an aggressive message is drive- reducing, it may serve to reduce feelings of aggression or hostility. Research purporting to show a cathartic effect supports such a position. Only the Feshbach (1961) study did not involve some aggressive encoding behavior on the part of subjects. Haer (1968) also supports such a proposition. He found that messages encoded by psycho- therapeutic patients had fewer aggressive remarks (deroga- tory, critical statements) after an expression of anger ("I am mad") than before such expression. Haer felt that release of anger reduced the drive for expression of aggression. It seems equally likely that the expression of aggression itself reduces the drive for further aggres- sion. This leads to the following hypotheses: H2a: For angered subjects, encoding an aggressive message that is directed toward the source of anger results in a lower level of hostility than no activity. For angered subjects, encoding an aggressive message that is not speci- fically directed toward the source of anger results in a lower level of hostility than no activity. 2b: For angered subjects, encoding an aggressive message that is directed toward the source of anger results in a lower level of hostility than encoding an aggressive message that is not directed toward the source of anger. 2c: 18 These hypotheses state that encoding an aggressive message will serve to reduce feelings of hostility, especially if the message is specifically directed toward the source of anger. However, this does not determine if the predicted reduction in hostility is due to the aggres- sive content of the encoding or the act of encoding itself. Perhaps any act of encoding may serve to reduce levels of hostility, in that it serves as an intervening activity following the occurrence of anger. However, it is unlikely that such activity would be as effective as that which allows release of a drive (as in the aggressive encoding situation) or directly reduces the drive (as in the aggres- sive relevant situation). This leads to the final three hypotheses: H2d: For angered subjects, encoding an aggressive message that is directed toward the source of anger results in a lower level of hostility than encoding a non-aggressive message. 2e: For angered subjects, encoding an aggressive message that is not speci- fically directed toward the source of anger results in a lower level of hos- tility than encoding a non-aggressive message. Hzf: For angered subjects, encoding a non-aggressive message results in a lower level of hostility than no activity. In summary, it is predicted that if subjects are frustrated (angered) through an insult technique, they will exhibit a greater amount of subsequent hostility toward the -.. .‘ v lune 1... I on. “c ‘V is ...',._ . ~ I '* ‘. :AH" 'Vigi "i" -" b. :'RA 9“": of, L‘, i. ‘Q IA‘ 0 (IF 19 source of frustration when they decode an aggressive mes- sage than when they do nothing (Hypotheses la and lb). Furthermore, those subjects who decode an aggressive message concerning the source of frustration will exhibit greater hostility than those who decode an aggressive message concerning someone not related to the source of frustration (Hypothesis 1c). Subjects who encode an aggressive message on the other hand, will exhibit l§§§_hostility than subjects who do nothing (Hypotheses 2a and 2b). Furthermore, those subjects who encode an aggressive message concerning the source of frustration will exhibit less subsequent hostility toward the source of frustration than those who encode an aggressive message concerning someone not related to the source of frustration (Hypothesis 2c). Finally, subjects who encode an aggressive message, specifically directed toward the source of frustration or not, will exhibit less subsequent hostility than those who encode a nonaggressive message (Hypotheses 2d and 2e). Subjects who encode a nonaggressive message will exhibit less hostility toward the source of frustration than those who do nothing (Hypothesis 2f). 20 These nine hypotheses have the following order on an aggression scale. Minimum'Aggression Encode - Specific /\ Encode - Nonspecific Encode - Nonaggressive Control (Time) Decode - Nonspecific Decode - Specific \V Maximum Aggression CHAPTER II RESEARCH METHODS Overview Subjects were frustrated by one experimenter. A second experimenter then induced one of six experimental manipulations. An experimenter evaluation measure and a hostility inventroy were administered afterwards. Groups were compared by t_- test. Subjects A total of 118 subjects were obtained from five business letter writing classes and one introductory com- munication class at Michigan State University. In addition, one business letter writing class and two introductory communication classes were used to determine the efficacy of the frustration manipulation in pre-test situations. Data were collected over a three week period. Procedures Antecedent Condition: Frustration Several methods have been used to frustrate sub- jects. Feshbach (1955) and Worchel (1957) used an insult 21 22 technique whereby one experimenter made disparaging remarks towards students while they performed a bogus task. This method has the advantage that it can be used on a group and was chosen for the present study. I In the first manipulation check, all subjects filled out a personality inventory that contained 12 items tapping general hostility. These were items selected from the aggression scales of Buss-Durkee (1957) and Sears (1961). The composition of this measure is described later in this chapter. Subjects then took a bogus pro- ficiency test. In the frustration condition, this task was introduced by the experimenter in an insulting, demean- ing manner. For example, the frustrator said, I realize college students seldom give a damn about anything concerned with knowledge unless a grade is attached, but I would hope that this would be an exception. All subjects then filled out a second personality inventory which contained the same hostility items. No differences were found on either the hostility items or the experimenter evaluation. However, subjects had not been explicitly told to evaluate only the frustrat- ing experimenter. Evaluations of the second, more positive, experimenter may have contaminated the results. For this reason, a second manipulation check was undertaken. Both the evaluation sheet (See Appendix A) and the hostility inventory (See Appendix B) were revised, the latter being *eduo 1 1'9"". .9: Hub “Q. ' I!“ Quay .-A a“: ‘M 23 reduced to six items. The insult technique was also revised and the test-retest procedure was eliminated. In the second manipulation check the class instruc- tor left the room after introducing the two experimenters. The frustrating experimenter (the author) then introduced the study as an attempt to correlate certain personality characteristics with how well people perform on proficiency tests. He explained that grades would not be affected but asked students to try hard anyway, feeling that "students don't give a damn about research." In general, the frus- trator questioned their motivation with a series of insulting remarks. The complete insult technique is in Appendix C. In the manipulation check, the second experimenter then handed out the personality inventory containing the hostility items. When this was completed, the experimenter evaluation sheet was handed out as a standard procedure in the Communication Department. The second experimenter also reminded them that only the first experimenter was to be evaluated. The results of the second manipulation check appear in Table l and Table 2, on page 24. All items concerned with a rating of the exper- imenter and the experiment showed a difference in hoStility. In the frustration condition, the experimenter was rated as having performed less satisfactorily and being 24 mo.v a pm so.~uu AN5.NI.D.mV Nm.HH Amb.mu.n.mv m¢.w 15.525 3.53; Amm.mu.a.mv mo.HH AHh.Nu.Q.mv om.m xmccH vasommm xoonH ucoeunommm cowumnumdhm cowpmnumSHMIcoz muonso>cH hpflaflnmom ”muaommm amououm .N manna mo.vm .mmumn .m~.~uu. malz mauz Anm.¢u.o.mv Aum.mu.o.mv . mm.ma «hm.oa Ammumv xoonH mh.a om.H Amnav noummmou Hon Hooucnao> mm.m mm.a “mlav omwmaamno Bo: mm.m em.H Amlav maonommmflnmm Enemnmm m mm.a mm.H ANIHV mxuamuc assesses mH.m mo.m AmIHV mawcssuuoz 30m ma.~ mm.~ Amuflc mauumauonunma mesa nose 30m coaumnumoum cowumnumnnmncoz mmcmm mfionH mmcflumm Hmucmfiwnomxm cams "mpHsmom ummuon .H manna 25 less qualified. The experiment was rated less worth- while, subjects were less likely to have liked parti— cipating in the study and were less likely to volunteer .for future research with the experimenter. An index constructed of these items showed significantly more Inostility (negative evaluation) in the frustration condition than in the non-frustration condition. The item "Was there anything you disliked about this research?" was not used in the index since it did not differentiate conditions. Thus the range of the index was from 5-23, 23 indicating greatest hostility. This index was the principal dependent measure in 'this study. The mean differences were large and significant. General Hostility Inventory A second set of measures was also used in the (study. These were six items selected from the aggres- ssion scales of Buss-Durkee (1957) and Sears (1961). {Pwo items, At times I feel that I get a raw deal out of life. When I look back on what's happened to me, I can't help feeling resentful. were selected from Buss-Durkee items and formed a Egasentment Index. Both items had seven steps varying from "strongly agree" to "strongly disagree." 26 Three other items, You have to stand up for your rights--even to the extent of fighting--if you want to get along in the world. I can often think of a good reason for hitting someone. If an older boy is mean to a younger one, the younger one has a perfect right to get even, even in some secret or sneaky way. formed an Assault Index. These items also consisted of seven steps and varied from "strongly agree" to "strongly disagree." In both submeasures, the higher the score the more aggressive the response. As Table 2 indicates, subjects in the frus- tration condition did not indicate significantly more aggressiveness on the hostility measures than in the non-frustration condition. Results on these indices were therefore regarded as exploratory and are reported separately from the Experimenter Evaluation measure. The entire experiment depended on the success of the frustration manipulation. Therefore, support- ive evidence independent of the evaluation ratings is provided. First, during the manipulation all classes showed a reaction to the frustration. Many glared at the frustrator during the manipulation and handed back the proficiency tests in a general spirit of uncooperativeness. While this does not constitute 27 direct evidence of anger, it does indicate a state of high arousal on the part of the subjects. Second, several subjects made comments on the evaluation sheets. "I felt like I should salute the man," "Is he going to come back with a rubber hose," "This guy doesn't instill much coOperation in students," and "Doesn't seem to think much about research" are typical examples. Third, during the debriefing many subjects indicated a suspicion that "something was going on." Only in the Decode-Specific group, however, did they connect the evaluation sheet to the frustration manip— ulation. More will be said of this later. All other groups indicated that "something had happened" which bothered them but they didn't know exactly what or why. They did indicate that the behavior of the frustrator was "unusual" and perhaps "put on," but they seemed offended by that as much as anything else. In no class did anyone indicate awareness that the general hostility measure was in any way related to the frustration manipulation. Fourth, the encoded messages were examined in terms of aggressiveness. All the encoded messages about the frustrator (Encode-Specific) were highly negative. They felt the frustrator "knows little about motivation," is "the worst research conductor 28 that I've ever been present with," "treated (us) like dirt," and "What a dink." Seven messages did indicate that the whole thing might have been planned, but these people seemed upset and did not indicate awareness of the reason for manipulation. Finally, two subjects were eliminated from the sample. One was deleted because the subject indicated awareness of aggression research; the second because the subject refused to participate. In one class, the Decode-Nonspecific condition, five people walked out of the room during the frustration manipulation. While this caused trauma for the experimenter, such behavior does provide indirect support for the success of a socially realistic manipulation. This left an N of 111. 'Experimental Conditions Classes used in the experiment were randomly assigned to one of six experimental conditions. In all six conditions, subjects were frustrated and the frustrator left the room after collecting the pro- ficiency tests. The second experimenter then made the experimental manipulation. After manipulations, subjects completed the hostility inventory and the rating sheet. 29 Three conditions involved an encoding task. This was to determine the relative efficacy of encod- ing an aggressive message that was relevant to the frustration experience versus one that was not relevant versus encoding a nonaggressive message. In the first condition (Encode-Specific), subjects were told by the second experimenter that he had heard several student complaints concerning the manner in which the first experimenter conducted research. To see of "something could be done about it," he had subjects write an "evaluation" of the frustrator (See Appendix D for a complete set of instructions). In the second condition (Encode-Nonspecific), subjects were told that the experimenter had heard several complaints concerning Vice-President Agnew and the way he handled campus dissenters. Students were asked to write an evaluation of Agnew's dealings with college students. In the third encoding condition (Encode- Nonaggressive), subjects were asked to write a short statement on the single most important reason the recent student strike had failed on the MSU campus. The fourth group (Time-No Activity) was the principal control group. The second experimenter excused himself and also left the room. He returned 30 in five minutes and administered the hostility inven- tory and the evaluation sheet. The final two groups were both decoding groups. In one (Decode-Nonspecific), the second experimenter told the class that he had prepared an evaluation of the way Vice-President Agnew handles student dissent- ers (See Appendix E). He asked students to read it and then administered the dependent measures. In the sixth condition (Decode-Specific), subjects were told that complaints had been received concerning the first experimenter and that an evalu- ation of him had been prepared (See Appendix E). Students were asked to read it. Its content paral- leled that of the other decode condition. The dependent measures were then administered. Dependent Measures Experimenter Evaluation The principal measure used in this study was a series of questions asking subjects to rate the experiment and the experimenter on a series of scales (See Appendix A). The measure is similar to that used by Feshbach (1955). Subjects responded on a five-step scale to these questions: How much did you like participating in the study just conducted? 31 How worthwhile was it to participate in the study just conducted? What is your reaction now to the person who conducted this research? Did he perform in a satisfactory manner? In-your opinion, how qualified was the per- son who conducted the study in which you participated? A fifth question, with three foils, asked, If you were asked by the researcher to volun- teer for another study he was conducting, would you volunteer? These responses were coded and summed to yield an index with a range of 5-23. The higher the score, the more dissatisfaction with the experimenter and the experiment. This index was the principal measure of hostility in this study. The Assault Index and, the Resentment Index previously described were also administered. The experimental treatments in this study in order of data collection appear on the following page. 32 .mc0fluapnoo Hmucoawnomxo ecu ca HHHuz cm puma mane .mummHOAMHmm o» ommomou muoonnnm o>Hm .coumomon .Nuflsoxnom mo mmocoumsm topmoflccfl poonnom oco .oDMQAOHuHmm ou nonsmou poonnSm doom 0 x Amfifiav Houucoo x Away GOAUMOHGSEEOU m x oamaoommlocoooa x Ammv mmmcwmom m x oamaoommcozprooma x Acmav mmocwmsm v x m>wmmoummmcozlopoocm x Anmmv unocwmnm m x oamaoomeOZImpoocm x Away mmocwmom N x oumuomdmumooocm x Ammac mmmeumsm H x x AmHV cowumoacsaeoo unmououm x II ANHV coaumoflcnafiou Humououm whommmz ounmmoz muwaflumom coHumasmaacz muHHflumom cowumuumdum z mmmau moonw ll CHAPTER I I I RESULTS The results are divided into several sections. The first four sections all deal with the Evaluation Index, the principal dependent variable. First, the results of the instigation to aggression or decoding hypotheses are reported. Second, the results of the catharsis or aggres- sive encoding hypotheses are described. Third, the results of the intervening task or non-aggressive encoding hypo- theses are reported. Fourth, results with the Control groups are compared to the Frustration group used in the pretest. Finally, the secondary results with the general hostility measures are presented. Instigated Aggressiongypotheses Three hypotheses predicted that subjects would be more hostile after decoding an aggressive message. Results are in Table 3. 33 34 Table 3. Instigated Aggression Hypothses--Evaluation Index Source N i’ S.D. t} p Decode-Nonspecific 14 17.64 (4.16) 3.76 <.001 Time 16 12.38 (3.52) Decode-Specific 23 10.78 (4.47) Time 16 12.38 (3.52) -l.l9 n.s. Decode-Specific 23 10.78 (4.47) Decode-Nonspecific 14 17.64 (4.16) -4.64 <.001 *The two-tailed test is used since both tails of the distribution are of interest to the investigation (Edwards, 1966, p. 96). Table 3 shows that subjects who read a negative evaluation of Vice-President Agnew (Decode-Nonspecific, See Appendix F) gave a significantly more hostile evaluation of the frustrating experimenter than subjects who did nothing (Time). This supports Hypothesis lb. However, subjects who read a negative evaluation of the frustrator (Decode- Specific) did not give a significantly more hostile evalu- ation of the same frustrator when compared to the time condition. When the Decode-Specific condition is compared to the Decode-Nonspecific condition, it is seen that those who read an aggressive message concerning the frustrator gave a significantly less hostile evaluation than those who read a negative message concerning Agnew. This is 35 significant in a direction opposite to that predicted. Catharsis Hypotheses Hypotheses 2a-c are conditions under which a cathartic effect was predicted. Results are shown in Table 4. Table 4. Catharsis Hypotheses-—Evaluation Index >cH muwawumomlumomocuommm mwmuocumu .m OHQMB 41 Table 9 reports the results for the intervening task hypotheses. No significant differences were found on the Resentment Index. On the Assault Index, those who. encoded a nonaggressive message were significantly more hostile than those who did nothing. Again this is in a direction opposite to that predicted. However, those who encoded an aggressive message, either specifically relevant to the source of frustration or not, were significantly lg§§_hostile than those who encoded a nonaggressive message. This supports hypotheses 2d and 2e. On the evaluation measure, no significant differences were found for these latter two comparisons. The results for the nine hypotheses are summarized in Table 10. This is done only for the evaluation index, the principal dependent variable. 42 mo.v mo.mu Amv.mv Hm.NH mm 0>Hmmonmmmc021oooocm 1mm.mc mm.oa ma oumuomdmumeooam mo.v mm.~u Am¢.mc Hm.~H mm m>ummoummaaoznmooocm 1m~.mc AH.oH ma oumwommmaozumeooam nos.mc mm.m Ga mane Hoo.v ms.m lms.mv Hm.~H mm o>ummmummaaozumeooam xoUcH uHsmmmd .m.a sm.au Aao.mv ss.m mm m>ummmummac02umeoocm 1mm.Hc om.m ma unmnommmumeooam .m.c om.H| Ado.mv nh.m mm o>flmmoummmcozIoooonm lma.mv om.m ma oumuoommaozumeooam .m.c No.H xsm.~c Hm.m ea c. mafia Ado.mv hh.m mm obflmmoummchZIoooocm NOUGH “Gmfiflmmmm m .m . .o.m .M z (mounom muoucm>nH muwafiumomllmomonuommm xmma mnwao>uounH .m manna 43 Table 10. Summary of Results: Evaluation Rating Hypotheses by Comparison Groups Probability l. Decode-Nonspecific vs. <.001 Time 2. Decode-Specific vs. n.s. Time 3. Decode-Specific vs. <.001* Decode-Nonspecific 4. Encode-Nonspecific vs. <.01* Time 5. Encode-Specific vs. <.001* Time 6. Encode-Specific vs n.s. Encode-Nonspecific 7. Encode-Nonaggressive vs. <.001* Time 8. Encode-Nonspecific vs. n.s. Encode-Nonaggressive 9. Encode-Specific vs. n.s. Encode-Nonaggressive *Supported in direction opposite to that predicted. truism} 3.; 41“ CHAPTER IV DISCUSSION The present research investigated the role of decoding and encoding aggressive messages on subsequent feelings of hostility. Previous research suggested that reading aggressive messages would increase hostility, while writing such messages would reduce the aggressive drive instigated by an act of frustration. None of the experimental evidence supported the latter proposition. In general, both encoding and decoding behaviors led to greater subsequent feelings of hostility when compared to no activity. Only the Decode-Specific condition was not significantly different from the Time condition. This may have been due to awareness of the manipulation in that experimental condition. These findings support the theoretic orientation of Berkowitz and refute the catharsis rationale of Feshbach. Encoding an aggressive message did not reduce aggressive drive, at least in terms of attitude toward the frustrator. In fact, the data support an opposite conclusion. However, no support was found for the Berkowitz formulation that aggressive messages would increase 44 45 aggression to the extent they contain cues related to the source of frustration. DecOding a source-related aggressive message led to less hostility than decoding a nonspecific aggressive message. Methodological Issu§§_ Methodological considerations can be discussed in terms of both sampling and measurement procedures. The possibility of experimenter and group bias is included in the sampling issues; operationalization of the independent. variables as well as measurement of the dependent variables are included in measurement issues. Each will be discussed in turn. Sampling Since intact groups were used in the present study, groups were randomly assigned to independent manipulations. Despite this, however, the possibility of both experimenter and group bias remains. It's possible that the frustration manipulation was not constant across all groups. It is also possible that differences between independent manipu- lations are related to differences between class groups. Both communication and business students were used in the study. Since the frustration manipulation was socially significant for the experimenter as well as the subject 46 in that students were insulted, it is possible that an order effect existed such that the manipulation was different over time. Another possibility remains. Both the Frustration- Only and Control conditions used communication classes. The instructors of these classes indicated that they had not discussed aggression literature and subjects indicated no awareness of such literature in the debriefing session. However, the experimenter was presented as a member of the Communication Department. Subjects in these conditions may have been more sympathetic, and thereby less hostile. Further research investigating these issues is warranted. Measurement As previously indicated, subjects in the Decode- Specific condition may have been less hostile than expected due to poor Operational procedures. In the debriefing, subjects indicated that they became aware of the frustra- tion manipulation when the second experimenter handed out the prepared message negatively evaluating the frustrator. They felt that such preparation was unlikely, that the frustration procedure was therefore planned, and as a result did not show greater hostility on the evaluation message. A better procedure might be to orally present such a message, thereby exhibiting less overt planning. 47 A second problem in operationalization concerns the Encode-Specific and Encode-Nonspecific conditions. It was predicted that therewould be less hostility when the message is specifically related to the source of frustration. One would still expect differences between these conditions even if encoding leads to greater hostility. The difference in hostility should be magnified when the message is specific. No such difference was found. One possibility is that since both conditions lead to greater hostility than a Control condition, the range of response is restricted to the more hostile part of the scale. Both means are close to 17 on-a scale ranging to 23, and this may represent a psychological "ceiling" effect. Another possibility previously indicated is that both groups became similarly more hostile, but for dissimilar reasons. The Encode-Nonspecific condition was operation- alized as encoding a negative message concerning Vice- President Agnew. While this is not relevant to the present source of frustration, previous frustrations may have been recalled which were displaced toward the frustrator. Since the NonaggressiveFEncoding also may have brought to mind a previous frustration, differences among the three encoding conditions may have been collapsed. Any encoding activity may be frustrating when its purpose is not obvious and the individual is already in a frustrating situation. Further research is needed to determine if 48 (1) frustrations can be aggressively displaced and if (2) intervening non-relevant activities increase frustration. The final methodological consideration involves the dependent measures. Both the evaluation measure and the assault index have been shown to be sensitive to experiments on aggression (the latter for the first time). Further psychological research is needed, however, investigating the relation between various types of frustration and various dimensions of hostility. What sorts of frustra— tion, for example, are likely to make people more resent- ful? If subjects are frustrated because of the difficulty of some task, such as a proficiency test, are they likely to displace their hostility toward the examiner when asked to evaluate him? Evidence from this study indicates that only certain dimensions of hostility are affected by particular types of frustration. Theoretical ImplicationS‘ Perhaps the most significant aspect of this research is that encoding aggressive messages can lead to an increase in subsequent aggression. Several theoretical issues remain unanswered. Message cues may be important to the extent that they increase a state of arousal. Berkowitz argued that cues are important to the extent that they reduce inhibi- tions against aggression or relate the aggressive message 49 to the source of frustration. The present research suggests the possibility that a message may arouse previous frustrations which are then directed toward the present source of frustration. It may be that any message, or any activity, which raises arousal increases the predisposition to aggress. Messages containing sexual aggression or aggressive humor may be cases in point. If this is the case, then any activity that is arousing may lead to aggression. Frustration and previous aggression may simply be subcategories of a more general variable called arousal. Aggressive cues may be effective for that reason. Activities such as play and competition may also be aggression inducing. Further research is needed to determine the relationship between frustration, aggressive cues, and states of general arousal. Second, catharsis is that special instance where aggressive graze is reduced. Inhibition of aggressive drive does not constitute catharsis. Nor does exhibiting less aggression than another group constitute catharsis. Both groups might be considered aggressive when compared to a pre-frustration level of aggression. Interpretation of much of the previous research on catharsis is made difficult in that evidence for less aggression is often taken as evidence for catharsis. Because of this, it is more appropriate to ask what conditions are likely to lead to greater or lesser aggression rather than to ask what 50 conditions are likely to lead to catharsis. The appropriate communication question is to ask what effect various types of messages and message activities have on subsequent aggression. In the latter framework, what sort of messages and message activities are likely to result in less.hostility and aggression? It may be that hostility (aggressive feelings) cannot be reduced unless aggressive drive is reduced or inhibited. An act of aggression may be required to reduce aggressive drive. Berkowitz and Holmes.(l960), for example, found that evaluations of the frustrator (hostility) were less unfavorable when they previously were allowed to aggress (shock) the frustrator. Aggression may reduce hostility, but does hostility increase or decrease aggression? The Feshbach research found less hostility when subjects were asked to engage in aggressive encoding activity. However, subjects were told that their evalu- ations would be shown to the department chairman. This may have inhibited hostility. A similar phenomenon may have occurred here with respect to the two communication classes. While they were not told that the chairman~ would see the evaluations, they may have not wished to cause difficulty for someone in their own department. This leads to the consideration that perceived "possibility of punishment" may be an important variable. The aggressive 51 drive may be inhibited or reduced depending on the extent to which the frustrator is likely to be punished. Such an interpretation is supported by the Berkowitz and Holmes (1960) study. Third, since no clearcut differences were found between various types of encoding and decoding behavior, discussion on this point must be speculative pending further research. It does seem clear, however, that the key communication question concerns the cues in the message and the situation itself which create an atmosphere of aggression. Perhaps the Encode-Nonspecific and nonaggres- sive messages contained cues related to previous frustra- tions. Agnew may have been frustrating to college students. In the Nonaggressive Encoding condition, subjects were asked to encode their opinions as to why a student strike failed on the campus. Since they could vent their hostil— ity concerning the strike in such a message, this may have been a poor operationalization of Nonaggressive Encoding. While the results support such a conclusion, two things should be noted. If this is another condition of aggressive (non- specific) encoding, the encoded messages across these conditions should be similar. However, a perfunctory analysis indicates major differences. The aggressive messages display a predominance of name-calling. The messages concerning the strike, however, tend to cite 52 reasOns for the failure rather than make aggreafiye charges. Among the phrases "student apathy," "too many issues," "poor organization," and "most students want an education," none seem to clearly connote aggressiveness. A fourth area concerns the extent to which the frustrator is punished. Subjects may be willing to shock a frustrator but be less willing to have him dismissed from a department. The effect of telling subjects that evaluations will be reviewed by a department chairman needs to be investigated. The extent to which punishment is direct or indirect also needs to be investigated. . Hitting someone may be different from writing a negative evaluation concerning him. i Finally, the question remains as to what message cues are related to the reduction of further aggression. This and previous research indicates that aggression can be increased if an atmosphere of aggression is created through message content. The aggressive content may involve someone fighting, reading or writing negative evaluations of someone, including the frustrator, and even writing messages concerning other (possibly frustra- ting) events. The question remains as to whether an "atmosphere of nonaggression" can also be created. Would messages concerning the negative aspects of aggression, for example, lead to an inhibition of aggres- sion? If subjects were told that the victor in a fight 53 scene (who represents the subject rather than the frustra- tor) later was punished for his aggression, would this. lead to less aggression? What if subjects were asked to read or write positive rather than negative evaluations? To avoid dissonance, such subjects may become less aggres- sive. What of nonaggressive (and non-frustrating) messages? Can humorous nonaggressive message activities, for example, serve to reduce the aggressive drive? It is questions such as these which remain to be investigated. BIBLIOGRAPHY BIBLIOGRAPHY Albert, R. S. "Role of Mass Media and the effect of aggressive film content upon children's aggressive responses and identification choices." Genet. Psychol. Monographs, l957,'§§; 221-285. Bandura, A., Ross, D., and Ross, S. "Transmission of aggression through imitation of aggressive models." J. Abnorm. Soc. Psychol., 1961, 63: 575-582. ’ , Ross, D. and Ross, S. "Imitation of film- mediated aggressive models." JigAbnorm. Soc. Psychol., 1963a, 66: 3-11. , Ross, D. and Ross, S. "Vicarious reinforcement and reinforcement learning." J. Abnorm. Soc. Psychol., 1963b, 61; 601-607. ‘ Berkowitz, L. "Some factors affecting the reduction of overt hostility.“ 'J. Abnorm. Soc.Psychol., 1960, . Aggression: A Social Psychological Analysis. New York: McGraw-Hill, 1962. r . "Aggressive cues in aggressive behavior and hostility catharsis." Psychol. Rev., 1964, 11: 104-122. . "Some aspects of observed aggression." J. Pers. Soc. Psychol., 1965, 2; 359-369. , Corwin, R., and Heironimus, M. "Film violence and subsequent aggressive tendencies."x Public Opinion Quarterly! 1963, 21: 217-229. , and Geen, R. G. "Stimulus qualities of the target of aggression: A further study." JLiPers. Soc. Psychol., 1967, 5; 364-368. , and Holmes, D.S. "A further investigation of ostility generalization to disliked objects." 54 55 Berkowitz, L., and Rawlings, Edna. "Effects of film violence on inhibitions against subsequent aggression." J. Abnorm. Soc. Psychol., 1963, 66; 405-412. Breuer, J. and Freud, S. "On the physical mechanism of hysterical phenomena: preliminary communication." In S. Freud, Vol. II, Standard edition, 1893. Buss, Arnold H. The Psychology of Aggression. New York: John Wiley, 1961. . "Physical aggression in relation to different frustrations." J. Abnorm. Soc. Psychol., 1963, 67: l-7. , and Durkee, Ann. "An inventory for assessing different kinds of hostility." J. Consult. Psychol., 1957, 27(4): 343-349. Carmichael, Carl W. "Attitude change as a function of the relevance of communications and their sources to frustrating experiences." Unpublished doctoral dissertation, University of Iowa, 1965. Dollard, J.; Dobb, L.; Miller, N.; Mower, 0.; and Sears, R. Frustration and Aggression. New Haven: Yale, 1939. Edwards, Allen L. Experimental Design in Psycholggical Research. New York: Holt, Rinehart and Winston, 1966. Feshbach, S. "The drive—reducing function of fantasy be- havior." J. Abnorm. Soc. Psycholq 1955, 523 3-11. . "The stimulating versus cathartic effects of a vicarious aggressive activity." J. Abnorm. Soc. Psychol., 1961, 63: 381-385. . "The function of aggression and the regulation of aggressive drive." Psychol. Rev., 1964, 11: 257-272. , Stiles, W. B. and Bitter, E. "The reinforcing effect of witnessing aggression." J. E_p. Res. Person., 1967(2): 133-139. Geen, Russell, G. and Berkowitz, L. "Name mediated aggres- sive cues properties." J. Pers., 1966, 34(3): 456-465. 56 Haer, J. L. "Anger and retaliation to aggression in psychotherapy groups." J; Soc. Psychol., 1968, ‘76(1): 123-127. Hartman, D. P. "Influence of symbolically modeled instru- mental aggression and pair cues on aggressive behavior." J. Pers. Soc. Psychol., 1969, 11(3): 280-288. McClelland, D. C. and Apicella, F. S. "A functional classification of verbal reactions to experi- mentally induced failure." J. Abnorm. Bog; Psychol., 1945, 40: 376-390. Miller, N. E. "The frustration-aggression hypothesis." Psychol. Rev., 1941, 48; 337-342. Pastore, N. "The role of arbitrariness in the frustration- aggression hypothesis." 'J. Abnorm. Soc. Psychol., 1952, 41: 738-741. Pepitone, A. and Reichling, G. "Group cohesiveness and the expression of hostility." 'Hum. Relat., 1955, 3; 327-337. Rosenbaum, M. E. and de Charms, R. "Direct and vicarious reduction of hostility." J. Abnorm. Soc. Psychol., 1960, 62; 105-111. Sears, Robert R. "Relation of early socialization experi- ences to aggression in middle childhood." ‘J. Abnorm. Soc. Psychol., 1961. Thibaut, J. W. and Coules, J. "The role of communication in the reduction of interpersonal hostility." "J. Abnorm. Soc. Psychol., 1952, 41; 770-777. Worchel, P. "Catharsis and the relief of hostility." J;_ Abnorm. Soc. Psychol., 1957, 55; 238-243. ‘ -:--—u.. n. A. . -nqun' ~ ‘ . ‘ ..‘ t _ u" ’ V ' 2-32;...” APPENDIX A Experimenter Evaluation .' any»: ‘ i - ;-1..&-r Experimenter Evaluation APPENDIX A. onnm uoz , oz no» «Hooucnao> no» panoz .mcfipooocoo mm3 on monum Hocuocm How HoopcsHo> ou Honoumommn on» an ooxmm ouo3 50» NH Ham pa uoz mum> uoz munm uoz Sauamuam snm> mooummfiofluumm no» coflcz CH monum ecu wouosocoo 0:3 common ogu mm3 ooflmwamso soc .GOAcho H50» :H Has he noz anm> uoz musm uoz nanzmsom mum> «mosque muouommmwumm m cw EHOMHmm or can mconmomou mac» ooponocoo 0:3 cowhmm map on so: coauomon Moo» ma umn3 oz _ _ . mow «concomou on» unonm.ooxwamflo no» mcflcuhcm muocu mH mum> umnsmsom ousm uoz mnm> uoz Haa,u¢ uoz moouonocoo umsfl mosum on» Ca mummwowuumm on pa mm3 oafln3£pnoz 30m Has 94 uoz sons mnm> uoz musmnoz mauuag a non: mnm> moouooosoo umsn hogan may as mcflummaofiunmm mxHH no» paw cone 30m .fl .omgonwsoo umsm soumommn on» mnflcumocoo msowummso 3mm o no» gas 0» mxfla UH503 m3 302 57 APPENDIX B Hostility Inventory (fl Hostility Inventory APPENDIX B. ooumd ooumomfio hamcouum ouflso nonsosom ounm uoz nonsoeom ouwno mamconum .Hnmucomou mcflaoom ado: u.coo H .oE ou cocoons: m.nocs so xoon xooa H coca ooumd ooumomwo mamconum onflno pocsosom ouom uoz uo£3oEom oufino mamcouum .ucouuomefl moanuofiom 0o on uco3 on sOmHom o How ma omaa :H moan» came one oonmomfio oonmm hamcouum opflno uo£3oEom ounm uoz nonzofiom ouwso mamconum .mcfiooon oeflu oHoE ocomm ou uco3 H ooumd ooumomflo mamsouum ouflso poc3oEom ounm uoz voc3oeom ouwso mamnouum .ua How oocmacom on Maouflcflmoo canocm on .oasn unouHomEH so coxoun no: sownom o sons oonmmmfin ooumd mamaonum ouano umc3ofiom ounm poz pon3oeom ouflno mamnouum .Hooooa o mo mowuflaflnwmcommou mnflxou one Goa mcflosoafioo oxaa H oonomao ooumd mamGOHpm ooflno umn3ofiom ouom uoz voc3osom opflno mamsouum .omwa mo #90 Hooo sou o uom H Hosp doom H mofifiu um 58 I fl~AWV IH‘E lewélw. vanN3\Golq~ .N ~ I! I"! \fl.-~ .,u'> ‘4...- ‘IU 5 ...V J. -.l‘\(l 59 oond ooumomHo mHmcouum ouHoO Honzofiom onnm uoz non3ofiom ouaoo mHmcouum .c3o as HHoo on mmcflcu o>mn coo oeoa o :30 on oxHH oHno3 H ooumomHo ooumd aHmconum ouHoo nocSoEom ounm uoz non3ofiom ouHoo hHmnonum .ocooEOm mnHuuHc Hon cowoou ooom o no stcu coumo coo H oonmd ooumomwn mHmcouum ouHoo possoaom oHom uoz poc3ofiom oHHso mHmnouum .ho3 mxoonm Ho #ouoom oEOm CH no>o .co>o pom ou ucmHu uoomuom o mos oco Homaoom onu .oco Hoodoo» m on soofi mH hon HooHo no HH oonmmmHo ooumd mHmcouum oano uozsoaom ousm uoz DonsoEom oano mHmconum .onoom cuHB Conn oGOHo oEHu osomm Hocuou oHno3 H oonmomHn ooumd mHmcouum oano uon3osom ounm uoz nonzofiom oano mHmcouum .UHH03 onu nH mGOHo pom 0» pass so» HHulmcHuanm Ho ucouxo on» on co>ollmuanH use» How no woman on o>oc now ooumomHn ooumd hHmcouum ouano pocsoaom ounm uoz pon3osom ouano aHmconum .comoon ocoaon HHommfi o>HHo on uso3 no: op H .NH .HH .OH .5 APPENDIX c Insult Technique U; ._.L' .unut l . nv‘ : IAIOQ , Ifiv‘flw IDQoVOI a DAV-AI ‘ . bio-Vt. c . a . ‘Mn :5 gnu .5. "aw "vnI - - "‘QAQ I ‘- -"'\ .“‘~ a...." .fi' "A id. ' .‘ :0 ‘ ¢°55 APPENDIX C. Insult Technique "My name is Roger Haney. The research you are about to participate in is being conducted by the Commun- ication Department to determine the relationship between performance on proficiency tests and certain types of personality characteristics. The tasks are fairly simple and straightforward so they shouldn't be too difficult for you. "Students around here are always worried about grades--in fact it seems that's all they're worried about-- so we're handing out index cards with letter-number combin- ations. When you get your forms put the code number on them instead of your name or student number. That way you won't have to worry about grades--I trust that's not too difficult?" (Experimenter hands out cards.) "All right, first I've got the proficiency tests for you to take." (Begins to hand out.) "We've found eight minutes to be plenty of time so try to concentrate. I realize students don't give a damn about research but do the best you can. Answer sheets and pencils are provided so don't mark up the test.booklets--and return the pencils when you're finished; I don't know why people like to keep those. Begin as soon as you get the test." 60 . p .3. . 1.1" Q 4" U 0 1' an n. E. .. IIs- n a. 4 flood. .. .,,e l 61 During test, Experimenter walks up and down aisle. At end of four minutes, "You.have four minutes left. At this rate you're not going to finish so you'd better speed it up." At the end of eightnminutes, "All right, time's up. Hand in your test booklets and answer sheets im- mediately. I can't have any stalling." The experimenter collects the tests and then tells the assistant that he's taking those back to the office. On the way out, without looking up, "Thanks a lot." APPENDIX D Encoding Manipulations APPENDIX D. Encoding Manipulations Encoding Specific "I've heard a lot of-complaints lately about the way Mr. Haney conducts research, and before we do the inventory I'd like you to take a few minutes to write an evaluation of him--whether or not you think he's the sort of person to conduct research in front of a classroom. He won't see it so if you'd take.out a piece of paper, I think something can be done. Write an evaluation of Mr. Haney. Encoding Nonspecific "I've heard a lot of complaints lately about the way Vice-President Agnew handles campus dissenters and before we do the inventory I'd like you to take a few minutes to write an evaluation of him. We'd like you to make it as negative as possible yet believable. Write an evaluation of Agnew's dealing with college students. Encoding_Nonaggressive "I've heard a lot of statements lately concerning why the student strike on the campus here failed and before we do the inventory, I'd like you to write a statement of the single most important reason the strike failed." 62 APPENDIX E Decoding Manipulations .. - a I' . . t' ‘ Jul-"‘5’ APPENDIX E. Decoding Manipulations Decode-Specific After working on.the-PPI project for several weeks, it has become evident that_the-proper motivation of students cannot be gained by the sort of person conducting this research. In fact, he can only be considered detrimental to the project. He shows no respect for students and often- times criticizes in many ways. In doing so, his own com- petence becomes questionable.. He is arrogant, inconsider- ate and rude. His classroom-behavior is completely inappropriate. For this reason, he should not be allowed to conduct research in front of a class. Decode-Nonspecific After observing Vice-President Agnew's behavior with respect to student protestors, it has become evident that he is not the sort of person to best motivate students toward peaceful dissent. In fact, he can only be considered detrimental to the cause of peaceful dissent. He shows no respect for students and criticizes in many ways. In doing so, his own competence becomes questionable. When inter- acting with students, Agnew is arrogant, inconsiderate and rude. His behavior is completely inappropriate and for this reason he should not receive the consideration of peaceful dissenters. 63