Effects of impuisivity, depression, provocation, and time on aggressive behavior

Effects of impuisivity, depression, provocation, and time on aggressive behavior

JOURNAL OF RESEARCH IN PERSONALITY 20, 158-171 (1986) Effects of Impulsivity, Depression, Provocation, and Time on Aggressive Behavior DANIEL J. ...

1003KB Sizes 0 Downloads 65 Views

JOURNAL

OF RESEARCH

IN PERSONALITY

20, 158-171 (1986)

Effects of Impulsivity, Depression, Provocation, and Time on Aggressive Behavior DANIEL

J. HYNAN AND JOSEPH E. GRUSH Northern

Illinois

University

Impulsive and nonimpulsive males, who were temporarily depressed or nondepressed, were provoked or nonprovoked prior to participation in a communication task. The task was structured so that the male subjects could administer 25 shocks to a male confederate. As predicted, impulsive males who were depressed administered more intense shocks than did nonimpulsive males who were depressed. In contrast, impulsive and nonimpulsive males who were not depressed did not differ. On early shock trials, impulsive males also gave shocks of longer duration than did nonimpulsive males. Other findings showed that attitudes toward administering painful shocks and depressed mood together accounted for 51% of the variance in the shock intensity scores. In contrast, neither attitudes nor moods mediated aggressive responding on the shock duration measure. Theoretical and practical implications of the findings were discussed. It was concluded that emotionally triggered impulsivity can have a significant impact on human aggression. 0 1986 Academic

Press. Inc.

Considerable research has focused on the context in which human aggression occurs. The general assumption guiding this approach is that situational factors are the primary determinants of aggression (Baron, 1977). Thus, investigators have attempted to assess the influence of a variety of environmental factors and experimental conditions on subsequent aggressive responding. For example, research shows that environmental factors such as noise (Donnerstein & Wilson, 1976), heat (Baron & Bell, 1976), and crowding (Freedman, 1975) can significantly increase aggressive behavior. Somewhat similarly, experimentally induced conditions such as frustration (Worchel, 1974) and provocation (Baron, 1977) have been shown to increase aggressive responding. In contrast to the situational approach, other researchers have adopted an individual difference approach to the study of aggression. The general assumption underlying this approach is that aggressive behavior is dispositional in nature because it tends to generalize across situations (Eron, Requests for reprints should be sent to Joseph E. Grush, Department of Psychology, Northern Illinois University, DeKalb, IL 60115. 158 0092-6566186 $3 .OO Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved.

IMPULSIVITY

AND

AGGRESSION

159

1980) and to remain stable over time (Olweus, 1979). Consistent with this approach, research has shown that a variety of personality traits such as anxiety, guilt, Type A behavior pattern, and locus of control can have a determinative intluence on the occurrence of human aggression (for a review, see Dengerink, 1976). To some extent, the situational and the trait approaches to aggression research are integrated in an interactional viewpoint. Based on a review of the empirical literature, for example. Baron (1977) concluded that dispositional variables are more likely to moderate the effects of situational factors than to exert condition-free main effects upon aggression. Perhaps Berkowitz’s (1983) theory of impulsive aggression provides the clearest theoretical statement of the interactional position regarding aggression. Specifically, Berkowitz contends that events of an aversive nature automatically elicit many of the expressive motor reactions and schematic memories that are associated with experiences of anger. These associations in turn spontaneously evoke an instigation to attack, especially in those individuals who are already disposed to act in this manner. In other words, the joint occurrence of aversive stimuli and predisposed individuals produces aggression. An additional area of aggression research has been concerned with the subjective processes that might mediate aggressive behavior. For example, Baron and Bell (1976) have proposed that, up to a certain point, negative affect from various emotional states may increase aggressive responding. In contrast, other evidence suggests that attitudes may mediate aggressive responding. For example, Geen (1983) has reviewed evidence which shows that the effects of television violence on aggression are mediated, in part, by attitudes toward the television programs. In one such study (Huesmann, Eron, Klein, Brice, & Fischer, 1983), changes in attitudes about televised violence were weakly but significantly correlated with changes in aggressive behavior. The primary purpose of the present research was to investigate Berkowitz’s (1983) theoretical position that aggression can be caused by aversive events that impinge upon persons predisposed to aggress. A secondary purpose of the present study was to examine the possible mediating role of social attitudes and depressed mood on aggressive behavior. To accomplish these goals, dispositionally high- and low-impulsive males, who were temporarily depressed or nondepressed, were also situationally provoked or nonprovoked prior to their participation in a communication task. The task was structured such that subjects could administer 25 shocks of various intensities and durations to a confederate. Subjects’ depressed moods and their attitudes toward administering shocks of various intensities and durations were also assessed. Impulsivity was studied as a predisposing trait because it has been conceptually linked to dangerous actions in general (Kipnis, 1971; Shapiro,

160

HYNAN

AND

GRUSH

1965) and empirically related to aggressive behaviors in particular (Messer & Brodzinsky, 1979). Provocation in the form of interpersonal insult was used as an external stimulus because personal attack is one of the most reliable elicitors of aggression (Baron, 1977). Depression in the form of induced mood (Velten, 1968) was investigated as an internal state because it is a common emotional problem (American Psychiatric Association, 1980) that can be conceptualized on a continuum form normal mood to severe disorder (Blatt, Quinlan, Chevron, McDonald, & Zuroff, 1982). Provocation and depression were also studied because they appear to be aversive in Berkowitz’s (1983) sense of having motoric or somatic, expressive or emotional, and behavioral components associated with aggression. In this regard, the case of provocation is prototypical and straightforward. According to Berkowitz (1974, 1983), personal insult spontaneously leads to expressive motor reactions which are often associated with anger and can also lead to aggressive behavior. While the case for depression is more tenuous, there is evidence to suggest that it can also be aversive. For example, Frost, Graf, and Becker (1979) showed that depression effects from the Velten procedure are due more to the suggestion of bodily sensation statements than to cognitively devaluative statements. If induced somatic sensations are construed as conditioned stimuli that automatically elicit the negative mood state (Frost et al., 1979, pp. 961-962), then the process by which depressed mood is formed closely parallels the process by which provoked anger is aroused. The empirical link between induced depression and aggressive behavior is provided by Hynan’s (1979) research which shows that temporarily depressed subjects displayed more aggression than nondepressed subjects. Hypothesis 1 was that impulsive subjects who were provoked and/or depressed would display more aggression than nonimpulsive subjects who were provoked and/or depressed. In contrast, impulsive and nonimpulsive subjects who were unprovoked and/or nondepressed would not differ in their aggressive behavior. This hypothesis was derived from Berkowitz’s (1983) theory which predicts that external or internal events of an aversive nature stimulate interpersonal attack, especially by persons who are predisposed to act in an aggressive manner. In addition, this hypothesis was based on empirical research which shows that the effects of impulsivity typically interact with the effects of some other factor. For example, high-impulsive subjects have displayed more aggression than low-impulsive subjects when they are provoked (Barr-as & Grush, 1981), exposed to aggressive models (Kipnis. 1971), or engage in fantasy aggression (Messer & Brodzinsky, 1979). Hypothesis 2 was that attitudes toward administering intense shocks would mediate shock intensity scores, while depressed mood would mediate shock duration scores. This hypothesis was also derived from theory and research. According to a theory of reasoned action. for example,

IMPULSIVITY

AND

AGGRESSION

161

intentions to engage in a specific behavior are partially determined by attitudes toward that behavior (Ajzen & Fishbein, 1980). Since shock intensity represents an intentional type of aggression (Berkowitz & Alioto, 1973), it was expected that attitudes would be primarily related to this mode of responding. In contrast to the central role of evaluative beliefs in attitudes, emotional factors are more prominent in depression (Izard, 1977). Since shock duration represents an emotional type of aggression (Berkowitz, 1974, p. 166), it was expected that depressed mood would be primarily related to this mode of responding. METHOD Design and Participants Eighty undergraduate males voluntarily participated in the experiment which had a 2 (high or low impulsive) x 2 (depressed or nondepressed) x 2 (provoked or nonprovoked) between-subjects design. Participants were preselected on the basis of their scores on an impulsivity scale (Grush, Barras, & Hynan. 1986) that had been administered earlier in the semester.’ Specifically, individuals whose scores were in the upper and lower quartiles of a sample distribution of 363 college males were considered high- and low-impulsive subjects, respectively. Subjects in the various high- and low-impulsive conditions were first matched on their impulsivity scores and then randomly assigned to the other experimental conditions. Eight psychology majors played the role of the confederate. Each confederate served equally often in each treatment condition.

Procedure Overview. Taped instructions informed subject and confederate that there would be three main stages to the experiment. In the first, subject and confederate would get acquainted by exchanging autobiographical letters and written impressions of one another. In the second stage, participants would receive a treatment that was designed to influence their moods. In the final stage, one person would act as a “stimulator” and the other as a “responder” in a nonverbal communication task. Informed consents were then obtained and a faked lottery conducted. The lottery was “rigged” so that the subject was always the stimulator and the confederate was always the responder. Appararus and instructions. After the lottery, the confederate was taken to another room ostensibly to write his autobiographical letter and to receive taped instructions. The experimenter then oriented the subject to the modified Buss Aggression Machine (BAM) that is described in detail by Buss (1961). Specifically, the subject was told that his task was to communicate a visual concept to his partner by presenting a scheduled pattern of lights. The confederate’s ostensible task was to decide on the correct response by pushing one of two buttons. The subject was also told that when his partner made an accurate communication, positive feedback in the form of a correct light was to be flashed to his partner. When the partner made an incorrect communication, the subject was to choose which of 10 intensities of shock to administer to him. To increase credibility concerning shock, the subject received three sample shocks of I The IS-item lmpulsivity Scale of Grush et al. (1986) includes some items from existing scales (e.g., Kipnis, 1971), other items based on a discussion of the impulsive personality (Shapiro, 19651, and still other items that measure delay of gratification. The self-report measure employs mixed response formats, has a coefficient a of .68, and has item-total correlations that are positive in sign and moderate in magnitude (M = .40).

162

HYNAN

AND

GRUSH

low to moderate intensity (Levels 1. 3, and 5). The subject was then asked to rate Shock Level 5 on a 6-point scale of painfulness. If necessary, the main rheostat of the BAM was adjusted until the subject rated Shock Level 5 as “moderately unpleasant.” Provocation manipulation. After the BAM instructions were completed, the subject was given instructions and a form to write his autobiographical letter. The letters of the subject and the confederate were exchanged while the two were still separated. The experimenter then requested that both subject and confederate express their honest reactions in a written response to their cosubject’s letter. The confederate’s standardized reply was mildly insulting for half the subjects and mildly positive for the other half. The confederate’s (handwritten) replies were modified from Kingsbury (1976). The mildly insulting reply read as follows: I might as well be honest, I’m not too impressed with you. I really don’t think you were honest in what you wrote. The only thing I’ve learned about your personality is that you’re probably a phoney, and that turns me off. I don’t mind if people have problems, everybody does. But if you can’t admit that you aren’t perfect, then you’re nothing but a put on. I guess this experiment is a flop since I’m not really interested in getting to know you. If you would forget about looking so great and were just yourself, you may someday find the acceptance you seem to be looking for. The mildly positive reply was worded as follows: This is kind of difficult to write since I don’t really know that much about you. I’d say you gave a pretty respectable description of yourself. Just about everything you wrote is something that I can identify with. I would say that we have some things in common. I guess there’s a lot of people around that are just like you or me, but I just have some preconceived ideas about them. I never get to find out they’re just like me. Without knowing you or anybody for a time, in my opinion, it’s very difficult to pass judgment about you-but from what I read you seem to be a nice guy. Mood induction. After the subject read the confederate’s reply, he was told that his mood treatment would begin. The experimenter ostensibly gave the confederate directions for his mood treatment in the adjacent room by means of taped material. When the experimenter returned, the subject received one of two types of mood induction. After reading preparatory instructions, the subject read aloud a set of 60 statements that were designed to be depressing or a set of neutral statements that served as a control condition (Velten, 19681.’ Since Velten’s procedure might produce only a transitory type of affect. the mood induction procedure came after the provocation manipulation and before the aggression procedure. Aggression procedure. The experimenter briefly reviewed the subject’s role as “stimulator,” responded to questions, and made clarifications as needed. The experimenter also indicated that the experiment would terminate whenever the responder made five correct responses in a row or when 60 trials were completed. The subject then attached finger electrodes to the confederate and read brief instructions concerning the use of the apparatus. The * With three exceptions, the instructions and mood statements were the same as those used by Velten (1968): (a) the instructions included information about the tasks that would follow: (b) four statements involving anger were revised because a relatively pure state of depression was desired; and (c) one statement about Viet Nam was changed to rehect current concerns about the economy.

IMPULSIVITY

AND

AGGRESSION

163

communication task began and the confederate responded on each trial according to a prearranged schedule. Specifically, the confederate always made the same 25 errors before responding correctly on Trials 46-50. During this phase of the experiment, the confederate’s actual task was to record the intensity of each shock and the cumulative duration of every five shocks.’ The only shocks actually given during the experiment were those that the subject received as samples.

Assessments Mood. Immediately after the subject completed the mood procedure, he completed a form of Izard’s (1977, p. 126) Differential Emotions Scale (DES). Specifically, the subject rated the extent to which each of 30 adjectives accurately described his current affective state. These ratings were made on 7-point scales that ranged from very slightly (1) through moderately (4) to very srrongly (7). Depressed mood was measured by the “distress” subscale which is made up of the average ratings of items concerned with being downhearted. sad, and discouraged. The distress subscale was used because Izard (1977) has shown that it is sensitive to both depressed states and long-term clinical depression. A second assessment of the subject’s depressed mood was taken immediately after the aggression procedure. This was done to determine whether the depressed mood had endured throughout the aggression procedure. Attitudes. Once the second DES assessment was completed, the subject filled out an attitude scale with a semantic differential format. Specifically, the following five items were used to assess the subject’s attitudes toward delivering Shock Intensity Levels 4 and 7 to the confederate: good-bad; useful-useless; helpful-harmful; worthwhile-worthless; and fair-unfair. Identical items were used to assess the subject’s attitudes toward administering shock durations of 0.5 and I.0 seconds to the confederate. Two magnitudes of shock intensity were examined because there was no a priori basis for knowing whether attitudes toward higher or lower intensities would mediate aggressive responding on the intensity measure. Similarly, two levels of shock duration were included to determine whether attitudes toward longer or shorter durations might mediate aggressive responding on the duration measure. Some filler concepts (e.g., “instructions”) were also rated to take the focus off of shock intensity and shock duration. Manipulation checks. Following the attitude assessments, the subject completed a brief questionnaire on which he used 7-point scales to rate his anger and dislike for the confederate as well as his level of frustration experienced during the experiment. The latter rating was taken to investigate whether effects due to mood or provocation might be attributable to differences in frustration associated with failure to teach the confederate more quickly. Four tiller items (e.g., subject’s “alertness”) were also included to help conceal the purpose of the questionnaire. Debriejing. After the manipulation questionnaire was completed, the subject was asked a number of open-ended questions about the experiment. In particular, an attempt was made to ascertain whether the subject might have had any suspicions about the true nature and purpose of the experiment.4 The subject was then thoroughly informed about the

’ The BAM used here has an indicator that shows the intensity of each shock, while its time clock only records cumulative shock duration. Thus, it would have been difficult for the confederate to record precisely the intensity and duration of each shock. Consequently, the statistical analyses only investigated a general temporal effect by looking at five blocks of five trials each. 4 Two judges, blind to subjects’ aggression responses, agreed that five subjects should be disqualified and replaced because their debriefing responses indicated that they were suspicious about the cover story and/or manipulations. Two other subjects were replaced because equipment malfunctions caused them to have incomplete data.

164

HYNAN

deceptions and the reasons for them. Each his feelings and experiences in the study. All mood before they left the laboratory.

AND

GRUSH

subject subjects

was also carefully questioned about indicated that they were in a positive

RESULTS Manipulation

Checks

Depression. To determine the effectiveness of the mood induction procedure, scores on the distress subscale of the DES were subjected to a 2 x 2 x 2 x 2 repeated measures analysis of variance. Results showed that greater distress was reported by the provoked (M = 3.23) than by the nonprovoked subjects (M = 2.57): F(l, 72) = 7.29, p < .Ol . Results also showed that significant main effects for depressed condition and assessment interval were qualified by a significant interaction between these two factors: F(1, 72) = 18.96, p < .OOl. On first assessment, depressed subjects reported more distress (M = 4.36) than did nondepressed subjects (M = 2.55): F(1, 78) = 25.48, p < .OOl. On second assessment, however, there was no difference between depressed (M = 2.41) and nondepressed (M = 2.28) subjects: F(1, 78) = 0.25, p > .60. In other words, the mood manipulation was initially successful, but it subsequently dissipated by the time the BAM procedure had been completed. Provocation. To test the effectiveness of the provocation manipulation, separate 2 x 2 x 2 analyses of variance were conducted on the scores for the dislike and anger ratings. As expected, subjects in the provoked condition expressed more dislike for the confederate (M = 4.95) than did subjects in the nonprovoked condition (M = 3.64): F(1, 72) = 25.50, p < .OOl. Provoked subjects were also more angry at the confederate (M = 5.07) than were nonprovoked subjects (M = 1.72): F( 1, 72) = 101.64, p < .OOl. No other main effects or interactions were significant. In other words, the confederate’s mild insult was successful in provoking the subjects who were the recipients of it. Frustration. To determine whether the effects due to the experimental conditions might be attributable to differences in frustration, the ratings on this scale were subjected to a 2 x 2 x 2 analysis of variance. None of the main effects or interactions were significant, p’s > .25. Thus, it appears that frustration was not responsible for any of the observed differences reported below. Aggressive

Behavior

Shock intensity. To examine the effects of the independent variables upon direct and intentional aggression, shock intensity scores were analyzed in a 2 x 2 x 2 x 5 mixed analysis of variance. The first three factors represent the experimental conditions, while the fourth factor represents five blocks of five trials each. As can be inferred from the means in

IMPULSIVITY

AND

165

AGGRESSION

Table 1, provoked subjects gave more intense shocks (M = 5.30) than did nonprovoked subjects (M = 4.39): F(1, 72) = 6.57, p < .013. The intensity of shocks also increased over trial blocks: F(4, 288) = 45.18, p < .OOl. These significant main effects are consistent with past findings for both provocation and trials (Baron, 1977; Buss, 1966). The analysis of variance also produced a significant impulsivity x depression interaction: F(I, 72) = 4.96, p <. 029. To specify the nature of this interaction, separate analyses of variance were conducted for the depressed and nondepressed conditions. In the depressed condition, impulsive subjects gave significantly more intense shocks (M = 5.46) than did nonimpulsive subjects (M = 4.23): F(1, 38) = 5.06, p < ,025. In the nondepressed condition, impulsive (M = 4.69) and nonimpulsive (M = 5.01) subjects did not significantly differ from one another: F(1, 38) = 0.49, p > .45. All other follow-up tests of the interaction were nonsignificant. The meaning and implications of this finding are discussed later. Shock duration. To determine the effects of the independent variables upon indirect and emotional aggression, shock duration scores were also analyzed in a 2 x 2 x 2 x 5 mixed analysis of variance. The only significant finding from this analysis was an impulsivity x blocks interaction: F(4, 288) = 3.69, p < .006. To specify the nature of this interaction, separate one-way analyses of variance were conducted for the impulsive versus nonimpulsive subjects on each of the five trial blocks. As can be seen in Table 2, the main effects for impulsivity were significant for Trial Blocks 1 and 2: F(1, 78) = 5.39, p < .03 and F(1, 78) = 4.46, p < .04, respectively. There were no significant differences between impulsive and nonimpulsive subjects on Trial Blocks 3, 4, and 5. In other words, impulsive subjects gave relatively longer shocks on earlier trial blocks and maintained that response pattern across all blocks. In contrast, nonimpulsive subjects became somewhat disinhibited over time and gave equally long shocks only on later trial blocks. TABLE MEAN

SHOCK

1

INTENSITY

SCORES

Subjects Condition Depressed Provoked Nonprovoked Nondepressed Provoked Nonprovoked Note. Intensity aggression.

Impulsives

Nonimpulsives

5.94 4.98

4.78 3.67

5.29

5.18

4.10

4.85

scores could range from 1 to 10, with higher scores indicating

more

166

HYNAN

AND

TABLE MEAN

SHOCK DURATION

GRUSH

2

SCORES FOR THE IMPLJLSIVITY

x

BLOCKS

INTERACTION

Trial blocks Subjects Impulsives Nonimpulsives

1

2

3

4

5

0.49, 0.40,

0.50, 0.41*

0.50, 0.46,

0.46, 0.45,

0.48, 0.48,

Note. Duration scores are in seconds, with higher numbers indicating more aggression. Within each block, means with different subscripts differ significantly at p < .05.

Mediating Processes To examine the roles of attitudes and depressed feelings as possible mediators of aggression, separate stepwise regression analyses were carried out which employed either mean shock intensity or mean shock duration scores (across all 25 trials) as the criterion variables. As described below, a variety of attitude and mood measures were employed as predictor variables. Shock intensity. Attitudes toward delivering Shock Level 4, Shock Level 7, and the difference between these scores (4 - 7) were used as predictor variables. For each attitude measure, a composite score was constructed by summing the ratings of all five evaluative scales (see Ajzen & Fishbein, 1980). Additional predictor variables included ratings of distress on the first DES assessment, the second DES assessment, and the difference between these scores (first - second). Attitudes toward delivering Shock Level 7 entered first in the regression equation (/3 = .62) and accounted for 43% of the variance in the shock intensity scores: F = 56.85, p < .OOl. Distress ratings on the second DES assessment entered second in the regression equation (/3 = .28) and accounted for an additional 8% of the variance: F = 11.28, p < .OOl . No other predictor variables had significant regression coefficients. These findings are consistent with the prediction that attitudes would primarily determine aggressive responding on the shock intensity measure. Shock duration. Attitudes toward administering a 0.5-s shock, a l.Os shock, and the difference between these scores (0.5 - 1.0) were used as predictor variables. Once again, five semantic differential ratings were used to construct the composite measures of attitudes. The distress ratings described in the previous analysis were also employed as predictor variables. Consistent with the experimental hypothesis, none of the attitude measures were significantly related to the mean shock duration scores. Contrary to the hypothesis, all of the mood measures also failed to predict the mean shock duration scores.

IMPULSIVITY

AND

AGGRESSION

167

DISCUSSION The primary purpose of the present research was to investigate whether depressed mood and/or provocation would elicit aggressive responding in impulsive subjects who are presumably disposed to act in this manner. As predicted, results showed that impulsive subjects who were depressed administered more intense shocks than did nonimpulsive subjects who were depressed. In contrast, impulsive and nonimpulsive subjects who were not depressed did not differ in the intensity of shocks that they delivered. In addition, results showed that impulsive subjects gave shocks of longer duration on earlier trial blocks than did nonimpulsive subjects. The present results show rather clearly that impulsivity can be an important personality trait in aggressive behavior. The finding that impulsivity interacted with depression to affect shock intensity scores is consistent with the results of other studies which show that the effects of impulsivity are typically “triggered” by some other mechanism (Barras & Grush, 1981; Kipnis, 1971; Messer & Brodzinsky, 1979). That depression served as the specific activating mechanism is also consistent with other results which show that depression can increase both aggression (Hynan, 1979) and hostility (Strickland, Hale, & Anderson, 1975). The fact that impulsivity interacted with trial blocks to affect shock duration scores supports the theoretical notion that impulsive subjects are initially predisposed to be more aggressive than nonimpulsive subjects (Barras & Grush, 1981). A secondary purpose of the present study was to examine the possible mediating role of social attitudes and depressed mood on aggressive behavior. As predicted, attitudes toward administering relatively painful shocks (Level 7) accounted for 43% of the variance in aggressive responding on the shock intensity measure. Depressed mood accounted for an additional 8% of the variance. Contrary to prediction, depressed mood alone and in combination with attitudes toward administering long shocks failed to account for a significant portion of the variance in shock duration scores. The finding that attitudes significantly predicted shock intensity scores is consistent with prevailing theories on the nature of social attitudes (Ajzen & Fishbein, 1980) and shock intensity measures of aggression (Berkowitz & Alioto, 1973). This finding is also similar to the results obtained in other recent investigations of the relationship between attitudes and aggression. For example, attitudes toward rape and other acts of sexual violence have been shown to predict males’ self-reports of the likelihood of engaging in rape (Briere & Malamuth, 1983), their reported use of other sexually coercive acts (Rapaport & Burkhart, 1984), and their actual aggressive behavior toward female confederates (Malamuth, 1983).

168

HYNAN

AND

GRUSH

The present findings should be viewed with some caution, however, because attitudes were assessed after, rather than prior to, the aggressive responding. Thus, it is possible that subjects simply inferred their attitudes from their behaviors (Bern, 1968). Some caution is also warranted because neither depressed mood nor social attitudes predicted shock duration scores. Despite these limitations, the present results are nevertheless important because they show rather clearly that both social attitudes and depressed mood can play some mediating role in deliberate acts of aggression. Implications

The overall pattern of findings obtained here fit most comfortably into Berkowitz’s (1983) theory of impulsive aggression. According to this theory, internal (depression) or external (provocation) events of an aversive nature automatically elicit many of the expressive-motor reactions and schematic memories that are associated with the experience of anger. These associations, in turn, spontaneously evoke an instigation to attack, especially in those individuals (impulsive) who are already disposed to act in this manner. Attack can be directed at the source of the aversiveness or at a stimulus (confederate) that is paired with the aversive event. Attack can occur rather instantaneously (duration scores), without any awareness or cognitive activity (uncorrelated attitudes) on the part of the aggressor. If cognitions enter the picture (intensity scores), they are parallel to (correlated with)-but not responsible for (causal ofl-the observed aggression. It should be noted that not all of the evidence obtained here supports Berkowitz’s (1983) theory. Specifically, there is some question as to why provocation failed to interact with impulsivity in determining aggressive responding. Perhaps provocation is such a robust elicitor of counterattack that aggression is a likely response regardless of the provoked person’s general predisposition to attack. In support of this notion, Baron (1977) points out that it is not at all uncommon in the aggression literature for strong situational manipulations to swamp the effects of individual difference variables. More recent support for a “suppression effect” of provocation comes from a study by Holmes and Will (1985). In that study, Type A’s and Type B’s were equally aggressive when provocation was present, but Type A’s were more aggressive than Type B’s when provocation was absent. To some extent, the present findings are also consistent with Bandura’s (1983) social learning theory of aggression. According to this theory, impulsive subjects would tend to engage in cognitions that exonerate their aggression. For example, impulsive subjects can justify their aggressive behavior by resorting to euphemistic labeling, e.g., “Shock Level 7 is

IMPULSIVITY

AND

AGGRESSION

169

and ‘helpful’ to provide clear feedback to my co‘good, ’ ‘worthwhile,’ subject.” In contrast, nonimpulsive subjects would tend to engage in cognitions that control their aggression, e.g., “Shock Level 4 or 5 is good enough to provide clear feedback.” Consistent with this interpretation, the results of the stepwise regression analysis showed that subjects’ evaluative beliefs about painful shocks were highly correlated with the actual shock intensities that they administered. Contrary to this view, the results of the stepwise regression analysis showed that subjects’ evaluative beliefs about long shocks were uncorrelated with the actual shock durations that they administered. Although the balance of evidence obtained here tends to support Berkowitz, it would be premature to conclude that his theory is correct and that Bandura’s theory is wrong. Further research will have to determine more clearly the extent to which impulsive individuals act with little prior thinking (Berkowitz, 1983) and/or engage in self-exonerating cognitions (Bandura, 1983). What is clear here, however, is that impulsive individuals did display more aggression than nonimpulsive individuals. The practical significance of this finding lies in the fact that it provides further supporting evidence for the impulsivity scale of Grush et al. (1986). This evidence is important because research on impulsive aggression has been hindered by ethical constraints on the use of highly aversive conditions to create impulsive states (Zillman, 1979) and logistical limits on the use of performance measures to identify impulsive individuals (Kagan, Rosman, Day, Albert, & Phillips, 1964; Porteus, 1965; for a review, see Messer, 1976). In other words, the availability of a relatively reliable and valid measure of impulsivity, that also offers economy of group administration and ease of computer scoring, can greatly facilitate future research on impulsive aggression. CONCLUSION

Impulsive violence is a controversial and highly important form of human aggression. Aside from its obvious theoretical interest, impulsive violence is also a phenomenon of considerable concern to society at large. For example, there is evidence to indicate that impulsivity is a major component of homicides (Berkowitz, 1974) and a contributing factor in child and spouse abuse (Zillman, 1979). In this regard, it is particularly interesting that depression triggered the impulsive aggression observed here. Given the prevalence of depressive episodes in the general population (American Psychiatric Association, 1980), one cannot help but wonder whether depression also serves to activate some nonlaboratory manifestations of impulsive violence. Hopefully, the present research will provide some impetus and one avenue to answer this and related questions in the future.

170

HYNAN

AND

GRUSH

REFERENCES Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall. American Psychiatric Association (1980). Diagnostic and statistica/ manual of mental disorders (3rd ed.). Washington, DC: Author. Bandura, A. (1983). Psychological mechanisms of aggression. In R. G. Geen & E. I. Donnerstein (Eds.), Aggression: Theoretical and empirical reviews: Vol. I. Theoretical and methodological issues (pp. l-40). New York: Academic Press. Baron, R. A. (1977). Human aggression. New York: Plenum. Baron, R. A., & Bell. P. A. (1976). Aggression and heat: The influence of ambient temperature, negative affect, and a cooling drink on physical aggression. Journal of Personality and

Social

Psychology,

33, 245-255.

Barras, J. A.. & Grush, J. E. (1981,

May). The injuence of dispositional and situational Paper presented at the meeting of the Midwestern Psychological Association, Detroit. Ml. Bern, D. J. (1968). Attitudes as self-descriptions: Another look at the attitude-behavior link. In A. G. Greenwald, T. C. Brock, &T. M. Ostrom (Eds.), Psychologicalfoundations of attitudes (pp. 197-215). New York: Academic Press. Berkowitz, L. (1974). Some determinants of impulsive aggression: Role of mediated associations with reinforcement for aggression. Psychological Review, 81, 165-176. Berkowitz. L. (1983). The experience of anger as a parallel process in the display of impulsive, “angry” aggression. In R. G. Geen & E. I. Donnerstein (Eds.), Aggression: Theoretical and empirical reviews: Vol. 1, Theoretical and methodological issues (pp. 103-133). New York: Academic Press. Berkowitz, L.. & Alioto, J. T. (1973). The meaning of an observed event as a determinant of its aggressive consequences. Journal of Personality and Social Psychology. 28, 206-217. Blatt, S. J., Quinlan, D. M., Chevron, E. S., McDonald, C.. & Zuroff. D. (1982). Dependency and self-criticism: Psychological dimensions of depression. Journal of Consulting and Clinical Psychology, 50, 113-124. Briere, J.. & Malamuth, N. M. (1983). Self-reported likelihood of sexually aggressive behavior: Attitudinal versus sexual explanations. Journa/ of Research in Persona&y, factors

on

aggressive

behavior.

17, 315-323.

BUSS, A. H. (1961). The psychology of aggression. New York: Wiley. Buss, A. H. (1966). The effect of harm on subsequent aggression. Journal Research

in Personality,

of Experimental

1, 245-255.

Dengerink, H. A. (1976). Personality variables as mediators of attack instigated aggression. In R. G. Geen & E. C. O’Neal (Eds.), Perspectives on aggression (pp. 61-98). New York: Academic Press. Donnerstein, E., & Wilson, D. W. (1976). Effects of noise and perceived control on ongoing and subsequent aggressive behavior. Journal of Personality and Social Psychology, 34, 774-778. Eron. L. D. (1980). Prescription for the reduction of aggression. American Psychologist, 35, 244-252. Freedman, J. L. (1975). Crowding and behavior. San Francisco: Freeman. Frost, R. 0.. Graf. M., & Becker, J. (1979). Self-devaluation and mood. Journal of Consulting and

Clinical

Psychology,

47, 958-962.

Geen, R. G. (1983). Aggression and television violence. In R. G. Geen & E. I. Donnerstein (Eds.), Aggression: Theoretical and empirical reviews: Vol. 2. Issues in research (pp. 103-125). New York: Academic Press. Grush, J. E.. Barras, J. A., & Hynan, D. J. (1986). A new’self-report measure ofimpulsivity. DeKalb, IL: Northern Illinois University. (ERIC Document Reproduction Service No. ED 263 168).

IMPULSIVITY

AND AGGRESSION

171

Holmes, D. S.. & Will, M. J. (1985). Expression of interpersonal aggression by angered and nonangered persons with the Type A and Type B behavior pattern. Journal of Personality

and

Social

Psychology,

48, 723-727.

Huesmann, R. W., Eron, L. D., Klein, R., Brice, P., & Fischer, P. (1983). Mitigating the imitation of aggressive behaviors by changing children’s attitudes about media violence. Journal

of Personality

and

Social

Psychology,

44, 899-910.

Hynan, D. J. (1979). The effects of mood on hostile aggression. Unpublished master’s thesis. Northern Illinois University. DeKalb. Izard, C. (1977). Human emotions. New York: Plenum. Kagan, I., Rosman, B. L., Day, D., Albert, J., &Phillips, W. (1964). Information processing in the child: Significance of analytic and reflective attitudes. Psychological Monographs, 78(1, Whole No. 578). Kingsbury, S. J. (1976). Attribution of intent of third party aggression and the reduction of hostile aggression. Unpublished doctoral dissertation, University of Miami, Coral Gables, FL. Kipnis, A. (1971). Character structure and impulsiveness. New York: Academic Press. Malamuth, N. M. (1983). Factors associated with rape as predictors of laboratory aggression against women. Journal of Personality and Social Psychology, 45, 432-442. Messer, S. B. (1976). Reflection-impulsivity: A review. Psychological Bulletin, 83, 10261052. Messer, S. B., & Brodzinsky. D. M. (1979). The relation of conceptual tempo to aggression and its control. Child Development, 50, 758-766. Olweus, D. (1979). Stability of aggressive reaction patterns in males: A review. Psychological Bulletin,

86, 852-875.

Porteus. S. (1965). Porteus Maze Test: Fifty years’ application. Palo Alto, CA: Pacific Books. Rapaport, K., & Burkhart. B. R. (1984). Personality and attitudinal characteristics of sexually coercive college males. Journal of Abnormal Psychology, 93. 216-221. Shapiro. D. (1965). Neurotic styles. New York: Basic Books. Strickland, B. R.. Hale, D. W., & Anderson, L. K. (1975). Effects of induced mood states on activity and self-reported affect. Journal of Consulting and Clinical Psychology, 43, 587. Velten, E. (1968). A laboratory task for the induction of mood states. Behavior Research and

Therapy,

6, 473-482.

Worchel. S. (1974). The effect of three types of arbitrary thwarting on the instigation to aggression. Journal of Personality, 42, 300-318. Zillman, D. (1979). Hostility and aggression. Hillsdale, NJ: Erlbaum.