7
0145-6008/94/1803-0657$3.00/0 ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH Vol. 18, No. 3 May/June 1994 Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal Amos Zeichner, Joseph D. Alien, Peter R. Giancola, and Jeffrey M. Lating Theorists have attempted to account for the relationship between alcohol intoxication and a wide range of aggressive behaviors by ascribing alcohol-related aggression to the disinhibiting effects of alcohol and to its disruptive effects on cognitive processes. Alloca- tion of attention and situational threat have been thought to mediate alcohol-related aggression. In the present study, 72 male social drinkers received either alcohol, a placebo, or a nonalcoholic bev- erage and were exposed to either threatening or nonthreatening personal information. Measuring levels and duration of aggressive responses on a Taylor-Buss aggression machine, intoxicated sub- jects were found to be generally more aggressive under threatening than under nonthreatening information conditions. Intoxicated sub- jects also reported relatively large increases in anger, depression, and tension following the aggression task, which appeared to be affected by the type of information received. These findings concur with and refine previous models suggesting that alcohol focuses the drinker’s attention to salient cues in threatening circumstances,thus increasingthe likelihood of aggressive behavior. Key Words: Alcohol, Aggression, Threat, Affective Arousal. HE RELATIONSHIP between alcohol consumption T and aggressive behavior has been studied extensively. ‘7’ Alcohol use has been implicated in a variety of aggressive behaviors, including domestic ~ i o l e n c e , ~ h~rnicide,~ sexual assault^,^ suicide: and traffic accident^.^ Controlled ex- perimental laboratory studies have also provided strong evidence supporting the hypothesis that alcohol tends to facilitate aggressive behavior.’-’’ Moreover, a large meta- analytic study recently concluded that “alcohol does in- deed cause aggressive behavior.”’ The evidence supporting the relationship between al- cohol consumption and aggression is indeed compelling. However, there appears to be a lack of consensus regarding the underlying mechanisms involved in the expression of alcohol-related aggression. Models put forth to account for this association have implicated alcohol’s disinhibiting and arousing properties, as well as alcohol-induced expec- tancies as determinants of alcohol-related aggression.”.’2 Of particular interest are recent models suggesting that alcohol-related aggression can be explained by the disrup- tive effects of alcohol on higher cognitive functions. From the Department of Psychology, University of Georgia, Athens, Received for publication May 7, 1993; accepted September 30, 1993 This research was supported by Grant AA06027 from the National Reprint requests: Amos Zeichner, Ph. D.. Department of Psychology, Copyright 0 I994 by The Research Society on Alcoholism. Georgia. Institute on Alcohol Abuse and Alcoholism. University of Georgia, Athens, GA 30602-30 13. Alcohol Clin Exp Res, Vol 18, No 3, 1994: pp 651-663 Zeichner and Pihl’ ascribe alcohol-related aggression to alcohol’s disruptive effects on information processing re- garding the consequences of one’s behavior. Similarly, Hull13 states that alcohol reduces self-awareness by dis- rupting the cognitive encoding of external and internal cues relative to the social appropriateness of a given be- havior. Others have suggested that alcohol interferes with the brain’s ability to simultaneously process a multitude of environmental cues resulting in selective attention to the most salient of situational cues.I4 These authors posit that alcohol facilitates aggressive behavior by allowing the individual to focus selectively on dominant provocative cues to the exclusion of less dominant inhibitory cues. A similar but more elaborate model posited by Steele and colleagues focuses on the allocation of attention to account for alcohol-related aggre~sion.”~’~ These authors claim that alcohol facilitates aggressive behavior by impairing higher cognitive processes of perception and thought. Ac- cordingly, alcohol promotes aggressive behavior by in- creasing one’s attention to the more salient aspects of a situation that encourages aggressive behavior, while deem- phasizing the less salient cues’ capacity to inhibit such behavior. A recent study demonstrated that both intoxicated and sober men allocated equal amounts of attention to non- threatening information. However, intoxicated men allo- cated more attention to threatening information than did the sober men.I7 These findings support Steele’s model in that intoxicated men allocate more attention to salient (threatening)information compared with nonsalient (non- threatening) information. A threatening or provocative situation has been shown to be an important determinant in eliciting aggressive behavior. Most studies tend to operationalize threat as the potential for receiving physical pain, usually in the form of noxious electrical shocks, from a competitor in an adversarial relationship. Studies have found that threat differentially influences aggressive behavior under sober, intoxicated, and placebo conditions. Threat of physical harm has been shown to increase aggressive behavior in the nonintoxicated state and even more so in the intoxi- cated state.”.” Given these findings, the purpose of this study was to test the moderating effect of personal threat on alcohol- related aggression and affective arousal using a task based on the Buss-Taylor aggression machine.*’ Given the well- 657

Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

Embed Size (px)

Citation preview

Page 1: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

0145-6008/94/1803-0657$3.00/0 ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH

Vol. 18, No. 3 May/June 1994

Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

Amos Zeichner, Joseph D. Alien, Peter R. Giancola, and Jeffrey M. Lating

Theorists have attempted to account for the relationship between alcohol intoxication and a wide range of aggressive behaviors by ascribing alcohol-related aggression to the disinhibiting effects of alcohol and to its disruptive effects on cognitive processes. Alloca- tion of attention and situational threat have been thought to mediate alcohol-related aggression. In the present study, 72 male social drinkers received either alcohol, a placebo, or a nonalcoholic bev- erage and were exposed to either threatening or nonthreatening personal information. Measuring levels and duration of aggressive responses on a Taylor-Buss aggression machine, intoxicated sub- jects were found to be generally more aggressive under threatening than under nonthreatening information conditions. Intoxicated sub- jects also reported relatively large increases in anger, depression, and tension following the aggression task, which appeared to be affected by the type of information received. These findings concur with and refine previous models suggesting that alcohol focuses the drinker’s attention to salient cues in threatening circumstances, thus increasing the likelihood of aggressive behavior.

Key Words: Alcohol, Aggression, Threat, Affective Arousal.

HE RELATIONSHIP between alcohol consumption T and aggressive behavior has been studied extensively. ‘7’

Alcohol use has been implicated in a variety of aggressive behaviors, including domestic ~iolence,~ h~rnicide,~ sexual assault^,^ suicide: and traffic accident^.^ Controlled ex- perimental laboratory studies have also provided strong evidence supporting the hypothesis that alcohol tends to facilitate aggressive behavior.’-’’ Moreover, a large meta- analytic study recently concluded that “alcohol does in- deed cause aggressive behavior.”’ ’

The evidence supporting the relationship between al- cohol consumption and aggression is indeed compelling. However, there appears to be a lack of consensus regarding the underlying mechanisms involved in the expression of alcohol-related aggression. Models put forth to account for this association have implicated alcohol’s disinhibiting and arousing properties, as well as alcohol-induced expec- tancies as determinants of alcohol-related aggression.”.’2 Of particular interest are recent models suggesting that alcohol-related aggression can be explained by the disrup- tive effects of alcohol on higher cognitive functions.

From the Department of Psychology, University of Georgia, Athens,

Received for publication May 7, 1993; accepted September 30, 1993 This research was supported by Grant AA06027 from the National

Reprint requests: Amos Zeichner, Ph. D.. Department of Psychology,

Copyright 0 I994 by The Research Society on Alcoholism.

Georgia.

Institute on Alcohol Abuse and Alcoholism.

University of Georgia, Athens, GA 30602-30 13.

Alcohol Clin Exp Res, Vol 18, No 3, 1994: pp 651-663

Zeichner and Pihl’ ascribe alcohol-related aggression to alcohol’s disruptive effects on information processing re- garding the consequences of one’s behavior. Similarly, Hull13 states that alcohol reduces self-awareness by dis- rupting the cognitive encoding of external and internal cues relative to the social appropriateness of a given be- havior. Others have suggested that alcohol interferes with the brain’s ability to simultaneously process a multitude of environmental cues resulting in selective attention to the most salient of situational cues.I4 These authors posit that alcohol facilitates aggressive behavior by allowing the individual to focus selectively on dominant provocative cues to the exclusion of less dominant inhibitory cues. A similar but more elaborate model posited by Steele and colleagues focuses on the allocation of attention to account for alcohol-related aggre~sion.”~’~ These authors claim that alcohol facilitates aggressive behavior by impairing higher cognitive processes of perception and thought. Ac- cordingly, alcohol promotes aggressive behavior by in- creasing one’s attention to the more salient aspects of a situation that encourages aggressive behavior, while deem- phasizing the less salient cues’ capacity to inhibit such behavior.

A recent study demonstrated that both intoxicated and sober men allocated equal amounts of attention to non- threatening information. However, intoxicated men allo- cated more attention to threatening information than did the sober men.I7 These findings support Steele’s model in that intoxicated men allocate more attention to salient (threatening) information compared with nonsalient (non- threatening) information.

A threatening or provocative situation has been shown to be an important determinant in eliciting aggressive behavior. Most studies tend to operationalize threat as the potential for receiving physical pain, usually in the form of noxious electrical shocks, from a competitor in an adversarial relationship. Studies have found that threat differentially influences aggressive behavior under sober, intoxicated, and placebo conditions. Threat of physical harm has been shown to increase aggressive behavior in the nonintoxicated state and even more so in the intoxi- cated state.”.”

Given these findings, the purpose of this study was to test the moderating effect of personal threat on alcohol- related aggression and affective arousal using a task based on the Buss-Taylor aggression machine.*’ Given the well-

657

Page 2: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

ZEICHNER ET AL. 658

documented link between alcohol and aggression and based on our findings that intoxicated men allocate more attention to negative (threatening) than to positive (non- threatening) cues, it was hypothesized that intoxicated men would act in a more aggressive manner when placed in a situation where they received personally threatening (salient negative) versus personally nonthreatening (non- salient positive) information. A self-report measure of affective arousal was also administered to assess changes in affect produced by the ingestion of alcohol and by participation in the aggression task under contrasting con- ditions of personal threat.

METHODS

Participants and Design Seventy-two male social drinkers between the ages of 2 1 and 34 (MWe

= 23.6) participated in the study and were paid $10.00 for their partici- pation. They were recruited as volunteers from the general population of a small southeastern U.S. university town by means of newspaper advertisements. Participants were randomly assigned to 1 of 6 conditions comprising a 3 x 2 Drug by Information design. Group assignment was as follows: Alcohol-Positive (n = 12), Alcohol-Negative (n = 12), Placebo- Positive (n = 12), Placebo-Negative (n = 12), Sober-Positive ( n = 12), and Sober-Negative (n = 12). Subjects with a score of 5 or above on the Michigan Alcoholism Screening Test2‘ (MMAST = 0.97) were excluded from participation. Also excluded were subjects in which participation was counterindicated by specific alcohol history and health screen ques- tionnaires.

Questionnaires All subjects completed four questionnaires during a separately sched-

uled screening session: ( I ) Drinking and medical history-A specially designed questionnaire that required subjects to provide information regarding the age at which they consumed their first drink (AFD), age when regular drinking began (ARD), drinking occasions/week (DOW), drinks/occasion (DPO), favorite beverage, parental history of alcoholism, and presence of diseases such as cardiovascular, renal, arthritis, and visual problems; (2) Repression-sensitization Scale-A 127-item ques- tionnaire derived from the Minnesota Multiphasic Personality Inventory (MMPI) designed to quantify an individual‘s response to threatening stimuli on a continuum ranging from repression (i.e., avoidance, denial) to sensitization (i.e., approach, intellectualization);22 (3) SeFmonitoring Scale-A 18-item scale designed to quantify a person’s regulation of expressive self-presentation and level of sensitivity to social cues;23 and (4) Cook-Medley Hostility Questionnaire-A 50-item true/false ques- tionnaire derived from the MMPI designed to assess predominantly unexpressed hostility.24

The three personality inventories were administered to increase the believability of the “computerized trait analysis” and to verify that random subject assignment achieved similar distributions of pertinent traits among the groups.

Profile of Moods States (POMS). This scale was administered during the experimental session to assess changes in mood. This is a standardized 64-item self-rating scale of six mood states: tension-anxiety, depression- dejection, anger-hostility, vigor-activity, fatigue-inertia, and confusion- bewilderment.25 Due to the high degree of association between alcohol and tension reduction, depression, and aggression, we were interested in assessing the effects of alcohol intoxication and aggressive behavior on the following subscales of the POMS: Anger, Depression, and Tension.

Drug Conditions Participants in the Alcohol group were administered 0.99 g/kg of 95%

USP alcohol in a 1:4 (alcohoktonic water) mixture divided into two

beverages of equal quantity. The beverages were served in tall glasses wrapped in terry-cloth holders and placed on a paper coaster and a saucer. Placebo group participants were administered two beverages as described, with the alcohol replaced by an equal quantity of tonic water. In addition, 1.0 ml of 95% USP alcohol was layered on the surface of each beverage, and the glass holder and paper coaster were sprayed with alcohol to maximize the strength of the placebo manipulation. A potent smell of alcohol lingered in the small experimental chamber throughout the session. Alcohol and Placebo subjects were told that they would receive an alcohol dose equivalent to up to five “shots” of vodka. This placebo manipulation has been used successfully in earlier studies.’*lo Participants in the Sober group were told that they would receive bever- ages containing no alcohol. They were administered two beverages of tonic in a quantity as described for the Placebo group, but without any alcohol manipulation. The smell of alcohol in the chamber was absent for the Sober group. All subjects were required to refrain from ingesting any food for at least 4 hr prior to testing.

Task Participants in the both Information groups were led to believe that

information relative to their personality traits would be made available to them by another subject (confederate) whose task it was to form an impression of the participant’s character based on data from question- naires completed by the participant earlier. Subjects were told that the conferderate’s capacity for attitude change was being tested in the exper- iment. To enhance the believability of the task, subjects were told that this evaluation technique is “commonly and reliably” used by corpora- tions during hiring interviews of prospective employees. Contingent on groups assignment, 30 positive or negative personality traits selected from Anderson’sz6 trait list (one trait/trial) were displayed on a Com- modore- 128 microcomputer screen placed in front of the participant. Examples of positive and negative traits are “bright,” “polite” and “stubborn,” “foolish,” respectively.* The participant was instructed to ”shape“ the confederate’s impression by administering to him bogus electrical shocks. All subjects were instructed to shape the confederate’s evaluation in a subjectively perceived positive direction. The bogus shock was administered by pressing keys on a numerical key pad of the commodore- I28 computer. Each of the I through 5 keys was interfaced with a volt meter that displayed incrementing needle deflections com- mensurate with higher key numbers pressed. Both the level of shock selected and the duration of shock key was pressed served as dependent measures of aggression. Shocks were generated by a Mark I1 Behavior Modifier (Fame1 Instruments, Grand Island, NE) shock generator. The subject received sample shocks of levels 1, 3, and 5 (1 .O, 1.85, and 3.25 mA, respectively). This was conducted to strengthen the bogus shock manipulation and to demonstrate to the subject that he could select level ” I ,“ which was a hardly noticeable shock. The 30 trials were interspaced by 5-sec rest periods. Each trait was displayed until the participant selected and pressed a shock key on his keyboard.

Procedure One week prior to the experiment, each subject was first given a

screening session at which he completed the MAST, and the alcohol history and personality questionnaires. Subjects were told that some of the information from the questionnaires (excluding the medical data)

*Estimates for this task’s validity can be garnered from Pearson correlation coeffients between shock intensity and the Cook-Medley Total hostility score, Aggression subscale, and Cynical hostility subscale scores. These are presented in “Results.” This task was also used by Lifhitz-Cooney and Zeichne?6 to demonstrate direrential attention allocation patterns in Type A and B individuals. Adjectives prepared for that study were judged by Jive undergraduates to have either ‘bositive” or “negative” connotation. A modified version of this task was also used by Zeichner et a1.17

Page 3: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

ALCOHOL AND AGGRESSION 659

will be used in the experiment. During a subsequent session, the subject was led to a 2.5 m x 2.5 m electrically shielded and sound-attenuated chamber and was seated in a chair facing a Commodore- 128 computer and monitor. After preliminary instructions were delivered and informed consent was obtained, the first blood alcohol concentration (BAC) level was taken from the participant using an Alcohol Countermeasures Sys- tems breathalyzer Model 75 (Denver, CO) to ensure subject sobriety. The participant also rated and entered his response on an 11-point Likert-type How Drunk scale presented to him on the computer monitor (0 = not drunk at all, 1 1 = the most drunk I’ve ever been). The participant then completed the POMS scale by responding via the computer. Next, the subject was served his two beverages by an assistant who was blind to the subject’s drinking condition.

The beverages were served in succession, and the participants were allowed 10 min to consume each drink while reading magazines. A 20- min absorption period followed the drinking period. A second BAC reading was taken and the How Drunk and POMS scales were readmin- istered. The trait information task then began. The task was followed by a third administration of the BAC to ensure that subjects were on the ascending limb of the BAC curve, and the How Drunk and POMS scales. At this point, all participants were debriefed. During debriefing, subjects were queried regarding the potency of their beverages (Alcohol and Placebo groups only), their opinion of the accuracy of traits they viewed, and the effectiveness of the shocks they administered. Participants in the Alcohol group were required to wait in a special lounge until their BAC had declined to 0.05% or lower. Subjects were seated alone when beverage consumption and task administration occurred.

RESULTS

Alcohol and Manipulation Checks BAC Level. All participants in the Alcohol group

achieved BAC values that were well within the “social drinking” range. BAC levels varied between 0.079% and 0.1 10% (MeAc = 0.092%, SD = 0.016) following the drink- ing period, and varied between 0.089% and 0.128% (MBAC = 0.107%, SD = 0.018) at the completion of the trait information task. No identifiable BAC levels were noted in the participants of either the Placebo or Sober groups.

“How Drunk” Scale. Participants in the Alcohol group indicated an average “drunk” score of 5.87 (SD = 1.35) directly before the trait task and an average “drunk” score of 5.67 (SD = 0.92) following the task. Placebo participants produced mean “drunk” scores of 4.92 (SD = 1.03) and 4.20 (SD = 0.88), respectively. Differences between the Alcohol and Placebo groups on the pre- and posttask scores were assessed using two-tailed t tests. Although the groups did not differ on the pretask score, the Placebo subjects rated themselves as less drunk after the trait task, t(68) = 2.98, p < 0.05.

DebrieBng Information. In general, Alcohol and Placebo subjects commented that they found the drinks strong but tasting badly. None of these subjects identified his bever- age as a placebo (or devoid of alcohol). Upon being asked how they felt about the traits they viewed, several subjects in the Positive condition made comments such as “I enjoyed getting this evaluation” or “I guess he wanted to make me feel good,” whereas some subjects in the Nega- tive group made comments such as ‘‘I would not get a job with these traits” or “this was a mean evaluation.” No

subjects indicated doubts relative to the trait manipula- tion.

Drinking History Questionnaire (DHQ Separate 2 (Positive, Negative) X 3 (Alcohol, Placebo,

Control) ANOVAs were performed on the AFD, ARD, DOW, and DPO variables of the DHQ. Means f SD for these variables were 17.83 (SD = 2.41), 19.85 (SD = 1.16), 1.62 (SD = 0.68), and 1.92 (SD = 0.78), respectively. No significant differences were found among the six experi- mental groups. Although beer was the preferred beverage among all subjects, x2 analyses found no differences in frequency distributions for favorite beverage among the groups. No subject reported parental history of alcoholism.

Personality Inventories Separate 2 x 3 ANOVAs were performed on the inven-

tories measuring hostility, repression/sensitization, and self-monitoring. None of the analyses revealed significant main effects or interactions.

Measures of Physical Aggression Administered shock intensities and durations were each

submitted to separate 3 (Drug) x 2 (Information Type) ANOVAs. Significant differences were further analyzed by post-hoc planned comparisons where appropriate.

Shock Intensity Level. Analysis of the shock intensity data revealed significant main effects for Drug, F(2,66) = 4.53, p < 0.02, and Information, F(1,66) = 16.27, p < 0.00 1. Post-hoc comparisons revealed that Alcohol sub- jects chose significantly higher shocks than did either Placebo, F( 1,66) = 9.06, p < 0.01, or Sober participants, F( 1,66) = 5.74, p < 0.0 1. Across drug conditions; partici- pants receiving Negative Trait information selected signif- icantly higher shocks than did those receiving Positive Trait information. No significant differences were found between the Placebo and Sober groups on this measure (Fig. 1, top panel).

Shock Duration. Analysis of the duration that partici- pants pressed the shock keys revealed significant effects for Drug, F(2,66) = 14.9 1, p < 0.00 1, Information, F( 1,66) = 16.79, p < 0.00 1, and a Drug x Information interaction, F(2,66) = 1 1.29, p < 0.001. Post-hoc comparisons revealed that both the Alcohol and Sober groups applied shocks for longer durations than the Placebo group, F( 1,66) = 14.2 1, p < 0.001 and F(1,66) = 28.16, p < 0.001, respectively. Further analysis of the interaction revealed that Alcohol subjects receiving Negative Trait information adminis- tered longer shock durations than did the Placebo subjects receiving either trait or the Sober group receiving Positive Traits ( p < 0.001). Alcohol subjects receiving Negative Traits also administered longer durations than their Posi- tive Trait counterparts, F(1,66) = 7.83, p c 0.01. Sober- Negative subjects administered longer shock durations

Page 4: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

660

1 2 -

1 -

2 0 8 - c 0 N 0 6 - D S

0 4 -

0 2 -

0 1

ZEICHNER ET AL.

SHOCK INTENSITY

ALCOHOL

-NEGATIVE W P O S l T l V E

n

PLACEBO SOBER INFORMATION TYPE

SHOCK DURATION

-NEGATIVE POSITIVE

ALCOHOL PLACEBO SOBER INFORMATIOPJ T Y P E

Fig. 1. Mean shock values (top) and shock durations (bottom) administered by subjects assigned to the Alcohol, Placebo, and Sober conditions following receipt of either Negative or Positive Trait information.

than those of all other groups, except the Alcohol-Negative group (all ps < 0.01; Fig. 1 , bottom panel).

POMS Separate 3 x 2 (Drug x Information) ANOVAs for

repeated measures were performed on three subscales of the POMS questionnaire. The POMS had been adminis- tered three times during the study to evaluate the effects of alcohol intoxication and aggressive behavior on Anger, Depression, and Tension.

Anger. The repeated measures analysis found significant Phase x Drug, F(4,132) = 10.98, p < 0,001 and Phase x Information, F(2,132) = 9.14, p < 0.001, interaction ef- fects. Post-hoc contrasts revealed that the increase in anger during the aggression task relative to its pretask baseline was significantly greater for the Alcohol group than for the Placebo, F( 1,66) = 3 1.09, p < 0.00 1 , or Sober, F( 1,66)

= 28.79, p < 0.001, groups. Across Drug conditions, subjects became more angry during the task relative to their pretask baseline when they received Negative Trait information than when they received Positive Trait infor- mation, F(1,66) = 16.56, p < 0.001 (Fig. 2).

Depression. A significant Phase effect for Drug, F(4,132) = 2.39, p < 0.05, was found by the repeated measures ANOVA. Post-hoc contrast revealed that across Informa- tion conditions, Alcohol participants reported a signifi- cantly larger increase in depressive mood following the aggression task, F( 1,66) = 5.13, p < 0.05, than either of the other groups. The largest, albeit nonsignificant, in- crease in rated depression occurred with the Alcohol group receiving negative traits (Fig. 2).

Tension. The repeated measures ANOVA revealed sig- nificant Phase x Drug, F(4,132) = 3.33, p < 0.01, and Phase x Information, F(2,132) = 4.75, p < 0.0 1 , interac- tion effects for the POMS measure. Post-hoc contrast revealed that both the Alcohol and Sober subjects reported significantly greater pre-posttask increases in tension rel- ative to the Placebo group, F(1,66) = 8.40, p < 0.01 and F( 1,66) = 12.43, p < 0.001, respectively. Tension increases were significantly higher for groups receiving Negative Trait information than for those receiving Positive Trait information, F( 1,66) = 10.03, p < 0.01 (Fig. 2).

To partially test the validity of the present aggression paradigm Pearson product-moment correlation, coeffi- cients were computed between the shock intensity meas- ure and the Cook Medley hostility scores under positive and negative information conditions. In the positive in- formation condition aggression correlated 0.65 with Total hostility, 0.47 with Aggressive behavior, and 0.40 with Cynical hostility. In the negative condition, the respective coefficients were 0.70,0.66, and 0.50.

DISCUSSION

The results of the present study demonstrated that in- toxicated participants selected significantly higher shock intensities relative to the other groups and applied shocks at longer durations than did Placebo participants. Partic- ipants exposed to threatening information selected higher shock levels than did subjects receiving nonthreatening information. With regard to shock duration, both intoxi- cated and sober subjects behaved similarly, with the ex- ception that intoxicated subjects were differentially af- fected by the type of information. Intoxicated subjects responded more aggressively to negative evaluation, whereas sober participants appeared equally aggressive to both types of evaluation.

These results add to earlier findings regarding the allo- cation of attention to situational threat.” Not only does the inebriate allocate equal attention to situational cues of threat when the level of threat is low and displays prefer- ence to cues of threat when the threat is high, but the intoxicated social drinker is able to discern between dif-

Page 5: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

ALCOHOL AND AGGRESSION 66 1

....................... .._ -rrr... ..........................................

.............................

1 ....................................... ........................ ...,..

..-l.. .mnrt----*y

8 ....................................................... pr /1.

........................ ..... . I ............................................................... ..........-.... ................... *T- c

0 Y

......................................................... ........................................... ......................................................

F ......................................................... ........................................... ...................................................... 1 ::

t lo 8 8 4

.................... . . I b /t ..................................................................................... .-T-T

......................... . . . . . . . . . . . . . .

n .....................

Fig. 2. Mean scores on the Anger, Depression, and Tension scales of the POMS during baseline (I), pre- (2), and posttask (3) phases for the Alcohol (top), Placebo (middle), and Sober (bottom) groups following the receipt of either Negative (left) or Positive (right) Trait information.

ferent types of threat. In this study, alcohol did not mask the fact that two types of personal evaluations, thus pre- sumably amounting to two levels of threat, were leveled

displayed only under conditions of high Personal threat. Moreover, these findings refine previous theories claiming that alcohol disrupts information processing. The present results also confirm earlier findings regarding strong as-

sociations between alcohol and aggression assessed in the lab~ratory. '~*~'~~* Finally, the present findings did not sug- gest that alcohol lowers the salience of self-relevant infor- mation, as has been previously reported.13 Although the Positive Information condition also contained aggression- relevant cues (i.e., the opportunity to deliver shocks), the fact that significantly higher levels of aggression were displayed in the Negative Information condition speaks to the robust nature of the data.

at the drinker. Rather, higher levels of aggression were

Page 6: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

662 ZEICHNER ET AL.

A somewhat different picture emerges in the absence of alcohol. Placebo subjects did not respond differentially to the two types of personal information and generally dis- played lower levels of aggression in comparison to the other two groups. This finding is most likely explained by the facts that Placebo subjects did not ingest the aggres- sion-facilitating substance and believing that one has con- sumed alcohol does not in itself facilitate aggressive be- havior. This is in agreement with earlier studies indicating that alcohol expectancies do not facilitate aggressive be- h a ~ i o r . ~ ~

Although it appeared from the debriefing data that the placebo manipulation was successful, no standardized scale was administered during the debriefing to confirm subjective reports. The intoxication ratings provided by the Placebo subjects on the “How Drunk” scale should be interpreted with caution, because some demand charac- teristics may have been produced by informing the sub- jects to expect alcohol content to match up to five shots of vodka and due to the open-ended nature of the scale. Although a measure of social desirability could have pro- vided pertinent data, the fact that no self-monitoring differences were detected among the groups increases our confidence that demand characteristics may have played only a minor role, if at all, in producing the present findings.

Data from the POMS provide additional information concerning the intoxicated subject’s behavior. Intoxicated participants reported experiencing the largest increases in anger and depression following the aggression task. These increases occurred predominantly in the negative infor- mation condition. This task-provoked anger may have been responsible, in part, for the threat-specific (e.g., greater personal threat conveyed by Negative Information) aggression displayed by the Alcohol subject^.^'

The high levels of tension reported by sober participants following the aggression task in either information condi- tion may help explain why these participants, without the aggression-facilitating effects of alcohol, selected relatively long shock durations and high shock intensities in re- sponse to Negative Trait information. The tension-reduc- ing effects of alcohol often discussed in the literature3’ were not seen in the present study; that is, no significant reduction in anger, depression, or tension were reported following the ingestion of alcoholic beverages. These find- ings are not surprising, however, given the report of Jo- sephs and Steelel’ that claimed that the tension-reducing properties of alcohol are only apparent when an individual is faced with a distracting stimulus. This was not the case in the present study.

In interpreting the results of this study, it is important to note that, although the two measures of aggression used (shock intensity and duration) are thought to both meas- ure the same behavioral c o n s t r ~ c t , ~ * ~ ~ they are, in fact, two different indices. Whereas shock intensity may be concep- tualized as a volitional or “active” index of aggression,

shock duration could be considered a subtle and more “passive” measure. It has been suggested that shock inten- sity and duration are respective measures of direct and indirect aggression given the increased cognitive effort required to discriminate different durations as opposed to different intensit ie~.~~ On the other hand, Be rk~wi tz~~ has suggested that shock duration is more likely a measure of impulsive aggression, given its sensitivity to emotional arousal. Accordingly, in the case of the intoxicated partic- ipants, experience of intense anger and tension in a situ- ation involving aggressive behavior did not affect their responding on a volitional active level as much as it affected their “passive” responding modus to the different threat conditions. These results are in accord with Berkow- itz’s theorizing that shock duration is sensitive to emo- tional arousal. Clearly, because no empirical work on this aggression index distinction has been reported yet, this interpretation is merely speculative.

In summary, the findings of this and the Zeichner et all7 study indicate that alcohol may not disrupt higher cognitive functioning in the straightforward “across-the- board” manner that previous models have indicated. It may be the case that, under conditions of high personal threat, individuals who allocate attention to Negative In- formation respond with higher levels of aggression. How- ever, this may not be the case in situations of low personal threat. This interpretation is supported by research on self- focused attention, which claims that individuals who self- focus more are increasingly attentive to their affective state.34 Therefore, it would not be unreasonable to argue that individuals faced with personally threatening infor- mation would react more aggressively than those encoun- tering personally nonthreatening information. This view concurs with the Steele and J o ~ e p h s ’ ~ ~ construct of “alco- hol myopia,” where alcohol is thought to facilitate aggres- sive behavior by increasing one’s attention to the more salient aspects of a situation (e.g., threatening informa- tion), whereas deemphasizing less salient aspects (e.g., social sanctions against aggression) that function to inhibit such behavior.

ACKNOWLEDGMENTS

We thank Paul Rasmussen, Charles Petrie, and Renuka Harper for their help in the data collection.

REFERENCES

1. Murdoch D, Pihl RO, Ross D Alcohol and crimes of violence: Present issues. Int J Addict 25:1065-1081, 1990

2. Pihl RO, Ross D: Research on alcohol related aggression: A review and implications for understanding aggression. Drugs Soc 1:105- 126, 1987

3. Gorney B: Domestic violence and chemical dependency: Dual problems, dual interventions. J Psychoactive Drugs 2 1 :229-238, 1989

4. Goodwin DW: Alcohol in suicide and homicide. Q J Stud

5. Johnson SD, Gibson L, Linden R Alcohol and rape in Winni- Alcohol 34: 144-156, 1973

peg, 1966-1975. J Stud Alcohol 39~1887-1894, 1978

Page 7: Alcohol and Aggression: Effects of Personal Threat on Human Aggression and Affective Arousal

ALCOHOL AND AGGRESSION 663

6. Frances RJ, Franklin J, Flavin D K Suicide and alcoholism. Ann New York Acad Sci 487:316-326, 1986

7. Lucker GW, Kruzich D, Holt MT, Gold JD: The prevalence of antisocial behavior among U.S. army DWI offenders. J Stud Alcohol 52:318-320, 1991

8. Taylor SP, Gammon CB: Effects of type and dose of alcohol on human physical aggression. J Pers SOC Psychol 32:169-175, 1975

9. Zeichner A, Pihl RO: Effects of alcohol and behavior contingen- cies on human aggression. J Abnorm Psychol 88:153-160,1979

10. Zeichner A, Pihl RO: Effects of alcohol and instigator intent on human aggression. J Stud Alcohol 41:265-276, 1980

I 1 . Bushman BJ, Cooper HM: Effects of alcohol on human aggres- sion: An integrative research review. Psych Bull 107:34 1-354, 1990

12. Graham K Theories of intoxicated aggression. Can J Behav Sci

13. Hull J G A self-awareness model of the causes and effects of alcohol consumption. J Abnorm Psychol 90586-600, 1981

14. Taylor SP, Leonard KE: Alcohol and human physical aggression, in Geen RG, Donnerstein EI (eds): Aggression: Theoretical and Empirical Reviews. New York, Academic Press, Inc., 1983, pp 77-101

15. Josephs RA, Steele CM: The two faces of alcohol myopia: Attentional mediation of psychological stress. J Abnorm Psychol 99: 1 15- 126, 1990

16. Steele CM, Southwick L Alcohol and social behavior. 1. The psychology of drunken excess. J Pers SOC Psychol 48: 18-34, 1985

17. Zeichner A, Allen JD, Petrie CD, Rasmussen PR, Giancola PR: Attention allocation: Effects of alcohol and information salience on attention processes in male social drinkers. Alcohol Clin Exp Res 17:727- 732, 1993

18. Taylor S , Gammon C, Capasso D: Aggression as a function of the interaction of alcohol and threat. J Pers Soc Psychol 34:938-941, 1976

19. Taylor S, Schmutte G, Leonard K, Cranston J: The effects of alcohol and extreme provocation on the use of a highly noxious electric shock. Motivation and Emotion 3:73-8 1, 1979

20. Taylor S P Aggressive behavior and physiological arousal as a function of provocation and the tendency to inhibit aggression. J Pers

12: 141-1 58, 1980

351297-310, 1967

2 1 . Selzer M: The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. Am J Psychiatry 127:1653-1658, 1971

22. Bryne D: The repression-sensitization scale: Rational, reliability, and validity. J Pers 29:334-349, 1961

23. Snyder M: Public Appearanm/Private Practices: The Psychol- ogy of Self-Monitoring. Philadelphia, Freeman, 1987

24. Cook W, Medley D: Proposed hostility and pharisaic-virtue scales for the MMPI. J Appl Psychol 238:414-418, 1954

25. McNair DM, Lorr M, Droppleman L F Manual for the Profile of Mood States. San Diego, ETS, 197 1

26. Anderson NH: Likableness ratings of 555 personality-trait words. J Pers Soc Psychol 9:272-279, 1968

27. Taylor SP, Gammon CB: Aggressive behavior of intoxicated subjects: The effect of third-party intervention. J Stud Alcohol 37:917- 930, 1976

28. Zeichner A, Pihl RO, Niaura R, Zacchia C: Attentional processes in alcohol-mediated aggression. J Stud Alcohol 43:7 14-724, 1982

29. Hull JG, Bond C F Social and behavioral consequences of alcohol consumption and expectancy: A meta-analysis. Psych Bull

30. Russell JA, Mehrabian A: The mediating role of emotions in

3 1 . Cappell H, Greeley J: Alcohol and tension reduction: An update on research and theory, in Blane HT, Leonard KE (eds): Psychological Theories of Drinking and Alcoholism. New York, Guilford Press, 1987,

32. Rogers R: Race variables in aggression, in Geen R, Donnerstein E (eds): Aggression: Theoretical and Empirical Reviews, vol 2. New York, Academic Press, 1983, pp 27-50

33. Berkowitz L Some determinants of impulsive aggression: Role of mediated associations with reinforcements for aggression. Psychol Rev

34. Scheier MR, Carver CS Self-focused attention and the experi- ence of emotion, attraction, repulsion, elation and depression. J Pers Soc

35. Steele CM, Josephs RA: Alcohol myopia: Its prized and danger- ous effects. Am Psychol 45:921-933, 1990

36. Lifshitz Cooney JL, Zeichner A: Selective attention to negative feedback in type A and type B individuals. J Abnorm Psychol 9 4 1 10- 112, 1985

99:347-360, 198 1

alcohol use. J Stud Alcohol 37:1508-1536, 1975

p~ 15-54

8 1: 165-176, 1974

Psycho1 35:625-636, 1977