toward a situational analysis of gender differences in aggression

12
Sex Roles, Vol. 8, No. 8, 1982 Toward a Situational Analysis of Gender Differences in Aggression i Shelagh M. J. Towson 2 and Mark P. Zanna 2 University of Waterloo The present study explored the notion that frustration in situations congruent with traditional sex roles will elicit higher levels of aggression from both women and men than will frustration in sex-role incongruent situations. In a 2 X 2 de- sign, female and male undergraduates read a "feminine" or "masculine" vignette describing an individual prevented by a classmate from practicing for an exami- nation in dance or body-building exercises. Subjects in gender-congruent con- dit~'ons perceived the examination as more important and advocated more ag- gressive responses than did subjects in gender-incongruent conditions. Males were significantly more aggressive than females only in the masculine situation; in the feminine situation, this difference was reduced or eliminated. The implications of these results for the interpretation of apparent sex differences in aggression are discussed. A pervasive stereotype in our society is that men are naturally more aggressive than women; and real world evidence confirms a higher incidence of male than of female aggression (e.g., Eron, Huesmann, Lefkowitz, & Walder, 1972; Frodi, Macaulay, & Theme, 1977; Macaulay, Note 1). While the hypothesis that this difference is a naturally occurring result of biological sex differences (Maccoby & Jacklin, 1974) is still open to debate, it seems clear that our society's uncriti- cal acceptance and reinforcement of this hypothesis results in gender differences in aggression. 1The present research was supported by Canada Council Grant S76-0344 to the second author and was conducted while the first author was supported by an Ontario Graduate Scholarship. The authors wish to thank Ann Werner for help in preparing and pilot testing the materials. This article was presented as a paper at the Association for Women in Psy- chology Conference in Dallas, March 8-11, 1979. 2 Correspondence should be sent to Shelagh M. J. Towson or Mark P. Zanna, Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G 1. 903 0360-25/82/0800-0903 $03.00/0 © 1982 Plenum Publishing Corporation

Upload: shelagh-m-j-towson

Post on 09-Aug-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

Sex Roles, Vol. 8, No. 8, 1982

Toward a Situational Analysis of Gender

Differences in Aggression i

Shelagh M. J. Towson 2 and Mark P. Zanna 2 University of Waterloo

The present study explored the notion that frustration in situations congruent with traditional sex roles will elicit higher levels o f aggression from both women and men than will frustration in sex-role incongruent situations. In a 2 X 2 de- sign, female and male undergraduates read a "feminine" or "masculine" vignette describing an individual prevented by a classmate from practicing for an exami- nation in dance or body-building exercises. Subjects in gender-congruent con- dit~'ons perceived the examination as more important and advocated more ag- gressive responses than did subjects in gender-incongruent conditions. Males were significantly more aggressive than females only in the masculine situation; in the feminine situation, this difference was reduced or eliminated. The implications o f these results for the interpretation o f apparent sex differences in aggression are discussed.

A pervasive stereotype in our society is that men are naturally more aggressive than women; and real world evidence confirms a higher incidence of male than of female aggression (e.g., Eron, Huesmann, Lefkowitz, & Walder, 1972; Frodi, Macaulay, & Theme, 1977; Macaulay, Note 1). While the hypothesis that this difference is a naturally occurring result of biological sex differences (Maccoby & Jacklin, 1974) is still open to debate, it seems clear that our society's uncriti- cal acceptance and reinforcement of this hypothesis results in gender differences in aggression.

1The present research was supported by Canada Council Grant S76-0344 to the second author and was conducted while the first author was supported by an Ontario Graduate Scholarship. The authors wish to thank Ann Werner for help in preparing and pilot testing the materials. This article was presented as a paper at the Association for Women in Psy- chology Conference in Dallas, March 8-11, 1979.

2 Correspondence should be sent to Shelagh M. J. Towson or Mark P. Zanna, Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G 1.

903

0360-25/82/0800-0903 $03.00/0 © 1982 Plenum Publishing Corporation

904 Towson and Zanna

But are these differences consistent - are men always more aggressive than women? Despite the prevalence of the "macho man" theme, a persistent cultural counterpoint is the expectation that in certain contexts women will be at least as aggressive as men: Mothers will go to any length to protect their children. A good woman will fight tooth and nail to keep her man. And behind every hen- pecked husband stands a verbally aggressive wife. Furthermore, hell hath no fury like a woman scorned.

That this image of women as aggressors has a factual basis is borne out by both real world and laboratory observations. An extensive literature indicates that women are as aggressive as men in marital conflict situations (e.g., Gelles, 1972; Straus, Gelles, & Steinmetz, 1974); and, tragically, mothers, perhaps even to a greater extent than fathers, abuse their children (Gelles, 1973; Gil, 1968, 1970; Radbill, 1974; Smith, 1975). A review of the experimental literature also supports the contention that men are not invariably more aggressive than wom- en. Frodi et al. (1977) reported that in 61% of 72 aggression experiments in- volving both female and male subjects, men were not consistently more aggres- sive than women across all conditions.

What factors characterize situations in which women behave aggressively? Both anecdotal and empirical evidence seem to indicate that female aggression is most likely to occur in situations involving the enactment of such traditionally gender-appropriate roles as wife and/or mother. It is suggested that this situa- tional specificity of female aggressiveness is due to two factors. First, such situa- tions may be more central to women's self-definitions and consequently may be perceived as more important and more involving than neutral or masculine con- texts. Second, female aggression in sex-role congruent settings may be regarded as less socially unacceptable than such behavior would be in other situations, so that women may be less anxious and less reticent about expressing it.

Experimental evidence cited by Frodi et al. (1977) in their excellent re- view provides tangential support for the hypothesized situational importance factor. Males and females respond differently to a variety of frustrating cues and situations; for example, men are more aggressive and women are more ner- vous in crowds (Freedman, Levy, Buchanan, & Price, 1972; Schettino & Borden, 1976; Stokols, Rall, Pinner, & Schopler, 1973). However, telephone callers who hang up without saying anything make women more angry than men (Harris, 1974). Women also respond more aggressively to insult than do male subjects (Buvinic, 1976). Finally, while men report that physical aggression makes them most angry, women are most angered by unfair treatment (Novaco, 1975). Al- though none of these studies deals with differential responses to situations de- fined specifically in terms of sex-role appropriateness, they do argue for the consideration of gender-related factors in the interpretation of responses to cues and contexts defined a priori as equally frustrating for both sexes.

As for the second factor, the greater social acceptability of female aggres- sion in sex-role appropriate contexts, many studies indicate that women experi-

Gender and Aggression 905

ence more anxiety about aggression than do men (e.g., Matarazzo, 1954;Kagan & Moss, 1962; Wyer, Weatherly, & Terrell, 1965). There is also evidence that adults judging adult behavior are less approving of female than of male aggres- siveness (e.g., Broverman, Vogel, Broverman, Clarkson, & Rosenbrantz, 1972; Costrich, Feinstein, Kidder, Maracek, & Pascale, 1975; Dertke, Penner, Hawkins, & Suarez, 1973). According to several researchers (Bandura, 1973; Grusec, 1972; MaUick & McCandless, 1966), girls learn aggressive verbal and physical responses, which they demonstrate only if they perceive the situation as permitting females to act aggressively. Extending this observation to adult aggression research, Frodi et al. (1977) found that most of the studies in which there are no sex differences in aggression provide adequate justification for the aggressive behavior elicited. Similarly, increasing the salience of sex-role requirements may have significant effects. For male subjects, both evocation of manliness norms (Borden, 1975) and negation of the taboo on aggression against women (Hedges, 1970) increase the amount of aggressive behavior displayed.

The present study was designed to explore the notion that if men and women were faced with a frustrating situation involving traditionally masculine activities, the men would define the situation as more important and respond much more aggressively than would the women, thus duplicating the "natural" aggression pattern. However, if both women and men were asked to respond to a frustrating situation involving traditionally feminine activities, the women would define the situation as more important and their level of aggression would equal or exceed that of the men.

Of course, comparatively low levels of female aggression in tile masculine condition could be due to greater anxiety about expressing anger in a masculine context, rather than to a perception of the masculine situation as less important than the feminine one. This possibility was considered and controlled for in various ways. As participants in a social perception study, subjects were not re- quired to commit themselves publicly to a particular course of action. And the frustrating situation described to subjects was one in which some form of ag- gression was undoubtedly justified. Finally, despite the somewhat mixed results regarding sex differences in preferred response mode (see Frodi et al., 1977), subjects were able to indicate their feelings regarding the appropriateness of dif- ferent kinds of aggression - verbal, physical, prosocial, and passive.

METHOD

Subjects

The experiment and one replication were conducted. In the experiment, 21 males and 22 females in an introductory social psychology class served as

906 Towson and Zanna

subjects in one in-class session; in the replicat ion, 31 males and 31 females drawn

f rom a vo lun ta ry subject poo l associated wi th their i n t roduc to ry psychology

course par t ic ipated in mixed-sex group sessions o f 4 to 10 subjects.

Design

The s tudy consisted o f a 2 (sex o f subject) X 2 (sex-role or ienta t ion o f

task) factorial design. The lat ter variable was manipula ted by having each subject

read a vignette describing a frustrat ing s i tuat ion involving the per formance o f

ei ther t radi t ional ly masculine or t radi t ional ly feminine s k i l l s - b o d y building

or dance exercises, respectively.

Procedure

Subjects were asked to read the vignet te careful ly and to answer severn

quest ions about the probable react ions o f a typical college s tudent o f their own

sex to the s i tuat ion described. Subject a n o n y m i t y was assured.

Vignettes. Each vignet te consisted o f five paragraphs describing a frustrat-

ing sequence o f events involving mode rn dance exercises in the feminine task

condi t ions and bodY-building exercises in the mascul ine task condit ions. The si-

tua t ion was described as fol lows:

You are taking a co-educational exercise course which includes a section on dance (body building). This section of the course is designed to increase your grace (physical strength), improve your posture (physical endurance) and teach you to express emotions with your body (muscle control). You would like to do your best on the practical exam being given at the end of this section, not only to pass the course but also for your own satisfaction. On the day of the test, you go to the gym early to warm up. The [same-sex] student scheduled to take the test immediately after you is using the only available warm-up room (set of equip- ment), so you ask when s/he will be finished as you really need to practice. The student says s/he will be finished in 20 minutes and that s/he thinks the exam times have been rescheduled, so that s/he now precedes you. Half an hour later, after running some errands, you check the exam schedule and find that it has not been changed. Very upset, you return to the warm-up room to find the same student still there. After explaining that no schedule change has been made, you request that s/he leave the room so you can use it. The student responds by say- hag that it was your responsibility to check the schedule, but that s]he will leave the room after you have changed clothes. Upon returning to the warm-up room after changing, you find the door closed and locked with only five minutes to go before your exam. You run back to the dressing room, hoping to find the other student, who may have the key, and as you enter the dressing room, you see the student locking her/his locker.

Dependent Measures. The quest ionnaire consisted o f 16 i tems, each o f

which was answered on a Liker t - type scale. 3 Students were asked to respond to

3 Because of a clerical error, the scales, although all approximately the same length, had dif- ferent numbers of points, ranging from 21 to 26. Two additional scales had 50 points.

Gender and Aggression 907

all questions as they thought a typical same-sex college student would respond as a participant in the situation described in the vignette.

Six items were included as manipulation checks in order to ascertain whether female and male subjects perceived the feminine and masculine tasks, respectively, as more important. Three of these items dealt with the general importance of doing well on the practical test; related items measured the extent to which subjects regarded good performance on each of three specific task-rel- evant skills as important. For subjects in the body building, masculine task con- ditions, these skills were physical strength, physical endurance, and muscle con- trol. In the modern dance, feminine conditions, the skills were grace, correct posture, and expression of emotions through body movement.

The main dependent measures consisted of 6 items assessing subjects' feel- ings regarding the appropriateness of various assertive/aggressive responses in the situation described. Thoughts on the demonstration of a general, undifferenti- ated "assertiveness" in this situation were assessed with 2 items. An additional 4 items required judgments of the appropriateness of specific kinds of aggression - physical (e.g., pushing, shaking, striking); verbal (e.g., name calling); indirect, nonobvious (e.g., downgrading the student to other students, preventing the student from performing well in the future by monopolizing the exercise room); and appeal to legitimate authority (e.g., reporting the student to the relevant authority). The nonobvious, or "passive," aggression Category was included to represent the nonconfrontational ways of expressing hostility sometimes de- scribed as "backbiting" or "cattiness" and traditionally regarded as typical of women rather than of men. An appeal to legitimate authority was conceptual- ized by the authors as a form of "prosocial aggression," defined as aggression "used in a socially approved way for purposes that are acceptable to the moral standards of the group" (Sears, 1961, p. 471). Sears and his colleagues (Sears, 1961; Sears, Ran, & Alpert, 1965) found that prosocial aggression was more characteristic of girls than of boys; but since research on both passive and pro- social aggression is negligible, the inclusion of these categories was exploratory.

Finally, 4 items assessed subjects' perceptions of the uncooperative stu- dent's actions and intentions in sabotaging the protagonist's attempts to prac- tice. Subjects were asked whether they thought this student had been inten- tionally misleading about the schedule change, had taken the key, had wanted to maximize his/her own performance, and had wanted to prevent the other stu- dent from doing his/her best on the exam.

RESULTS

All dependent measures underwent 2 (sex of subject) × 2 (sex-role orienta- tion of task) × 2 (experiment/replication) unweighted means analyses of vari- ance. These analyses indicated significant main and interaction effects for experi- ment/replication on 4 items. Since these effects did not qualify the description

9011 Towson and Zanna

of the main results, the data were collapsed across this dimension; 2 (sex of sub- ject) × 2 (Sex-role orientation of task) unweighted means analyses of variance were then performed on all dependent measures.

Manipulation Checks

In order for the predicted aggressive reactions to occur, males and females in sex-role congruent conditions had to judge the described task as more impor- tant than subjects in sex-role incongruent conditions. Responses to 4 of the 6 manipulation checks indicated that this differential perception had occurred. A significant interaction effect (F(1, 101) = 4.72, p < .05) reflected the fact that women taking the dance unit thought they would feel more positive about placing in the top 10% of the class (M = 20.75) than did women completing the body-building unit (M = 19.42). For men, however, more positive feelings were anticipated by those taking body building (M = 19.83) than by those completing the dance course (M = 19.00). On all 3 items measuring perceived importance of task-relevant skills, significant interaction effects indicated the same pattern of results. As the specific pairings of "masculine" and "feminine" skills were ar- bitrary, these items were combined to form a masculine and feminine skills in- dex. The analysis of this index yielded a significant interaction effect (F(1,101) = 17.44, p < .01), reflecting once again that women regarded feminine skills as more important (M = 22.02) than masculine ones (M = 18.75), while the oppo- site was true for men (Ms of 17.58 and 21.03 for dancing and body building, re- spectively).

Only two manipulation checks failed to demonstrate the required effect. Females regarded good performance on the test as significantly more important in both masculine and feminine task conditions (Ms of 18.56 and 20.02, respec- tively) than did males (Ms of 17.81 and 16.82, respectively) (F(1, 101) = 6.55, p < .05). And neither main nor interaction effects were significant on the item dealing with hypothetical placement in the bottom 10% of same-sex students taking the test.

Muin Dependent Measures

Table I shows that, consistent with the hypothesis, male and female sub- jects regarded physical and verbal aggression as more appropriate in sex-role congruent situations (F(1, 101) = 4.38, p < .05, and F(1,101) = 7.44, p < .01, for physical and verbal aggression, respectively). Although males approved of both types of aggression more than did females (F(1,101) = 20.93, p < .01, and F(1, 101) = 22.00, p < .01 , for physical and verbal aggression, respectively), comparison of the relevant means indicates that the classic gender difference in aggression held only in the masculine task situation (t = 4.72, p < .001, for

Gender and Aggression 909

Table I. Mean Judgments of Aggressive Response Appro- priateness

t

Sex-role orientation of task Sex of subject Masculine Feminine

Physical response Male 8.94 6.42 Female 2.31 3.96

Verbal response Male 17.44 12.50 Female 7.96 10.02

Prosocial response Male 14.43 13.14 Female 11.42 12.46

Passive response Male 8.22 10.52 Female 4.81 8.09

aCell ns are as follows: male/masculine = 27, male/femi- nine = 25, female/masculine = 26, female/feminine = 27. All responses were made on 24-point scales; the higher the number, the more appropriate the aggressive re- sponse.

physical aggression and t = 5.25, p < .001, for verbal aggression): In the femi- nine task situation, this difference is reduced to a marginal level in the case of physical aggression (t = 1.73, p < .10) and eliminated in the case of verbal ag- gression (t = 1.36).

This reduction or elimination of gender differences in aggression seems to be attr ibutable to changes in both male and female responses. Males learning dance exercises regard physical aggression as marginally less appropriate ( t = 1.81, p < .10) than did those doing body-building exercises (Ms of 6.42 and 8.94, respectively). Females, on the other hand, regarded physical aggression as more appropriate in the dance context (M = 3.96) than in the body-building sce- nario (M = 2.31), although the difference between these means is not significant ( t = 1.17). The same pat tern held for verbal aggression, with men advocating a relatively high level when body building was involved (M = 17.44) and a signifi- cantly lower level ( 3 / = 12.50) for dance exercises (t = 2.71, p < .05). By com- parison, women approved more, although not reliably, of verbal aggression in defense of dance practice (M = 10.02) than for body-building exercises (M = 7.96, t = 1.14). The same pat tern of results is discernible in the mean responses to the suggestion of prosocial aggression, although the male-female difference was not significant even in the masculine condition.

910 Towson and Zanna

Consistent with their prior responses, females regarded passive aggression as marginally more appropriate (t = 1.92, p < .10) in the feminine (M = 8.09) than in the masculine condition (3/= 4.81). Males, deviating from their typical pattern, agreed with the female assessment and approved of passive aggression slightly more in the feminine than in the masculine condition (Ms of 10.52 and 8.22, respectively) although this difference was not significant (t = 1.25). As a result of this deviation, passive aggression was considered significantly more ap- propriate for male and female dancers than for male and female body builders F(1, 101) = 4.65, p < .05). Otherwise, results for passive aggression paralleled those obtained for the other aggressive response modes. Males approved of passive aggression more than did females (F(1, 101) = 5.09, p < .05), and comparison of means within conditions revealed that the classic gender differ- ence was marginally significant in the masculine condition (t = 1.87, p < .10) and eliminated in the feminine condition (t = 1.32).

Neither of the two questions regarding the general advisability of assertion differentiated subjects on the basis of task or sex. Perhaps the questions were too general or the word "assertion," used rather than the word "aggression," was open to too many interpretations.

Intention. The familiar interaction effect occurred on 2 of the 4 items dealing with the obstructive student's probable motivations. Male and female subjects for whom task was congruent with sex role were more convinced (Ms of 21.74 and 22.93, respectively) than those in incongruent conditions (Ms of 20.34 and 20.15, respectively) that the offending student had wanted to maxi- mize his/her own performance (F(1, i01) = 6.65, p < .05). These students also tended to be more convinced (Ms of 16.98 and 18.72, respectively) than their incongruent sex-role counterparts (Ms of 16.48 and 15.08, respectively) that the student had taken the exercise room key (F(1,101) = 3.75, p < .10).

Males and females in the feminine condition were more sure (Ms of 21.80 and 22.78, respectively) that the student had deliberately misled them (F(1,101) = 4.94, p < .05) than subjects in the masculine condition (Ms of 21.24 and 19.35, respectively). Regarding the idea that the student had monopolized the room deliberately to prevent the protagonist from doing his/her best, no signifi- cant main or interaction effects occurred.

DISCUSSION

The results support the hypothesis that both males and females define skills congruent with traditional sex roles as more important than incongruent ones. This differential attribution of importance is reflected in the f'mding that males advocated significantly more physical and verbal aggression than females in the masculine condition, a replication of the classic gender difference. Men in

Gender and Aggression 911

this condition also regarded passive aggression as a marginally more appropriate response than did women, despite the feminine sex-role connotations of this form of aggression. Thus, the typical pattern of significantly greater male aggres- sion held in the masculine condition, defined by females as relatively less impor- tant. In the feminine condition, defined as more important by the female sub- jects, this difference in aggression level was reduced for physical, and eliminated for verbal and passive, aggression.

Although both increased female and decreased male physical and verbal aggression in the feminine, as compared to the masculine, condition contributed to the significant interaction, only the decrease in the male subjects' responses was consistently reliable. Despite this corroboration of the hypothesis, the fact remains that although women were as aggressive as men in the feminine condi- tions, women were never m o r e aggressive than men. A possible explanation is that the experimental context may not have been completely successful in allay- ing women's sex-role specific aggression anxiety. Macaulay (Note 1) concludes that women's aggression anxiety is a complex mix of internal and social factors, and perhaps the examination component of the frustrating situation, with its explicit competitiveness, served as an anxiety-eliciting cue for women, but not for men. Tangential support for this suggestion is provided by female subjects' being significantly more concerned about good performance on both the femi- nine and masculine tests than were the males.

The consistency of the male-female aggressive expression pattern across the four response modes suggests that whatever factors contribute to females' lesser aggressiveness, this phenomenon is not the result of gender differences in response preference. Although folk wisdom suggests that women prefer verbal and indirect modes of aggression (Macaulay, Note 1), while men prefer physical ones, the data indicate that women approved less of every form of aggression than did men. In addition, examination of the order of the response mode means within each condition indicates that females and males shared a preference for verbal and prosocial, rather than physical and passive, aggression.

Perceived importance of sex-role congruent tasks may also have mediated subjects' interpretations of the obstructing student's intentions. Subjects in con- gruent sex-role conditions were more certain in their judgments than their incon- gruent sex-role counterparts, attributing the student's monopolization of the room to a desire to maximize his/her own performance. Congruent subjects were also more convinced that the student had deliberately taken the key to the locked warm-up room. Perhaps the fact that sex-role congruent subjects per- ceived the task as more important led them to regard the other student's activi- ties - the source of the frustration - as more arbitrary and unfair than did in- congruent sex-role subjects. As a consequence, the former's advocacy of greater aggression may have been based on a perception of the frustration as both higher in magnitude and less justified.

912 Towson and Zanna

Interpretation o f these results is limited by the fact that this was a social perception study. Regardless o f the subjects' degree o f involvement in the situa- tion described to them, their reactions undoubtedly would have been different had they been directly involved. The use o f a social perception procedure was dictated in part by ethical considerations; given the high traumatic potential of aggression, its manipulation in the laboratory should be undertaken only after extensive preliminary investigation. Social perception studies have distinct ad- vantages, however; this technique permits the assessment o f a greater number of dependent variables in more ways than would be possible using active partici- pants. Moreover, it allows direct determination of the extent to which subjects perceive various aggressive responses to be appropriate reactions to frustrating circumstances.

Based on these results, some tentative conclusions may be reached regard- ing the interpretation of past research and possible directions for future research. It is probable that gender differences in many frustration-aggression experiments have been due to female subjects' perceiving the frustrating situation as less im- portant and experiencing it as less involving than male subjects. Similarly, mani- pulations designed to minimize aggression anxiety may have been more effective for men than for women. Future aggression research should include measures assessing the effects o f these two factors. In addition, situational importance should be manipulated. If tile situation is sex-role specific and important and in- volving enough, female aggression may be found to exceed male aggression. Finally, future research on gender differences in aggression should include a mea- sure of subjects' sex-role orientation. 4 It seems probable that subjects' reactions to frustrating situations are determined in part by their self-definitions and atti- tudes regarding appropriate female and male behavior. Thus, for example, wom- en and men whose definitions o f sex roles are essentially androgynous may react less aggressively than more traditionally oriented subjects, not because androgy- nous persons are innately less aggressive, but because sex-role specificity no longer mediates perceived situational importance for them.

4In the present study, subjects completed the Personal Attributes Questionnaire (PAQ; Spence, Helmreich, & Stapp, 1975) and analyses of variance were conducted only on the data from subjects who were classifiable as "traditional," or nonandrogynous according to this scale. These analyses did not differ in any appreciable way from those including all subjects, however, so they are not reported here. The lack of significant effects unique to nonandrogynous subjects may be due to the characteristics of the PAQ. This scale was designed to measure instrumental and expressive personality traits, rather than sex-role orientation per se, and recent evidence (Helmrich, Spence, & Holahan, Note 2; Spence & Helmreich, 1978) suggests that men and women who are sex-typed in personality are not necessarily sex-typed on various sex-role attitudes or preferences. A measure assessing sex- role attitudes directly, such as the Attitudes toward Women Scale (Spenee, Helmreich, & Stapp, 1973) might have produced quite different results.

Gender and Aggression 913

REFERENCE NOTES

1. Macaulay, J. On women, anger and aggression. Unpublished manuscript, Wisconsin, 1977. (Available from 314 Shepard Terrace, Madison, Wisconsin 53705).

2. Helmreieh, R. L., Spence, J. T., & Holahan, C, K, Psychologicalandrogyny and sex-role flexibility: A test of two hypotheses. Unpublished manuscript, University of Texas at Austin, 1978.

REFERENCES

Bandura, A. Aggression: A social learning analysis. Englewood Chffs, N.J.: Prentice-Hall, 1973.

Berkowitz, L., & LePage, A. Weapons as aggression-eliciting stimuli. Journal of Personality and SocialPsychology, 1967, 7, 202-207.

Borden, R. J. Witnessed aggression: Influence of an observer's sex and values on aggressive responding. Journal of Personality and Social Psychology, 1975,31, 567-573.

Broverman, I., Vogel, S., Broverman, D., Clarkson, F., & Rosenkxantz, P. Sex-role stereo- types: A currant appraisal. Journal of SocialIssues, 1972, 28(2), 59-78.

Buvinic, M. Sex differences in the aggressive consequences of expressing feelings or opinions (Doctoral dissertation, University of Wisconsin, 1975), Dissertation Abstracts Inter- national, 1976, 36, 1498B. (University Microfilms No. 75-12, 890)

Costfich, N,, Feinstein, J., Kidder, L., Maracek, J., & Pascale, L. When stereotyPes hurt: Three studies of penalties for sex-role reversals. Journal of Experimental Social Psychology, 1975, 11, 520-530.

Dertke, M. C., Penner, L. A,, Hawkins, H. L., & Suarez, C. The inhibitory effects of an observer on instrumental aggression. Bulletin of the Psychonomic Society, 1973, 1, 112-114.

Eron, L. D., Huesmarm, L. R., Lefkowitz, M. M., & Waider, L. O. Does television violence cause aggression? American Psychologist, 1972, 27, 253-263.

Freedman, J. L., Levy, A. S., Buchanan, R. W., & Price, J. Crowding and human aggressive- ness. JournalofExperimentalSocialPsychology, 1972, 8, 528-548.

Frodi, A., Macaulay, J., & Thome, P. R. Are women always less aggressive than men? A review of the experimental hterature. Psychological Bulletin, 1977, 84, 634-660.

GeUes, R. J. The violent home: A study of physical aggression between husbands and wives. Beverly Hills, Calif.: Sage, 1972.

Gelles, R. J. Child abuse as psychopathology. American Journal of Orthopsychiatry, 1973, 43, 611-621,

Gil, D. C. Incidence of child abuse and demographic characteristics of persons involved. In R. E, Heifer & C. H. Kempe (Eds.), The battered child. Chicago: University of Chica- go Press, 1968.

Gil, D. C. Violence against children. Cambridge, Mass.: Harvard University Press, 1970. Grusec, J. E. Demand characteristics of the modeling experiment: Altruism as a function of

age and aggression. Journal of Personality and Social Psychology, 1972, 22, 139-148. Harris, M, B. Aggressive reactions to a frustrating phone call. Journal of SocialPsychology,

1974, 92, 193-198. Hedges, L. E. Aggressive responses as a function of target cues and the possibility of retalia-

tion (Doctoral dissertation, University of Iowa, 1969). Dissertation Abstracts Inter- national, 1970, 30, 3385B-3386B. (University Microfilms No. 69-21,696)

Kagan, 1., & Moss, H. A. Birih to maturity: A study in psychological development. New York: Wiley, 1962.

Maccoby, E. E., & Jacklin, C. N. The psychology of sex differences. Stanford: Stanford University Press, 1974.

Mallick, S. K., & McCandless, B. R. A study of catharsis of aggression. Journal of Personality and Social Psychology, 1966, 4, 591-595.

914 Towson and Zanna

Matarazzo, J. D. An experimental study of aggression in the hypertensive patient. Journal of Personality, 1954, 22, 423-447.

Novaco, R. W. Anger control. Lexington, Mass.: Lexington Books, 1975. Radbill, S. K. A history of child abuse and infanticide. In R. E. Heifer & C. H. Kempe

(Eds.), The battered child (2nd ed.). Chicago: University of Chicago Press, 1974. Schettino, A. P., & Borden, R. J. Sex differences in response to naturalistic crowding: Af-

fective reactions to group size and group density. Personality and Social Psychology Bulletin, 1976, 2, 67-70.

Sears, R. R, Relation of early socialization experiences to aggression in middle childhood. Journal of Abnormal and Social Psychology, 1961, 63, 466-492.

Sears, R. R., Rau, L., & Alpert, R. Identification and child rearing. Stanford: Stanford Uni- versity Press, 1965.

Smith, S. M. The battered child syndrome. London: Butterworth, 1975. Spence, J. T., & Helmreich, R. L. Masculinity and femininity: Their psychological dimen-

sions, correlates, and antecedents. Austin: University of Texas Press, 1978. Spence, J. T., Helmreich, R., & Stapp, J. A short version of the Attitudes toward Women

Scale (AWS). Bulletin of the Psychonomic Society, 1973, 2, 219-220. Spence, J. T., Helmreich, R., & Stapp, J. Ratings of self and peers on sex role attributes and

their relation to self-esteem and conceptions of masculinity and femininity. Journal of Personality and Social Psychology, 1975, 32, 29-39.

Stokols, D., Rall, M., Pinner, B., & Schopler, J. Physical, social, and personal determinants of the perception of crowding. Environment and Behavior, 1973, 5, 87-115.

Straus, M. A., Geiles, R. J., & Steinmetz, S. K. Theories, methods, and controversies in the study of violence between family members. In S. K. Steinmetz & M. A. Straus (Eds.), Violence in the family. New York: Dodd, Mead, 1974.

Wyer, R. S., Weathedy, D. A., & Terrell, G. Social role, aggression, and academic achieve- ment. Journal of Personality and Social Psychology, 1965,1,645-649.