Violence against female partners: Direct and interactive effects of family history, communal orientation, and peer-related variables

Gail M. Williamson

University of Georgia

Jay G. Silverman

Harvard University

Journal of Social and Personal Relationships, 18 (4) 535-549, 2001

ABSTRACT

Data from 172 male university students provide evidence for the role of communal orientation (dispositional beliefs that relationship partners should be mutually responsive to each other's needs) in predicting violence against female partners. Associating with peers who verbally endorse and behaviourally model dating violence predicted participants' violent behaviour. In addition, communal orientation, a personality trait not previously investigated in this context, was both directly and interactively related to dating violence. Men low in communal orientation were more likely to physically abuse their dating partners and to associate with peers who endorsed violence against female partners and who were themselves abusive. Moreover, less communal men appeared to be more subject than their highly communal counterparts to the influence of peers.

KEY WORDS: abuse. aggression. communal orientation. dating violence

Violence against female partners does not occur exclusively in the domain of marital relationships (Kropp & Hart, 1997). Women in heterosexual dating relationships are subject to similar, if not higher, levels of abuse (Koss et al., 1994; Makepeace, 1981; Stark & Flitcraft, 1992). For example, it is estimated that over 725,000 adolescent women are abused by male dating partners each year (Snyder & Sickmund, 1995). In a national study, one-third of college women reported experiencing physical violence in heterosexual relationships (White & Koss, 1991). Moreover, a growing percentage of female abuse-related homicides are perpetrated by male dating partners rather than husbands (Browne & Williams, 1993).

Individual differences not previously investigated in this context may be important in predicting men's violent behaviour against female relationship partners. In addition, personality characteristics may be most useful when conceptualised not solely as main effects but, rather, as factors that interact with situational variables (e.g., Blass, 1991; Lewin, 1936; Milgram, 1974; Mischel, 1968; Staub, 1999). The personality variable of interest in this study was communal orientation, and being high in this trait should be antithetical to abusing another person. However, individuals low in communal orientation may be more likely to engage in partner violence depending on situational cues, operationalised in this study as variables (i.e., witnessing paternal violence and associating with peers who verbally and behaviourally endorse abusive behaviour) previously shown to support an ecological perspective on relationship violence (e.g., DeKeseredy, 1988; DeKeseredy & Schwartz, 1993; Silverman & Williamson, 1997). We assumed that communal orientation would interact with ecological variables, that ecological variables would interact with each other, and that these interactions would account for significant variance in whether female relationship partners are abused beyond that explained by simple main effects.

Ecological variables

In several studies (e.g., Bernard & Bernard, 1983; O'Keefe, 1997; Riggs & O'Leary, 1996; Sack, Keller, & Howard, 1982), exposure to violence in the home has predicted violent behaviour among young men. However, in other research, witnessing domestic violence did not discriminate those who physically abuse female dating partners from those who do not (e.g., DeMaris, 1987; Marshall & Rose, 1987; O'Keefe, Brockopp, & Chew, 1986; Sigelman, Berry, & Wiles, 1984; Stets & Pirog-Good, 1987). Focusing on the family may ignore other potentially important sources (e.g., peers) of antisocial behaviour socialization (Barber & Olsen, 1997; O'Keefe, 1997; Silverman & Williamson, 1997). In other words, family experiences alone may not adequately account for relationship violence (Belsky, 1993). Indeed, rather than directly predicting dating violence, witnessing domestic violence seems to exert indirect effects through attitudes that favour violent behaviour toward women, beliefs that may be fostered by associating with abusive peers (Silverman & Williamson, 1997).

There is evidence that peers contribute to and maintain attitudes about violence against women (Andrews & Bonta, 1994; Silverman & Williamson, 1997). The development of moral, sex-appropriate, socially responsive, and normative social behaviours may be peer dependent (e.g., Cooke & Apolloni, 1976; Fagot, 1985; France-Kaatrude & Smith, 1985; Keena, Loeber, Zhang, & Stouthamer, 1995; Walker & Taylor, 1991). Similarly, DeKeseredy (1988; DeKeseredy & Schwartz, 1993) proposed that when male peers sanction abusing women, an ideology is fostered that legitimises relationship violence. If a man's partner deviates from his peer group's ideals of male dominance, group members may provide 'informational support' that encourages partner abuse and, by modelling violence against women, further validate abusive behaviour.

Empirical evidence supports the DeKeseredy model. Male undergraduates are more likely to be abusive when their peers indicate that it is appropriate to do so (DeKeseredy, 1990; DeKeseredy & Kelly, 1993; Silverman & Williamson, 1997). In addition, young men whose peers engage in violence against female partners are more likely to be violent themselves (DeKeseredy & Kelly, 1993; Silverman & Williamson, 1997). On a more encouraging note, the DeKeseredy model implies that peers also can discourage abusing female partners. Indeed, the relation between having witnessed paternal violence against women and abusing one's own relationship partner is weakened by peers who do not support such behaviour (Silverman & Williamson, 1997). We proposed that a dispositional communal orientation toward relationships with others can buffer the association between exposure to male violence (paternal, peer-related, or both) and perpetrating dating violence.

Communal orientation

In conjunction with specifying rules that govern behaviour in different types of interpersonal relationships (e.g., Clark & Mills, 1979, 1993; Mills & Clark, 1982), Clark and her colleagues have identified dispositional differences in attitudes and beliefs about what qualifies as appropriate behaviour in close relationships (Clark, Ouellette, Powell, & Milberg, 1987). In particular, compared with those low in communal orientation, individuals high in this trait characteristically believe that they should meet others' needs and also that others should meet their needs (Williamson & Clark, 1989).

Through socialization, most people come to believe that mutual sensitivity represents the ideal in interpersonal behaviour (e.g., Clark & Mills, 1993), but individuals low in communal orientation do not possess these beliefs about how people ought to treat each other. Highly communal people are dispositionally inclined to understand another person's perspective (e.g., in times of conflict), be sympathetic to that perspective, and remain attentive to the other's needs while, at the same time, expecting the other person to treat them in similar fashion. Thus, abusing a dating partner should grossly violate the 'golden rule' beliefs of someone high in communal orientation. Less communal individuals, conversely, are unlikely to take the other's perspective, feel sympathetic to that perspective, or be attentive to the other's needs. Of particular relevance is that less communally oriented people may be insensitive to their partners' needs and, especially during disagreements, focus instead on assuring that their own needs are met.

Evidence indirectly supports this perspective. Lack of empathic feeling for others contributes to instrumental violence (e.g., Cornell et al., 1996), and people low in communal orientation are less empathic than are those high in this trait (Clark et al., 1987). But, the constructs of empathy and communal orientation are distinct both conceptually and empirically (i.e., share only about one-third variance, Clark et al., 1987). That they are related at all reflects the component of communal orientation specifying that attending to others' needs is appropriate behaviour. That the relation is no more than moderate reflects the fact that measures of empathy do not assess the component of communal orientation specifying that others should be responsive to one's own needs. However, this element of mutuality is important because interpersonal relationships that are mutually responsive tend to be less conflicted, more satisfying, and highly valued (e.g., Clark & Mills, 1993; Hays, 1984, 1985; Sperberg & Stabb, 1998).

Other results also suggest that communal orientation is important in predicting violence against female partners. First, attitudes condoning such behaviour are related to relationship violence (e.g., Kropp & Hart, 1997; Silverman & Williamson, 1997). Second, violent men are less likely to take their partner's perspective or provide emotional support in response to her personal problems (Holtzworth-Munroe, Stuart, Sandin, Smutzler, & McLaughlin, 1997). And, third, low communal orientation fosters less attentiveness to a partner's needs when the partner is in a sad mood (Clark et al., 1987). These findings fit well with our view of men dispositionally low in communal orientation, i.e., that their attitudes and beliefs lead them to be unsympathetic to (and, in fact, ignore) their partner's perspective and, consequently, be inattentive to their partner's needs. We propose that these factors then make less communally oriented men more likely to abuse their partners, particularly during periods of conflict or disagreement.

Hypotheses and goals

Witnessing paternal violence and associating with peers who endorse violence against women may be related to abusing one's own female relationship partners (e.g., DeKeseredy, 1990; DeKeseredy & Kelly, 1993; Hotaling & Sugarman, 1986; Silverman & Williamson, 1997; Straus, Gelles, & Steinmetz, 1980). Moreover, being exposed to violence in the family of origin may lead to associating with peers who both advocate and perpetrate partner abuse (Silverman & Williamson, 1997), maintaining and promoting dating violence (DeKeseredy, 1990; DeKeseredy & Kelly, 1993).

However, our primary intent was to demonstrate that dispositional communal orientation predicts which men are disposed to behave violently toward the women with whom they have intimate relationships. Young men who are low in communal orientation should be more likely to associate with peers who endorse and perpetrate violence against women, and these men also should be more prone to abusing their own female partners.

A second primary goal was to show that, beyond their simple main effects, interactions between communal orientation and other predictor variables can explain significant variance in dating violence. In this context, we expected that the men most likely to be violent with their female partners would be those low in communal orientation who had been exposed to attitudes and behaviours that endorse and model violence against women.

Method

Sample and procedure

The data employed in these analyses were selected from a larger study (see Silverman & Williamson, 1997, for descriptive data and preliminary analyses from the total sample) to yield a sample of 172 single, heterosexual college men who provided complete data on our variables of interest. Undergraduate men enrolled in introductory psychology classes received partial course credit for completing a packet of questionnaires in small group testing sessions. No personally identifying information was recorded, thereby assuring confidentiality. The vast majority (96.9%) were White, precluding analyses of the effects of ethnicity. Mean age was 20.6 years (SD = 2.4). Slightly over half (55.2%) were freshmen and sophomores, with the remainder being upperclassmen. All participants were currently (or recently had been) in heterosexual dating relationships for, on average, 16.3 months (SD = 15.2, range 1-84 months).

Measures

Outcome variable - Participant-perpetrated dating violence. The Conflict Tactics Scale (CTS; Straus, 1979) was adapted to assess physical violence perpetrated in the last year against female partners. The introduction to this instrument advised that 'No matter how well a couple gets along, there are times when they disagree, get annoyed with the other person, or just have spats or fights because they're in a bad mood or tired or for some other reason. They also use many different tactics in settling their differences. Below is a list of some ways that you might have behaved toward your girlfriends and/or dating partners in these circumstances.' Participants reported how often (0 = never, 6 = more than 20 times) they had used eight categories of violent behaviour (thrown something at; pushed, grabbed, or shoved; slapped; kicked, bit, or punched; hit or tried to hurt with something; beat up; threatened with a knife or gun; used a knife or gun) with their girlfriends or dates. Responses were summed to produce a measure of dating violence with a Cronbach's alpha of .92. Scores could range from ° to 48. The mean score was low (i.e., 0.9). However, there was adequate variability (SD = 3.8, range 0-42), and 19.2 % of the participants reported one or more of these behaviours in the last 12 months.

Ecological predictor variables

Witnessing paternal violence. The same CTS items and similar instructions with a yes (1) or no (0) response format assessed whether participants had witnessed their fathers or male guardians physically abuse their mothers or other adult female partners (e.g., wives or girlfriends). Responses were summed to create a measure of witnessed paternal violence with a possible range of 0-8 and a Cronbach's alpha of .82. The mean score was 0.5 (SD 1.2, range 0-6). One-fifth (20.3%) of this sample reported witnessing their fathers or male guardians perpetrate at least one type of violent behaviour against female partners.

Peer advice. A measure of 'informational support' devised by DeKeseredy and Kelly (1993) assessed advice received from peers to abuse one's dating partners. Participants were told that 'Friends often are a major source of all kinds of advice on how to handle relationships,' and then responded (0 = no, 1 = yes) to seven questions, each prefaced by 'Have any of your male friends told you that. . . .' (1) 'you should respond to your dates' or girlfriends' challenges to your authority by using physical force, such as hitting or slapping?' (2) 'it is alright for a man to hit his date or girlfriend in certain situations?' (3) 'your dates or girlfriends should have sex with you when you want?' (4) 'if a man spends money on a date, she should have sex with him in return?' (5) 'you should respond to your dates' or girlfriends' challenges by insulting them or putting them down?' (6) 'you should respond to your dates' or girlfriends' sexual rejections by employing physical force to obtain sex?' and (7) 'it is alright for a man to physically force a woman to have sex with him under certain conditions?' DeKeseredy and Kelly (1993) reported an internal reliability for the summed score of. 70, and Cronbach's alpha in this study was. 74. Although the mean was low in this sample (0.8), there was considerable variability (SD = 1.3, range 0-7), and almost two-fifths (39.50;0) reported receiving peer advice to abuse their female partners.

Abusive peers. The number of participants' male friends who mistreated their dating partners was assessed with responses to three scenarios. Specifically, the introduction to this measure stated that 'Friends often share dating experiences with each other, both good and bad.' Participants then indicated the number of male friends (1 = none, 5 = more than 10) who, to the best of their knowledge, (1) 'made physically forceful attempts at sexual activity with women they were dating which were disagreeable and offensive enough that the woman responded by crying, fighting, screaming or pleading,' (2) 'used physical force, such as hitting or beating, to resolve conflicts with their girlfriends and/or dating partners to make them fulfil some demand,' and (3) 'insult their dating partners and/or girlfriends, swear at them, and/or withhold affection.' DeKeseredy and Kelly (1993) reported an internal reliability for this instrument of .65; in the present sample, Cronbach's alpha was .75. The mean score was 4.5 (SD = 1.6, range 3-15), and almost three-quarters (73.8%) of the participants reported that at least one or two of their male friends used these types of behaviours against their dating partners.

Communal orientation. The Communal Orientation Scale (COS; Clark et al., 1987; Williamson & Schulz, 1990) consists of 14 statements measuring a dispositional desire to respond to others' needs (e.g., 'When making a decision, I take other people's needs and feelings into account,' 'I am sensitive to other people's feelings') and dispositional desire for others to respond to one's own needs (e.g., 'It bothers me when other people neglect my needs,' 'I expect people I know to be responsive to my needs and feelings'). Each statement was rated on a 5-point scale (1 = strongly disagree, 5 = strongly agree). Scores could range from 14 to 70, with higher scores indicating greater communal orientation. The COS has demonstrated adequate psychometric properties. For example, in a sample of 561 college students, Cronbach's alpha was .78, and test-retest reliability in a sample of 128 students at an 11-week interval was .68 (Clark et al., 1987). In this group of college males, Cronbach's alpha was .82. The mean COS score was 54.2 (SD = 5.8, range 27-68), indicating relatively high average levels of communal orientation.

Results

Bivariate analyses

To determine whether we should control for participant age, year in college, and length of dating relationship, we conducted zero-order correlations. As shown in Table 1, none of these variables was related to any of the variables of interest in this study, and, therefore, these factors were excluded from further analyses.

Also displayed in Table 1 are intercorrelations between our study variables. Previous research indicates that witnessing paternal violence and associating with peers who endorse violence against women are related to abusing one's own female relationship partners. Our data are, in part, consistent with these findings. That is, although witnessing paternal violence was not related to partner abuse in this sample, associating with peers who verbally endorsed and behaviourally modelled such behaviour were both significantly correlated with partner abuse. In addition, witnessing paternal violence was related to having peers who verbally endorsed analogous behaviour, and peer advice in favour of abusing women was associated with having peers who actually engaged in abusive behaviour.

However, as hypothesised, men higher in communal orientation were less likely to behave in abusive ways toward their partners. Moreover, results were consistent with our expectation that men higher in communal orientation would be less likely to associate with abusive peers but might, nevertheless, find themselves in these kinds of peer groups. Bivariate correlations supported the first part of this hypothesis. That is, COS scores were negatively related to peer advice and having abusive peers. Additional analyses supported the second part of this hypothesis, indicating that this trend was not universal. Specifically, among men scoring in the highest quartile on the COS (i.e., scores of 58 or above), almost two-fifths (39.5%) reported that their peers provided advice favouring abusive behaviour, and a high percentage (79.1 %) reported that some of their peers mistreated female relationship partners. Thus, even highly communal men may associate with abusive peers.

TABLE 1

One nonsignificant result in these bivariate analyses is particularly noteworthy. That is, there was no evidence for an association between witnessing paternal violence and communal orientation.

Hierarchical regression analyses

We predicted that, beyond their simple main effects, interactions among communal orientation and other predictor variables would explain significant variance in dating violence. Tests for moderation effects were conducted according to methods recommended by Baron and Kenny (1986), and the results of these analyses are shown in Table 2. Main effects of witnessing paternal violence, communal orientation, peer advice, and abusive peers were entered in the first step of a hierarchical regression equation. The second step consisted of all possible two-way interactions between these four variables, and the third step contained all possible three-way interactions. The four-way interaction term entered in the final step explained no additional variance (i.e., R²change = .00), and results for this step are not shown in Table 2.

TABLE 2

Main effects of lower communal orientation, more peer advice, and more peers who were themselves abusive (Step 1) explained 40% of the variance in dating violence. Two-way interactions entered in Step 2 explained an additional 39% of the variance. However, two-way interactions appeared to be qualified by a three-way interaction. Specifically, entry of the three-way interaction terms in Step 3 explained a small, but marginally significant, 1% of the variance (Fchange = 2.31, p < .06) in dating violence. More important, with the entry of the three-way interactions, all main effects became non-significant as did all two-way interactions with one exception. That is, the interaction between peer advice and peer abuse remained significant in Step 3. Of the possible three-way interactions, only the communal orientation X peer advice X peer abuse term was significant.

To examine the pattern of this three-way interaction, regression slopes (illustrated in Figure 1) were calculated as recommended by Jaccard, Turrisi, and Wan (1990) using cut-point scores for the 10th and 90th percentiles to represent the form of this interaction across the range of variables involved. As can be seen, the worst-case scenario appears to be in men dispositionally low in communal orientation who receive high levels of advice from their peers supporting such behaviour and whose peers abuse their relationship partners. Results also suggest that additive effects of aspects of peer influence may be all-important for men low in communal orientation. In this group, if peers did not both abuse their dating partners and verbally advocate dating violence, these men were less likely to be abusive.

FIGURE 1

Discussion

As in previous research (e.g., DeKeseredy, 1990; DeKeseredy & Kelly, 1993; Silverman & Williamson, 1997; Straus et al., 1980), associating with peers who advocate and perpetrate dating violence was related to abusing one's own female partners. In addition, having witnessed domestic violence in the family predicted associating with peers who endorse analogous behaviour. A plausible interpretation is that internalised social norms acquired in the family of origin (e.g., that abusing women is acceptable, and, perhaps, rewarded) lead men to gravitate toward peer groups that condone similar behaviour, further exacerbating the influence of being exposed to paternal violence against women.

These data extend previous findings by demonstrating that the personality trait of communal orientation has both direct and interactive utility in predicting which men are likely to abuse their female partners. These results are important for several reasons. First, as hypothesised, men who were less communally oriented (i.e., those who did not personally subscribe to societal ideals about how people should treat each other) were more likely to belong to peer groups who both verbally and behaviourally endorsed abusive behaviour toward women and also to admit that they had behaved in violent ways toward their own partners. A second notable aspect of these analyses is that they illustrate how interactions between predictor variables can promote our understanding of the complex processes involved in violent behaviour - information less likely to be obtained from simple main effects models.

Consider our results as an example. In both bivariate and multivariate analyses, being lower in communal orientation, receiving higher levels of advice from peers, and associating with peers who abuse women all emerged as direct predictors of physical violence against female partners. From these results, we might conclude that these three variables qualify as risk factors for violent behaviour against women. Because, together, their main effects accounted for 40% of the variance in violent relationship behaviour, this interpretation would not be erroneous. However, we argue that it is simplistic.

Indeed, we anticipated that interactions between predictor variables would account for significant variance beyond main effects and found evidence to support this hypothesis. After controlling for the main effects of witnessing paternal violence, communal orientation, peer advice, and peer abuse, interactions explained an additional 39% of the variance in dating violence. Thus, the total equation, including interactions, allowed us to account for 79% of the variability in violent relationship behaviour perpetrated in the last year by the college men in this sample.

These results clearly indicate that the predictive utility of the main effects of witnessing paternal violence, communal orientation, and peer-related variables, even in conjunction, is less than perfect. All men whose fathers or male guardians abuse women do not become abusive. The same is true for all men low in communal orientation and all men whose peers verbally and behaviourally sanction abusing female relationship partners. Rather, interactions yielded clues about which men are most likely to engage in this type of behaviour.

Moreover, we found evidence that peer variables operate in conjunction. That is, even men low in communal orientation were unlikely to he abusive when their peers did not both verbally and behaviourally endorse such behaviour. Thus, as proposed in the DeKeseredy model of peer influence (1988; DeKeseredy & Schwartz, 1993), peers who both advocate partner abuse and model abuse through their own behaviour may exert a powerful influence on young men. However, individuals high in communal orientation were far less vulnerable to this potent combination than were their less communal counterparts. In other words, our results also qualify the DeKeseredy model by suggesting that peer influence may matter most in certain segments of the population, in this case, men dispositionally low in communal orientation. In fact, the impact of these variables appears to be additive. That is, men low in communal orientation with peers who both advocated and perpetrated dating violence were more likely to physically abuse their own partners than were those who did not fall into all three of these categories.

The limitations of this study warrant noting several caveats. First, as in many previous studies (e.g., DeKeseredy, 1988; DeKeseredy & Kelly, 1993; Makepeace, 1981; Marshall & Rose, 1987; Sigelman et al., 1984; White & Koss, 1991), these men were college students, young (i.e., approximately 21 years old), most likely relatively affluent, almost exclusively White, and entirely heterosexual. Therefore, these results may not generalise to men in other demographic groups. However, if our sample could be considered at lower risk for dating violence based on, for example, level of education, this should work against, rather than for, finding results of the magnitude identified in this study.

Although it could be argued that our results are subject to the alternative explanation that men scoring high in communal orientation simply responded to all measures in a socially desirable fashion, there are reasons to believe that this was not the case. First, the Communal Orientation Scale is not related to the Crowne-Marlowe (1964) measure of social desirability (Clark et al., 1987). Second, this study was designed so that participants completed measures in-group testing sessions and were aware that no personally identifying information was recorded. Moreover, the highly communal group did not uniformly respond in socially desirable directions. Specifically, in post-hoc analyses, men high in communal orientation were neither more nor less likely than their less communal counterparts to admit that they had witnessed paternal violence or that some of their friends advocated abusing or, in fact, abused their female relationship partners. Thus, social desirability likely had only a negligible effect, if any.

Third, these are cross-sectional data that bear replication in longitudinal research to better test the directions of observed effects. A unidirectional path was assumed in this research, but it may be that critical changes occur over time such that behaviours originally fostered by external influences ultimately become internalised and, consequently, more accurately attributable to the person (e.g., Darley, 1999). In other words, more complex models are required to investigate reciprocal pathways and bi-directional associations.

A fourth point is that these data provide no information, other than null results, about how communal orientation develops. Witnessing paternal violence and communal orientation were not bivariately related nor did they interact in predicting dating violence, indicating that witnessing paternal violence does not mean that a man necessarily develops low communal orientation. What, then, determines communal orientation? According to Williamson and Clark (1989), it most likely depends on parental socialization through which children typically adopt orientations consistent with those of their primary caregivers. In a parallel fashion, attachment representations may be transmitted between generations (see Shaffer, 2000, for a review). That is, maternal models of self and others are associated with both their infants' and their own mothers' attachment representations (Benoit & Parker, 1994). Thus, both communal orientation and attachment style may be independent of fathers' behaviour and traits, but highly dependent on those of their mothers or other primary caregivers (see Staub, 1999, for a related discussion). There is some evidence that an insecure attachment style is moderately related to both low communal orientation (Clark, personal communication, 1999) and aggressive behaviour in dating and other relationships (e.g., Bookwala & Zdaniuk, 1998; Dutton, 1998). Our data further indicate that a less communal orientation is associated with more dating violence, at least when there is peer support for such behaviour. These results suggest several interesting and potentially important directions for future research. Among these is how high communal orientation, secure attachment, or both, develop in an abusive family environment. Evidence reveals that communal orientation and attachment are related but share only about 25% of the variance (Clark, personal communication, 1999). Thus, both may depend, to some extent, on caregiver responsiveness to the child's needs. But, clearly, the developmental process for each construct varies according to factors yet to be identified.

We cannot ignore the impact of historic, social and, legal inequalities between men and women in the development of low communal orientation and hostility toward female partners. As this study and others (e.g., DeKeseredy & Kelly, 1993) demonstrate, violence against women has been practised and endorsed by the families and peers of substantial percentages of men. Congruent messages from the media, law, and other institutions may further socialise some men to disregard the needs of others, especially those of female partners (Birns, Cascardi, & Meyer, 1994; Muehlenhard & Kimes, 1999; Senn, Desmarais, Verberg, & Wood, 2000; Staub, 1999).

It is also worth considering that the measure of communal orientation utilised in this study was global rather than relationship specific. Men may be abusive in their relationships with female partners but not with, for example, co-workers, friends, and other women (Adams, 1988; American Psychological Association, 1996; Saunders, 1992). Assessing communal orientation and behaviours in specific types of relationships is an interesting direction for future research. In this regard, we believe that people who are truly high in communal orientation tend to be communally oriented in all types of relationships. But, it may well be that those marginally high in communal orientation are more subject to behaving communally in some relationships (e.g., with peers) than others (e.g., with female partners).

More attention to socialising factors beyond the family of origin (e.g., peer influence) is also warranted (e.g., Belsky, 1993; Senn et al., 2000). The same is true for developmental processes that foster individual differences (e.g., communal orientation) that dispose men to mistreat their female partners. Perhaps most importantly, this study demonstrates that examining the interplay between predictors of violent relationship behaviour is a worthwhile endeavour.

References

Adams, D. (1988). Treatment models of men who batter: A profeminist analysis. In K. Yllo & M. Bograd (Eds.), Feminist perspectives on wife abuse (pp. 176-200). Beverly Hills, CA: Sage.

American Psychological Association. (1996). Violence and the family: Report of the American Psychological Association Presidential Task Force on Violence and the Family. Washington, DC: American Psychological Press.

Andrews, D. A., & Bonta, J. (1994). The psychology of criminal conduct. Cincinnati, OH: Anderson.

Barber, B. K., & Olsen, J. A. (1997). Socialization in context: Connection, regulation, and autonomy in the family, school, and neighbourhood, and with peers. Journal of Adolescent Research, 12, 287-315.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Belsky, J. (1993). Parent, infant, and social-contextual antecedents of father-son attachment security. Developmental Psychology, 32, 905-913.

Benoit, D., & Parker, K. C. H. (1994). Stability and transmission of attachment across three generations. Child Development, 65, 1444-1456.

Bernard, M. L., & Bernard, J. L. (1983). Violent intimacy: The family as a model for love relationships. Family Relations, 32, 283-286.

Bims, B., Cascardi, M. A., & Meyer, S. (1994). Sex-role socialisation: Developmental influences on wife abuse. American Journal of Orthopsychiatry, 64, 50-59.

Blass, T. (1991). Understanding behaviour in the Milgram obedience experiment: The role of personality, situations, and their interactions. Journal of Personality and Social Psychology, 60, 398-413.

Bookwala, J., & Zdaniuk, B. (1998). Adult attachment styles and aggressive behaviour with dating relationships. Journal of Social and Personal Relationships, 15, 175-190.

Browne, A., & Williams, K. R. (1993). Gender, intimacy, and lethal violence: Trends from 1976 through 1987. Gender and Society, 7, 78-98.

Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and communal relationships. Journal of Personality and Social Psychology, 37, 12-24.

Clark, M. S., & Mills, J. (1993). The difference between communal and exchange relationships: What it is and is not. Personality and Social Psychology Bulletin, 19, 684-691.

Clark, M. S., Ouellette, R., Powell, M. C., & Milberg, S. (1987). Recipient's mood, relationship type, and helping. Journal of Personality and Social Psychology, 53, 94-103.

Cooke, T., & Apolloni, T. (1976). Developing positive social-emotional behaviours: A study of training and generalization effects. Journal of Applied Behaviour Analysis, 9, 65-78.

Cornell, D. G., Warren, J., Hawk, G., Stafford, E., Oram, G., & Pine, D. (1996). Psychopathy in instrumental and reactive violent offenders. Journal of Consulting and Clinical Psychology, 64, 783-790.

Crowne, D., & Marlowe, D. (1964). The approval motive. New York: Wiley.

Darley, J. M. (1999). Methods for the study of evil-doing actions. Personality and Social Psychology Review, 3, 269-275.

DeKeseredy, W. (1988). Woman abuse in dating relationships: The relevance of social support theory. Journal of Family Violence, 3, 1-13.

DeKeseredy, W. (1990). Male peer support and woman abuse: The current state of knowledge. Sociological Focus, 23, 129-139.

DeKeseredy, W., & Kelly, K. (1993). Woman abuse in university and college dating relationships: The contribution of the ideology of familial patriarchy. Journal of Human Justice, 4, 25-52.

DeKeseredy, W., & Schwartz, M. D. (1993). Male peer support and woman abuse: An expansion of DeKeseredy's model. Sociological Spectrum, 13, 393-413.

DeMaris, A. (1987). The efficacy of a spouse abuse model in accounting for courtship violence. Journal of Family Issues, 8, 291-305.

Dutton, D. G. (1998). The abusive personality: Violence and control in intimate relationships. New York: Guilford Press.

Fagot, B. I. (1985). Beyond the reinforcement principle: Another step toward understanding sex-role development. Developmental Psychology, 21, 1097-1104.

France-Kaatrude, A., & Smith, W. P. (1985). Social comparison, task motivation, and the development of self-evaluative standards in children. Developmental Psychology, 21, 1080-1089.

Hays, R. B. (1984). The development and maintenance of friendship. Journal of Social and Personal Relationships, 1, 75-98.

Hays, R. B. (1985). A longitudinal study of friendship development. Journal of Personality and Social Psychology, 48, 909-924.

Holtzworth-Munroe, A., Stuart, G. L., Sandin, E., Smutzler, N., & McLaughlin, W. (1997). Comparing the social support behaviours of violent and non-violent husbands during discussions of wife personal problems. Personal Relationships, 4, 395-412.

Hotaling, G. T., & Sugarman, D. B. (1986). An analysis of risk markers in husband to wife violence: The current state of knowledge. Violence and Victims, 1, 101-124.

Jaccard, J., Turrisi, R., & Wan, C. K. (1990). Interaction effects in multiple regression. Newbury Park, CA: Sage.

Keena, K., Loeber, R., Zhang, O., & Stouthamer, M. (1995). The influence of deviant peers on the development of boys' disruptive and delinquent behaviour. Development and Psychopathology, 7, 715-726.

Koss, M. P., Goodman, L. A., Browne, A., Fitzgerald, L. F., Keita, G. P., & Russo, N. F. (1994). No safe haven: Male violence against women at home, at work, and in the community. Washington, D.C.: American Psychological Press.

Kropp, P. R., & Hart, S. D. (1997). Assessing risk of violence in wife assaulters: The Spousal Assault Risk Assessment Guide. In C. D. Webster & M. A. Jackson (Eds.), Impulsivity: Theory, assessment, and treatment (pp. 302-325). New York: Guilford Press.

Lewin, K. (1936). A dynamic theory of personality. New York: McGraw-Hill.

Makepeace, J. M. (1981). Courtship violence among college students. Family Relations, 30, 97-102.

Marshall, L. L., & Rose, P. (1987). Gender, stress, and violence in adult relationships in a sample of college students. Journal of Social and Personal Relationships, 4, 299-316.

Milgram, S. (1974). Obedience to authority. New York: Harper & Row.

Mills, J., & Clark, M. S. (1982). Communal and exchange relationships. In L. Wheeler (Ed.), Review of personality and social psychology (pp. 121-144). Beverly Hills, CA: Sage.

Mischel, W. (1968). Personality and assessment. New York: Wiley.

Muelenhard, C. L., & Kimes, L. A. (1999). The social construction of violence: The case of sexual and domestic violence. Personality and Social Psychology Review, 3, 234-245.

O'Keefe, M. (1997). Predictors of dating violence among high school students. Journal of Interpersonal Violence, 12, 546-568.

O'Keefe, M., Brockopp, K., & Chew, E. (1986). Teen dating violence. Social Work, 31, 465-468.

Riggs, D. S., & O'Leary, K. D. (1996). Aggression between heterosexual dating partners. Journal of Interpersonal Violence, 11, 519-540.

Sack, A. R., Keller, I. F., & Howard, R. D. (1982). Conflict tactics and violence in dating situations. International Journal of the Sociology of the Family, 12, 89-100.

Saunders, D. G. (1992). A typology of men who batter. American Journal of Orthopsychiatry, 62, 264-275.

Senn, C. Y., Desmarais, S., Verberg, N., & Wood, E. (2000). Predicting coercive sexual behaviour across the lifespan in a random sample of Canadian men. Journal of Social and Personal Relationships, 17, 95-113.

Shaffer, D. R. (2000). Social and personality development (4th ed.). Belmont, CA: Wadsworth. Sigelman, C. K., Berry, C. K., & Willes, K. A. (1984). Violence in college students' dating relationships. Journal of Applied Social Psychology, 5, 530-548.

Silverman, I. G., & Williamson, G. M. (1997). Social ecology and entitlements involved in battering by heterosexual college males: Contributions of family and peers. Violence and Victims, 12, 147-164.

Snyder, H. N., & Sickmund, M. (1995). Juvenile offenders and victims: A national report. Washington, DC: US Department of Justice.

Sperberg, E. D., & Stabb, S. D. (1998). Depression in women as related to anger and mutuality in relationships. Psychology of Women Quarterly, 22, 223-238.

Stark, E., & Flitcraft, A. H. (1992). Spouse abuse. In I. M. Last & R. B. Wallace (Eds.), Public health and preventative medicine (13th ed., pp. 1040-1043). Norwalk, CT: Appleton & Lange.

Staub, E. (1999). The roots of evil: Social conditions, culture, personality, and basic human needs. Personality and Social Psychology Review, 3, 179-192.

Stets, I. E., & Pirog-Good, M. A. (1987). Violence in dating relationships. Social Psychology Quarterly, 50, 237-246.

Straus, M. A. (1979). Measuring intrafamily conflict and violence: The conflict tactics scales. Journal of Marriage and the Family, 41, 75-86.

Straus, M. A., Gelles, R. I., & Steinmetz, S. (1980). Behind closed doors: Violence in the American family. Garden City, NY: Anchor Press.

Walker, L. I., & Taylor, I. H. (1991). Family interactions and the development of moral reasoning. Child Development, 62, 264-283.

White, I. W., & Koss, M. P. (1991). Courtship violence: Incidence in a national sample of higher education students. Violence and Victims, 6, 247-256.

Williamson, G. M., & Oark, M. S. (1989). The communal/exchange distinction and some implications for understanding justice in families. Social Justice Research, 3, 77-103.

Williamson, G. M., & Schulz, R. (1990). Relationship orientation, quality of prior relationship, and distress among caregivers of Alzheimer’s patients. Psychology and Aging, 5, 502-509.

The data reported in this article represent a portion of the second author's master's thesis research, which was supported by a University - Wide Fellowship from the Graduate School at The University of Georgia. Manuscript preparation was facilitated by Grant AG15321 from the National Institutes of Mental Health (G. M. Williamson, principal investigator) and funds from the Institute for Behavioural Research at The University of Georgia to the first author. Correspondence concerning this article should be addressed to Gail M. Williamson, Department of Psychology, The University of Georgia, Athens, GA 30602, USA. [E-mail: LGMW@arches.uga.edu]. Barbara Sarason was the Action Editor on this article.