Journal of Youth and Adolescence, Vol. 27, No. 6, 1998

Conflict Resolution Styles in Family Subsystems and Adolescent Romantic Relationships Maria Reese-Weber1 and Suzanne Bartle-Haring2 Received October 7, 1997; accepted May 18, 1998

The purpose of the present study is to examine how conflict resolution styles between one family dyad is related to other family dyads and how conflict resolution styles within these family dyads are related to conflict resolution styles in one relationship outside the family—adolescents’ romantic relationships. Late adolescents (n = 217) were asked to report their perceptions of three resolution styles used in interparent, mother-adolescent, father-adolescent, sibling, and romantic couple conflicts. Based on the cases with complete data (n = 163), path analyses indicated that both direct and indirect relationships exist between these dyadic relationships: (1) Resolution styles utilized in the interparent subsystem were found to have a direct relationship with motheradolescent and father-adolescent resolution and an indirect relationship with sibling and romantic couple resolution. (2) Mother-adolescent and father-adolescent resolution were found to have a direct relationship with sibling and romantic couple resolution. (3) Sibling resolution was found to have a direct relationship with romantic couple resolution for negative resolution styles, but not for the positive resolution style of compromise.

1Assistant

Professor, Department of Psychology, Illinois State University, Normal, Illinois. Received Ph.D. in family science from The Ohio State University. Major research interests are adolescent development and conflict management behaviors. To whom correspondence should be addressed at Department of Psychology, Illinois State University, Campus Box 4620, Normal, Illinois 61790-4620. 2Associate Professor, Department of Human Development and Family Science, Ohio State University, Columbus, Ohio. Received Ph.D. in family science from the University of Connecticut. Major research interest is family distance regulation.

735 0047-2891/98/1200-0735S15.00/0 © 1998 Plenum Publishing Corporation

736

Reese-Weber and Bartle-Haring

INTRODUCTION Conflict is an inevitable event in all interpersonal relationships (Shantz and Hartup, 1992). The management of conflict determines the degree to which a relationship will be influenced by these conflictual episodes. During adolescence interpersonal relationships extend from the family to peer relationships providing opportunities to observe and explore conflict management styles within the family and outside the family (Hartup, 1992). The purpose of the present study is to examine how conflict resolution styles between one family dyad is related to other family dyads and how conflict resolution styles within these family dyads are related to conflict resolution styles in one relationship outside the family—adolescents’ romantic relationships. With families, adolescents can practice resolution styles in relationships with their parents and siblings and observe resolution styles used during interparent disputes. Research indicates that resolution styles experienced and observed within the family are similar across dyads. Rubenstein and Feldman (1993) found three resolution styles in parent-adolescent conflict using the Conflict-Resolution Behavior Questionnaire. Behaviors such as apologizing, understanding the others' point of view, and/or working together to solve the problem have been considered a compromise style of resolution. A resolution style that includes hostility, authoritarian behavior, and/or escalation of the conflict has been labeled attack. Finally, an avoidance style of resolution includes ignoring the problem and/or withdrawing from the conflict. This research indicates that specific resolution styles are utilized by parents and adolescents during conflict. Consistent with the conflict resolution styles found in parent-adolescent conflict, research has indicated that similar resolution styles are used in both sibling and interparent conflicts. Graham-Bermann et al. (1994) found two resolution styles within sibling conflicts. One resolution style, positive problem solving, included behaviors such as discussing the disagreement calmly and gathering information to substantiate their own point of view, which is similar to the compromise style. Negative problem solving included insulting the other, similar to attack, and refusing to talk or leaving the room, which is similar to the avoidance resolution style in parent-adolescent relationships. Also, Katz and Gottman (1993) found two resolution styles in interparent conflict that are similar to those found with parentadolescent and sibling conflict. One style (mutually hostile) included both parents expressing hostility. This style was similar to each person using an attack style of resolution. The second style (demand-withdrawn) included one parent expression hostility (usually the mother), which is similar to the attack style, and one parent distancing themselves from the resolution proc-

Conflict Resolution Styles

737

ess (usually the father), which is similar to the avoidance style. Previous research on family dyads (i.e., parent-adolescent, sibling, and interparent) indicate that conflict resolution styles are not unique to a specific dyadic relationship, but are utilized across family subsystems. Although the resolution styles found in parent-adolescent, sibling, and interparent conflict are very similar, most research on conflict within the family has examined conflict in only one of these family subsystems. A few studies have examined how conflict in two or more family dyads may be related to adolescent outcome variables. Two studies that have investigated conflict in multiple family dyads suggest that the influence of interparent conflict on adolescent adjustment may be mediated through parent-adolescent conflict. For example, Brody and forehand (1993) examined the relationship of interparent conflict and mother-adolescent conflict to adolescents' alcohol and drug use. A direct relationship between interparent conflict and adolescent alcohol was not found, but mother-adolescent conflict did have a direct relationship with adolescents' alcohol use. Further, Cole and McPherson (1993) found that mother-adolescent and father-adolescent conflict mediated the relationship between interparent conflict and adolescents' depression symptoms. However, only father-adolescent conflict was significantly related to adolescent depressive symptoms. Taken together, these two studies suggest that the relationship between interparent conflict, parent-adolescent conflict, and adolescent outcome variables may best be explained by a mediator model with father-adolescent conflict as the predictor of adolescent outcome. Although this mediator model does not include the sibling relationship, previous research has found sibling conflict to be related to parent-adolescent and interparent conflict. Montemayor and Hanson (1985) found that adolescents reported similar resolution styles for both parent-adolescent and sibling conflict, suggesting that not only are conflict resolution styles consistent across studies, but within samples also. Further, Brody et al. (1992) suggest that increases in interparent conflict may have long-term implications for increases in sibling conflict as indicated by the finding that interparent conflict is related to sibling conflict concurrently and one year later. Because sibling conflict is related to both parent-adolescent and interparent conflict, inclusion of sibling conflict may provide additional information about the relationships of conflict styles among multiple family subsystems. In addition to examining the associations between resolution styles in multiple family subsystems, the relationship between resolution styles in family subsystems and resolution styles in adolescents' relationships outside the family has been examined. Romantic relationships were determined an appropriate nonfamily relationship to investigate because previous research

738

Reese-Weber and Bartle-Haring

has suggested that conflict is a significant event in late adolescents' romantic relationships. For example, Furman and Buhrmester (1992) found that although early and middle adolescents rated conflict with family members as occurring more frequently than conflict in relationships outside the family, late adolescents rated conflict with romantic partners as equally frequent as conflict with family members. Given that conflict occurs in these relationships, it is important to understand how conflict resolution styles in the family may impact conflict resolution styles in late adolescents' romantic relationships. Although research has not examined the link between conflict resolution styles in the family and conflict resolution styles within romantic relationships, research has assessed the relationship between aggressive behaviors observed and experienced in the family and aggressive behaviors displayed outside the family system. Prior to reviewing the literature on aggression, it should be stated that conflict and aggression are not identical. Shantz (1987) defined conflict as mutual opposition and aggression as behavior that is intended to harm another person. Given this cautionary note, the aggression literature should be viewed as an example of how behaviors within the family can be related to behaviors in relationships outside the family. Research on aggression has found interparent, parent-adolescent, and sibling aggressive behaviors to be related to aggressive behaviors displayed outside the family. For example, observing interparent aggression, displayed by either mother or father, has been found to be associated with late adolescents' aggressive behaviors towards nonfamily members (Mangold and Koski, 1990). Specific to romantic relationships outside the family, Foo and Margolin (1995) found that observing interparent aggression is related to expressing aggression in late adolescents' dating relationships. In addition, research has linked parent-adolescent aggression to nonfamily aggression suggesting that adolescents who experience physical or verbal aggression from their fathers are more likely to be aggressive toward others (Truscott, 1992). Finally, late adolescents' who were aggressive toward their siblings have been found more likely to be aggressive in nonfamily relationships (Mangold and Koski, 1990). In the literature on domestic violence or violence between partners, the "cycle of violence" is often discussed (i.e., Breslin et al., 1990; Hotaling and Sugarman, 1986). This cycle of violence suggests that those who are witnesses to and/or victims of violence in their families of origin will be more likely to use violence or be victims of violence in relationships outside the family. In reviewing the literature on this issue, however, it is clear that growing up in a violent home is not a sufficient factor to predict perpetrating or receiving violence in intimate adult relationships. In fact, Kauf-

739

Conflict Resolution Styles

man and Zigler (1987) report that only about 30% of those who witness interparental violence or experience violence at the hands of their parents also perpetrate violence as adults. Thus, even though there is a relationship between violence in the family of origin and violence in adult relationships, there may be other mediating influences. Although aggression is not by definition conflict resolution, the aggression literature provides evidence that behaviors in the interparent, mother-adolescent, father-adolescent, and sibling subsystems may impact behaviors in relationships outside the family system. The present study was conducted as an initial investigation to determine if the same links that exist regarding aggressive behaviors also exist regarding conflict resolution behaviors. Thus, the present study examined the relationship of conflict resolution styles in family subsystems to conflict resolution styles within adolescents' romantic relationships. Late adolescents' perceptions of mothers', fathers', siblings', romantic partners', and their own conflict resolution styles within interparent, mother-adolescent, father-adolescent, sibling, and romantic relationships were assessed. Based on the previous research, the first hypothesis was that the relationship between resolution styles in interparent conflict and romantic partner conflict would be mediated by the resolution styles of mother-adolescent and father-adolescent conflict. However, only father-adolescent, not mother-adolescent, resolution styles were hypothesized to be positively related to conflict resolution styles in adolescents' romantic relationships. Second, it was hypothesized that the relationship between resolution styles in interparent conflict and sibling conflict would be mediated through mother-adolescent and father-adolescent conflict. Third, sibling conflict resolution styles were expected to be significantly related to adolescents' resolution styles in romantic relationships. (See Fig. 1 for the general model.) Regarding specific conflict resolutions styles, compromising in family conflicts was hypothesized to be related to compromising in adolescents' romantic relationships. Finally, it was hypothesized that attack and avoidance resolution styles in the family would be related to attack and avoidance resolution styles in adolescents' romantic relationships, respectively. METHOD Sample and Procedures The sample for this study was a sample of convenience. Undergraduate students were given extra class credit for their voluntary participation in this study and assistance in recruitment of participants. Requirements for

740

Reese-Weber and Bartle-Haring

Fig. 1. The full model that will be tested for each conflict resolution style.

participation in this study included having at least one sibling and currently being in an exclusive romantic relationship that had been at least 1 month in duration. Students meeting these requirements completed and returned the study measures during class time. Also, most students volunteering as assistants were female and were thus asked to distribute the questionnaires to individuals of the opposite sex meeting the requirements (to include both males and females). Questionnaires distributed by student assistants were returned by mail. Three hundred questionnaires were distributed and 217 returned. The 217 participants (144 females and 73 males) had a mean age of 21.8 years. The majority of respondents were White (92%), with 82% from intact families. The remaining participants included 9% from stepfamilies, 8% from single families, and 1% from some other family structure. Participants completed measures on the sibling relationship in regards to the sibling closest in age to them. The average number of years between these siblings was 3.4 years, with 56% responding in regard to an older sibling and 44% to a younger sibling. Most siblings were biological siblings (94%) and the remaining were half siblings (3%) or step siblings (3%). Of these 217 participants, there were 163 cases with complete data (113 females and 50 males). Similar to the total sample, the mean age of

741

Conflict Resolution Styles

these participants was 21.9 years. The majority of participants were White (94%) and from intact families (85%). The mean spacing of siblings was 3.4 years with 57% reporting on older siblings and 43% reporting on younger siblings. Most siblings were biological (94%) with the remaining either half-siblings (2%) or stepsiblings (4%). Instruments Conflict-Resolution Behavior Questionnaire This 22-item questionnaire assesses three conflict-resolution behaviors: attack, avoidance, and compromise (Rubenstein and Feldman, 1993). Using a 5-point scale ranging from 1 = never to 5 = almost always, subjects rate how often a person engages in each behavior. The three subscale scores are computed by calculating the mean rating for each subscale. Rubenstein et al. (1993) reported alpha coefficients as .78, .73, and .77 for the attack, avoidance, and compromise scales, respectively. Since only simple revisions were made to this questionnaire to reflect behaviors of both adults and late adolescents (i.e., changing "playing video game" to "read a book" and "talk with brother or sister" to "talk to friend or spouse") the reliability, as expected, was similar to alpha coefficients reported previously. In the present study, alpha coefficients ranged from .81 to .87 for the attack scale, .63 to .79 for the avoidance scale, and .78 to .91 for the compromise scale. Although previous research has obtained parents' and adolescents' perceptions of the adolescents' conflict-resolution behaviors, this study asked adolescents' perceptions of their mothers', fathers', siblings', romantic partners', and their own conflict-resolution behaviors within reciprocal dyadic relationships (mother to father, father to mother, mother to adolescent, adolescent to mother, father to adolescent, adolescent to father, sibling to adolescent, adolescent to sibling, romantic partner to adolescent, and adolescent to romantic partner). A confirmatory factor analysis using structural equation modeling was conducted in order to ensure that the subscale items were significantly loading on their designated subscale. This was done for each dyadic scale (i.e., mother to father, father to mother). Two items did not load significantly on the avoidance subscale and also had high error terms. These items were deleted. Five dyadic scores were created by summing reciprocal scores: (1) Interparent = mother to father and father to mother; (2) Mother-adolescent = mother to adolescent and adolescent to mother; (3) Father-adolescent = father to adolescent and adolescent to father; (4) Sibling = sibling to adolescent and adolescent to sibling; and (5) Romantic couple = ro-

742

Reese-Weber and Bartle-Haring

mantic partner to adolescent and adolescent to romantic partner. In addition, each dyadic score was computed for each resolution style yielding 15 variables. RESULTS Correlations The correlations among conflict resolution styles across and within dyadic relationship are presented in Table I. Across dyadic relationships, the compromise resolution style in one dyadic relationship was positively correlated with compromise in other dyadic relationships (e.g., interparent compromise was related to mother-adolescent, father-adolescent, sibling, and romantic couple compromise). Likewise, attack and avoidance resolution styles in one dyadic relationship were positively correlated with attack and avoidance resolution styles, respectively, in other dyadic relationships. These correlations suggest that if a resolution style is utilized in one dyadic relationship that same resolution style is likely to be utilized in other dyadic relationships. Relationships among resolution styles indicated higher scores on compromise were associated with lower scores on attack and avoidance. These relationships indicated that compromise in one dyadic relationship was related to the lack of both attack and avoidance resolution styles in that same dyad and other dyads. In addition, attack and avoidance resolution styles were positively correlated indicating that higher scores for attack were related to higher scores for avoidance both within dyadic relationships and across dyadic relationships. Path Analyses The path models were estimated using the computer program LISREL VII. The resolution styles for each dyadic relationship were used as observed variables in structural equation modeling procedures. LISREL VII provides maximum likelihood estimates for the models, chi-square statistics, and a Goodness of Fit Index (GFI). Since the chi-square statistic is biased by sample size and model size (see Bollen, 1989), the chi-square to degrees of freedom ratio can also be used as an indicator of fit. A ratio between 1 and 5 has been suggested as an acceptable range for adequate fit (Marsh and Hocevar, 1988). The full models tested, which include all direct and indirect pathways, were exactly identified, thus, no fit indices could be calculated. Therefore,

Conflict Resolution Styles

743

744

Reese-Weber and Bartle-Haring

the full model is compared to the mediator model to determine whether a significant loss of fit occurred. If there is no significant loss of fit when using the mediator model, the mediator model is accepted. LISREL VII provides R2 for each equation and z values to determine the significance of paths. Finally, using structural equation modeling, disturbance terms can be allowed to correlate (this is not possible with more traditional path analysis procedures). In the present study, the disturbance terms for the mother-adolescent and father-adolescent dyadic variables were freed to correlate. These disturbance terms were correlated to account for the fact that more than just interparent conflict resolution may predict parent-adolescent resolution styles, and these other variables may be similar for both mother-adolescent and father-adolescent interaction patterns; thus the disturbance terms would have to be correlated. The disturbance terms may also include error variance, and some of this variance may not be random due to the fact that the subject was asked to complete the same items several times. Again, this would imply that the disturbance terms may be correlated. Figure 2 presents the maximum likelihood results of the full model (A) and mediator model (B) for the compromise resolution style. In the full model, the direct paths between interparent compromise and romantic couple compromise and interparent and sibling compromise were not significant. The chi-square for the mediator model was 1.52 with 2 degrees of freedom (p = .47), the GFI was .97, with a chi-square to degrees of freedom ratio of .78. All these indicate no significant loss of fit when using the mediator model. The z values indicated that all paths were significant in the mediator model, with the exception of sibling compromise to romantic couple compromise, which only approached significance (p = .04). This model suggests that the relationship between compromise resolution style in the interparent subsystem and compromise in the sibling or romantic dyads is mediated through the mother-adolescent and father-adolescent resolution style of compromise. The R2 for romantic couple compromise was .33 and was .35 for sibling compromise, indicating that approximately the same amount of variance was accounted for in the sibling and romantic couple variables. Figure 3 presents the maximum likelihood results of the full model (A) and mediator model (B) for the attack resolution style. The full model indicates that the direct path between interparent and romantic couple and interparent and sibling attack are not significant. The chi-square for the mediator model was 1.35 with 2 degrees of freedom (p = .51), the GFI was .98, with a chi-square to degrees of freedom ratio of .68. All these indicate no significant loss of fit when using the mediator model. The z values indicated that all paths were significant. This model suggests that

Conflict Resolution Styles

Fig. 2. Maximum likelihood path analysis results of the full model (A) and mediational model (B) for the compromise resolution style.

745

746

Reese-Weber and Bartle-Haring

Fig. 3. Maximum likelihood path analysis results of the full model (A) and mediational model (B) for the attack resolution style.

Conflict Resolution Styles

747

the relationship between attack resolution in the interparent and sibling relationships is mediated through the mother-adolescent and father-adolescent dyads. In addition, the relationship between interparent attack resolution and romantic couple attack resolution is mediated through the mother-adolescent and father-adolescent dyads, and also through the sibling relationship via the mother-adolescent and father-adolescent relationships. The R2 for romantic couple attack was .34 and was .37 for sibling attack. Finally, Fig. 4 presents the results of the full model (A) and mediator model (B) for the avoidance resolution style. Similar to the full models for compromise and attack, the direct paths between interparent and romantic couple and interparent and sibling avoidance are not significant. The chisquare for the mediator model was 2.83 with 2 degrees of freedom (p = .24), the GFI was .95, with a chi-square to degrees of freedom ratio of 1.42. All these indicate no significant loss of fit when using the mediator model. The z values indicated that all paths were significant. As with the attack resolution model, this model indicates that the relationship between avoidance resolution in the interparent and sibling relationship is mediated through the mother-adolescent and father-adolescent dyads. In addition, the relationship between interparent avoidance resolution and romantic couple avoidance is mediated through the mother-adolescent and fatheradolescent dyads and also through the sibling relationship via the motheradolescent and father-adolescent relationships. The R2 for romantic couple avoidance was .34 and .36 for sibling avoidance. DISCUSSION The present study examined two important questions: (1) What are the direct and indirect relationships between conflict resolution styles within family subsystems? (2) How are conflict resolution styles within the family related to conflict resolution styles in late adolescents’ romantic relationships? Regarding the first question of inquiry, the findings indicated that a direct relationship exists between interparent and both mother-adolescent and father adolescent conflict resolution styles. However, an indirect relationship existed between resolution styles in the interparent and sibling dyads with mother-adolescent and father-adolescent resolution styles as the mediating variables. These relationships were found for each resolution style, suggesting that this pattern existed for the process of resolution rather than a specific resolution style. One explanation for why interparent conflict had a direct relationship with parent-adolescent conflict may be the inclusion of the parent in both

748

Reese-Weber and Bartle-Haring

Fig. 4. Maximum likelihood path analysis results of the full model (A) and mediational model (B) for the avoidance resolution style.

dyadic relationships. Parents who used an attack resolution style in interparent disputes may have also used an attack resolution style in conflicts with their adolescents. Regarding how the adolescents' resolution style con-

Conflict Resolution Styles

749

tributes to parent-adolescent conflict, social learning theory would hypothesize that adolescents observe their parents’ attempts to resolve conflict in the interparent relationship and imitate those behaviors in their own conflicts (Bandura, 1989). Thus, both the parent and the adolescent may be utilizing resolution styles that are similar to those used to manage conflict in the interparent subsystem. Systems theory may be helpful in understanding the direct relationship found between parent-adolescent and sibling resolution styles. The systemic perspectives posit that what occurs in one family subsystem is likely to influence what occurs in other family systems (von Bertalanffy, 1968). In addition, family subsystems are hierarchically arranged, suggesting that the subsystems including the parents may have more influence over subsystems not including the parents. Parent-adolescent relationships would have more influence over sibling relationships, rather than the reverse, because parents are included in the first, but not the latter. Thus, the present results could be interpreted as supporting a systemic perspective of the family. Regarding how resolution styles in the family are related to resolution styles in late adolescents’ romantic relationships, the present study found that conflict resolution in the sibling relationship was significantly related to conflict resolution in the romantic relationships for the attack and avoidance styles, but only approached significance for the compromise style. These differences across resolution styles suggest that the influence of the sibling dyad on the romantic dyad may only exist for specific resolution styles rather than the entire process of resolution. For example, sibling resolution behaviors that have previously been labeled negative problem-solvin (Graham-Bermann et al., 1994), attack, and avoidance styles were found in the present study to be significantly related to those same resolution styles in the romantic relationship. In contrast, sibling resolution behaviors that have been labeled positive problem-solving (Graham-Bermann et al., 1994), or a compromise style, were not significantly related to resolution styles in the romantic relationship in the present study. It may be the case that these more negative resolution styles are “easier” or at least do not require as much time and patience, and thus tend to be more automatic responses, while compromise requires more time and must be less automatic. In addition to the relationship between sibling resolution styles in late adolescents’ romantic relationships, the present study found that the relationship between interparent and romantic resolution styles was mediated through the mother-adolescent and father-adolescent resolution styles. It is interesting to note that in many families, parents prefer not to argue in front of their children, assuming that this spares the children from becoming anxious, and perhaps hoping that these conflicts are not replicated.

750

Reese-Weber and Bartle-Haring

From the results of this study, it is not the conflicts between the parents that appear to influence how the offspring behave later on in their own outside relationships, but the conflicts between the parent and the child. The mediating influence of the parent-adolescent conflict resolution styles and the sibling adolescent conflict resolution style have implications for the study of more aggressive conflicts or violence. It appears that regardless of what parents do to each other, the more important patterns of behavior are those that the adolescents directly experience. So, if parents and adolescent use an attacking, avoiding, or compromising conflict resolution style, the adolescent is more likely to use that style in romantic relationships. If the adolescent and sibling use attack and avoidance, then the adolescent is more likely to use that style in romantic relationships. DeMaris (1987) and Marshall and Rose (1988) have demonstrated that individuals who received severe physical punishment as a child were significantly more likely to inflict violence on a dating partner. This was also related to sustaining violence for males. Thus the cycle of violence may be more likely to be perpetuated if the child is directly involved in the violence or aggressive behavior than when a witness only. It is also apparent from the results of this study that for more positive conflict behaviors to be perpetuated, the child needs to be involved in these sorts of interchanges with parents. Thus, it would appear from these results that when a child experiences more negative conflict resolution behaviors with a parent, but positive resolution with a sibling, the negative resolution style will be perpetuated in relationships outside the home. More research in this area is needed to explore this possibility. With larger samples, all three resolution styles could be used in a model to predict the resolution styles used in romantic relationships. This would provide information about what combinations of resolution styles in the parent-adolescent and sibling-adolescent relationships are predictive of the preferred resolution style in romantic relationships. Although the relationship between father-adolescent resolution styles and romantic couple resolution styles was expected, the significant relationship found in the present study between mother-adolescent and romantic couple resolution styles was not expected. Because previous research has not examined the relationship between family and nonfamily conflict resolution styles, the expected results were based on previous research examining related topics such as male adolescents’ aggressive behaviors (Truscott, 1992) and nonspecific conflict behaviors (Cole and McPherson, 1993). Previous research on these related topics has indicated that the father-adolescent, but not the mother-adolescent dyad, has a significant relationship with outcome variables. Truscott (1992) only included male adolescents, whereas the present study included both male and female adolescents. The possibility exists that male adolescents may be more likely to

751

Conflict Resolution Styles

imitate their fathers’ behaviors, rather than their mothers’ behaviors. Cole and McPherson (1993) did assess both male and female adolescents, but assessed conflict in general rather than the specific aspect of resolution. In addition, Cole and McPherson (1993) assessed the relationship between conflict and depressive symptoms, rather than the relationship between conflict behaviors in one dyad and conflict behaviors in another dyad. Therefore, the findings in the present study may be different from previous research due to differences in samples and variables examined. One strength of the present study was the inclusion of multiple family dyadic relationships. Adolescents provided information on the behaviors of each person within several dyadic relationships. However, because adolescents completed the measure for several dyadic relationships, rater bias or common methods variance most likely influenced the results. In other words, a portion of the variance in scores may be due to the idiosyncratic manner in which an individual responds to the measures (Sabatelli and Bartle, 1995). In addition, mothers, fathers, siblings, and romantic partners may perceive the resolution styles utilized in their relationships differently. Given these limitations, future research should include multiple perspectives when assessing dyadic relationship qualities. In addition to assessing multiple perspectives, future research in this area of inquiry should include examining the links between conflict resolution styles within diverse families. The findings of the present study were based on white, intact families, and should not be generalized to other populations because differences in family structure and/or ethnic group may alter the pattern found in the present study. For example, the relationship between conflict resolution styles within a single-parent family might exclude the interparent dyad or the dyadic relationship between the nonresidential parent and the adolescent. Likewise, the model found in the present study may not include important dyadic relationships for other ethnic groups (e.g., extended family). The present study does contribute significantly to the literature. First, as noted above, multiple family subsystems were assessed, rather than only one or two family subsystems. Second, the present study was the first attempt to link family and nonfamily resolution styles. In conclusion future research is needed that replicates the present study and provides more knowledge on how conflict resolution styles are related within families and to late adolescents’ resolution styles in romantic relationships.

REFERENCES Bandura, A. (1989). Social cognitive theory. Ann. Child Develop. 6: 1-60.

752

Reese-Weber and Bartle-Haring

Bollen, K. A. (1989). Structural Equations with Latent Variables. John Wiley & Sons, New York. Breslin, F. C., Riggs, D. S., O’Leary, K. D., and Aries, I. (1990). Family precursors: Expected and actual consequences of dating aggression. J. Intetpers. Vio. 5: 247-258. Brody, G. H., and Forehand, R. (1993). Prospective associations among family form, family processes, and adolescents’ alcohol and drug use. Behav. Res. Ther. 31: 587-593. Brody, G., Stoneman, Z., McCoy, K., and Forehand, R. (1992). Contemporaneous and longitudinal associations of sibling conflict with family relationship assessment and family discussions about sibling problems. Child Develop. 63: 391-400. Cole, D. A., and McPherson, A. E. (1993). Relation of family subsystems to adolescent depression: Implementing a new family assessment strategy. J. Family Psychol. 7: 119-133. DeMaris, A. (1987). The efficacy of a spouse abuse model in accounting for courtship violence. J. Family Issues 8: 291-305. Foo, L., and Margolin, G. (1995). A multivariate investigation of dating aggression. /. Family Viol. 10: 351-377. Furman, W., and Buhrmester, D. (1992). Age and sex differences in perceptions of networks of personal relationships. Child Develop. 63: 103-115. Graham-Bermann, S. A., Cutler, S. E., Litzenberger, B. W., and Schwartz, W. E. (1994). Perceived conflict and violence in childhood sibling relationships and later emotional adjustment. J. Family Psychol. 8: 85-97. Hartup, W. W. (1993). Conflict and friendship relations. In Shantz, C. U., and Hartup, W. W. (eds.), Conflict in Child and Adolescent Development. Cambridge University Press, New York. Hotaling, G., and Sugarman, D. (1986). An analysis of risk markers in husband to wife violence: The current state of knowledge. Viol Victims 1: 101-124. Katz, L. F., and Gottman, J. M. (1993). Patterns of marital conflict predict childrens’ internalizing and externalizing behaviors. Develop. Psychol. 29: 940-950. Kaufman, J., and Zigler, E. (1987). Do abused children become abusive parents? American J. Orthopsychiat. 57: 186-192. Mangold, W. D., and Koski, P. R. (1990). Gender comparisons in the relationship between parental and sibling violence and nonfamily violence. J. Family Violence 5: 225-235. Marshall, L. L., and Rose, R. (1988). Family of origin violence and courtship abuse. J. Counsel. Develop. 66: 414-418. Marsh, H. W., and Hocevar, D. (1988). A new, more powerful approach to multitrait-multimethod analyses: Application of second-order confirmatory factor analysis. J. Appl. Psychol. 73: 107-117. Montemayor, R., and Hanson, E. A. (1985). A naturalistic view of conflict between adolescents and their parents and siblings. J. Early Adoles. 3: 83-103. Rubenstein, J. L., and Feldman, S. S. (1993). Conflict-resolution behavior in adolescent boys: Antecedents and adaptational correlates. J. Res. Adolesc. 3: 41-66. Sabatelli, R. M., and Bartle, S. E. (1995). Survey approaches to the assessment of family functioning: Conceptual, operational, and analytical issues. J. Marriage Family 57: 10251039. Shantz, C. U. (1987). Conflicts between children. Child Development 58: 238-305. Shantz, C. U., and Hartup, W. W. (1992). Conflict in Child and Adolescent Development. Cambridge University Press, New York. Truscott, D. (1992). Intergenerational transmission of violent behavior in adolescent males. Aggres. Behav. 18: 327-335. von Bertalanffy, L. (1968). General System Theory. George Braziller, New York.