|Home | About | Journals | Submit | Contact Us | Français|
This study examined associations between interparental conflict and adolescents’ romantic relationship conflict. High school seniors (N=183) who lived with married parents completed questionnaires about their parents’ marriage and their own romantic relationships. A subset of 88 adolescents was also observed interacting with their romantic partners. Adolescents’ perceptions and appraisals of interparental conflict were related to the amount of conflict in romantic relationship and adolescents’ conflict styles. Adolescents’ appraisals of interparental conflict (i.e., self-blame, perceived threat) moderated many of the associations between interparental conflict and conflict behavior with romantic partners. The patterns of moderated effects differed by gender. These findings suggest that the meanings boys and girls ascribe to interparental conflict are important for understanding how family experiences contribute to the development of romantic relationships.
A small but growing body of research suggests that exposure to high levels of interparental conflict could interfere with youths’ social development. Widely accepted as a risk factor for psychological maladjustment, interparental conflict has more recently been linked to difficulties in relationships with siblings, peers, and romantic partners (Kinsfogel & Grych 2004; Parke et al., 2001; Steinberg, Davilla, & Fincham, 2006; Stocker & Richmond, 2007). The current study builds on this literature by examining examine how interparental conflict is associated with the incidence and management of conflict in adolescents’ romantic relationships.
Learning to form and sustain romantic relationships is a key developmental task of adolescence, and building effective conflict skills is fundamental to this task (Shulman, Tuval-Mashiach, Levran, & Anbar, 2006). Unlike family relationships, where conflicts are frequently resolved through submission or escalation, romantic relationships require adolescents to manage differences by balancing their own needs with those of the partner and the relationship (Laursen, Finkelstein, & Betts, 2001). The voluntary, egalitarian nature of romantic relationships motivates most adolescents to negotiate mutually agreeable solutions to disagreements. Negotiation maximizes the potential for relationship preservation and can even strengthen relationships (Hartup, 1992). In contrast, destructive conflict tactics can be hurtful to relationships and individuals, with approximately 35–40% of adolescents experiencing physical or emotional abuse from romantic partners (Wolfe et al., 2001).
Studies of adult romantic relationships indicate that exposure to interparental conflict could shape the development of romantic relationship conflict skills. Interparental conflict is a significant predictor of subsequent distress and conflict in offspring’s marital relationships (Amato & Booth, 2001; Conger, Cui, Bryant, & Elder, 2000). Adults whose parents use verbal and physical aggression are more likely to use similar conflict styles with their own romantic partners (Ehrensaft et al., 2003; Martin, 1990; Reese-Weber & Bartle-Haring, 1998). Although less is known about the significance of interparental conflict for adolescents’ romantic relationships, marital interactions could be particularly salient as adolescents are learning to establish and maintain romantic relationships (Gray & Steinberg, 1999). Parents’ marriage is often the romantic relationship to which adolescents are most immediately and intimately connected. Despite some differences, both dating and marital relationships are voluntary, egalitarian relationships that involve emotional and physical intimacy. Lessons learned from observing parents’ marital interactions might be informative to adolescents as they seek to build more intimate and reciprocal romantic relationships (Simon, Bouchey, & Furman, 2000).
Adolescents exposed to high levels of interparental conflict might anticipate and potentiate conflict with romantic partners. Interparental conflict appears to sensitize youth to interpersonal conflict and foster hypervigilance to conflict cues (Davies, Myers, Cummings, & Heindel, 1999, Fosco, DeBoard, & Grych, 2007; Grych & Fincham, 1990). This sensitivity could generalize to romantic relationships and manifest as expectations for conflict and aggressive responding to minor or ambiguous signs of conflict (Crick & Dodge 1994). In support of this idea, adolescents exposed to marital violence have more conflictual romantic relationships, more often use aggressive conflict styles, and more often view dating aggression as justifiable (Kinsfogel & Grych, 2004; Linder & Collins, 2005; Reitzel-Jaffe & Wolfe, 2001). Yet even intense interparental conflict is not necessarily aggressive, and non-aggressive marital conflict could also predispose adolescents to destructive conflict behavior. In fact, exposure to interparental hostility predicts adolescents’ use of verbal attack and mild psychological aggression with romantic partners (Stocker & Richmond, 2007). Likewise, interparental conflict could predispose adolescents to non-aggressive, destructive conflict tactics. For instance, adolescents exposed to interparental conflict may seek to minimize conflict altogether by avoiding negative expressions, yielding to partners’ preferences, or expressing positive affect when conflict cues are perceived. Although effective for reducing conflict, such strategies are likely to be ineffective for resolving inevitable disagreements and could undermine the development of intimacy (Gottman & Krokoff, 1989; Simon, Kobielski, & Martin, 2008).
These findings suggest that interparental conflict could sensitize adolescents in ways that increase both the potential for romantic relationship conflict and adolescents’ use of various destructive conflict styles. The current study examines a range of romantic relationship conflict styles, including physical aggression, conflict engagement, compliance, withdrawal, and positive problem solving. We expected that greater exposure to interparental conflict would be associated with more conflictual romantic relationships, less use of problem solving, and greater use of destructive conflict styles with romantic partners.
Understanding the effects of interparental conflict requires attention to the meanings youth ascribe to it. According to the cognitive-contextual framework, youth subjectively appraise the meaning of interparental conflict for their personal well-being (Grych & Fincham, 1990). Negative appraisals, including perceived responsibility for parents’ conflict (self-blame) or beliefs that conflict will escalate and result in harm (threat), predict poor psychological functioning (Cummings & Davies, 2002; Gerard, Buehler, Franck, & Anderson, 2005; Grych, Harold, & Miles, 2003). If such appraisals generalized to romantic relationships, they might also predict romantic relationship functioning. Although the generalization of interparental conflict appraisals to romantic relationships has yet to be demonstrated, research by El Sheikh and Harger (2001) demonstrates that generalization is indeed possible. In their study, youths made similar appraisals about conflict between unfamiliar adults as for conflict between their own parents. To the extent that negative conflict appraisals are generalized, they could sensitize adolescents to actual or ambiguous conflict cues with romantic partners and negatively bias their appraisals of situations or partners (Fosco, et al., 2007). Distorted social information processing during interactions with romantic partners could then increase the potential for conflict and ineffectual conflict management (Crick & Dodge, 1994). Accordingly, we expected that negative appraisals of interparental conflict would be associated with more conflictual romantic relationships, less use of positive problem solving, and greater use of ineffectual conflict styles.
Debate exists as to whether appraisals mediate or moderate the associations between interparental conflict and youth outcomes. In the mediation model, appraisals are conceptualized as direct responses to interparental conflict that explain how and why interparental conflict leads to maladaptive outcomes. Although several studies support a mediation model (Gerard et al., 2005; Grych, Fincham, Jouriles, & McDonald, 2000; Grych et al., 2003), others indicate that appraisals moderate associations between interparental conflict and youth outcomes (El-Sheikh & Hager, 2001; El-Sheikh, Harger, & Whitson, 2001; Kerig, 1998; Rogers & Holmbeck, 1997; Rossman & Rosenberg, 1992). Unlike mediators, moderators do not speak to issues of causality. Instead they explicate for whom and under what conditions interparental conflict is associated with negative outcomes. Significant moderated effects would signify that the impact of interparental conflict on adolescents’ conflict in romantic relationships varies depending on the nature of adolescents’ conflict appraisals.
We hypothesized that a moderation model would better suit our sample of late adolescents. Whereas children's appraisals could reflect situational coping to interparental conflict, adolescents’ appraisals could reflect dispositional coping (Kerig, 2001). In childhood, when cognitive and coping resources are limited, negative appraisals could be normative, direct responses to intense interparental conflict (Fosco, et al., 2007). During adolescence, advances in abstract thinking, perspective-taking, and coping skills enable youth to make more differentiated appraisals. Over time, these skills are refined in ways that allow many adolescents to understand better that they are not to blame for parents’ conflict and to regulate emotional responses more effectively. By late adolescence, individual differences in conflict appraisals could reflect stable differences in exposure and responses to interparental conflict. Such dispositional characteristics are believed to moderate links between interparental conflict and youths’ outcomes (Davies and Windle, 2001; Grych & Fincham, 1990).
In support of this idea, data by Wadsworth and colleagues indicate developmental differences in the function of coping responses in adjustment to stress (Wadsworth, Raviv, Compas, & Conner-Smith, 2005). Coping responses mediated links between stress and adjustment through early adolescence, and later became more stable responses that moderated stress-related outcomes. To our knowledge, no one has assessed whether conflict appraisals serve as moderators or mediators in late adolescence. Based on the literature reviewed, we hypothesized that appraisals of interparental conflict made by late adolescents would reflect dispositional responses that moderate associations between perceptions of interparental conflict and conflict in romantic relationships. Associations between interparental and adolescents’ romantic relationship conflict were expected to be stronger for adolescents who made more negative appraisals of interparental conflict than for those with more benign evaluations.
Davies and Lindsay (2004) have argued that differences in gender socialization could lead boys and girls to react differently to interparental conflict. Whereas boys’ socialization generally emphasizes agency, assertion, and individual well-being, girls’ socialization typically stresses interpersonal connectedness (see Ruble & Martin, 1998). Gender roles become increasingly differentiated during adolescence and could contribute to differential reactivity to interparental conflict (Davies & Lindsay, 2004). Boys’ concern with agency and autonomy could lead them to focus on their own needs during conflict with romantic partners, and those who make negative conflict appraisals might be more likely to view power assertive conflict styles, such as verbal attack or physical aggression, as acceptable ways to achieve their goals (Dadds et al., 1999; Simon & Martin, 2006). In support of this idea, interparental aggression appears to be a better predictor of dating aggression for boys than girls (Foo & Margolin, 1995; Kingsfogel & Grych, 2004). Alternatively, withdrawal may be an attractive means for boys to maintain their personal well-being and regulate angry arousal or aggressive impulses (Kurdek, 1995; Markman, Silvern, Clements, & Kraft-Hanak, 1993). For girls, concerns for interpersonal connectedness could heighten sensitivity to the negative outcomes of conflict and lead them to focus on partners’ needs or the relationship itself (Davies & Lindsay, 2004). Girls who make negative conflict appraisals of interparental conflict could be more likely to view compliance as an acceptable way of achieving these goals (Dadds et al., 1999). Accordingly, we expected that interparental conflict and negative conflict appraisals would be associated with conflict engagement, physical aggression and withdrawal among boys and with compliance among girls.
The current study is among the first to examine how adolescents’ perceptions and appraisals of interparental conflict are associated with conflict styles in romantic relationships using observational and self-report measures. The central hypotheses of this study were: 1) adolescents’ perceptions and appraisals of interparental conflict would each be associated with romantic relationships characterized by greater conflict, less positive problem-solving, and more destructive conflict behavior; 2) adolescents’ appraisals of interparental conflict would moderate these associations; and 3) the pattern of moderated effects would vary by gender.
Participants were 183 high school seniors ranging in age from 17 to 19 years old (M = 17.6 years) who participated in a larger study on adolescent romantic relationships (N = 260). The 98 girls and 85 boys included in the study represent the subset of adolescents who were living in a household with two parents who had been married at least two years (M = 19.53 years; range = 2 – 40 years). Most (88%) lived with two natural (i.e., biological or adoptive) parents; 12% lived with a natural parent and a stepparent. The sample was ethnically and socioeconomically diverse, with 68% Caucasian, 13% African American, 12% Hispanic, and 4% Asian. All identified as heterosexual. All reported having a romantic relationship lasting at least one month, and 85% reported at least one romantic relationship lasting at least three months. Seventy-four percent of the participants were in a relationship at the time of their enrollment in the study.
Participants were recruited from two school districts of a large Western metropolitan city through advertisements in school newspapers and letters sent to families of high school seniors in these schools. Participating adolescents came to the laboratory for a series of three interviews about close relationships and completed two sets of questionnaires. Those who were in a romantic relationship of at least 6 months were observed in a series of discussions with their partner during a fourth session (n=88). Participants were paid $60 for completing the interviews and questionnaires, and an additional $20 if they also completed the observation session.
Adolescents’ perceptions of interparental conflict were assessed using the Conflict Properties scale of the Children's Perceptions of Interparental Conflict Scale (CPIC; Grych, Seid, & Fincham, 1992). The scale includes 19 items assessing the frequency, intensity, and resolution of interparental conflict (e.g., “I often see or hear my parents arguing.”). Items were rated on a seven-point Likert scale with higher scores reflecting higher levels of interparental conflict. The scale is significantly correlated with parental reports of interparental conflict and reliable in adolescent samples (Bickham & Fiese, 1997; Grych, Raynor, & Fosco, 2004). Cronbach’s alpha for the current sample was .91.
The Threat and Self-Blame scales from the CPIC were used to assess adolescents’ appraisals of interparental conflict. The Threat scale includes 12 items that tap beliefs that conflict will escalate and result in harm and perceived ability to cope (e.g., “When my parents argue, I’m afraid something bad will happen.”). The Blame scale includes 9 items that assess self-blame for marital disputes (e.g., “It is usually my fault when my parents argue.”). The two scales are valid and reliable measures of youths’ subjective evaluations of conflict (Bickham & Fiese, 1997; Grych et al., 1992). In the current sample, Self-Blame and Threat were highly correlated (r = .68), so they were averaged to create a single scale for negative appraisals of interparental conflict (Cronbach’s alpha = .85).
The amount of conflict in adolescents’ romantic relationships was assessed using the Negative Interactions factor score from the Network of Relationships Inventory (NRI; Furman & Buhrmester, 1985), which consists of two 3-item scales that measure the degree of conflict and antagonism in romantic relationships. Mean scores from the two scales were averaged to create the factor score, which ranged from 1 to 5 with higher scores reflecting more negative interactions (Cronbach’s alpha = .87).
Conflict styles were assessed with the Conflict Resolution Style Inventory (CRSI; Kurdek, 1994), which contains 16 items tapping the frequency of four resolution styles: 1) conflict engagement (e.g., “throwing insults and digs”), 2) compliance (e.g., “giving in with little attempt to present my side of the issue”) 3) withdrawal (e.g., “reaching a limit, shutting down, and refusing to talk any further”), and 4) positive problem-solving (e.g., “focusing on the problem at hand”). Each style has been identified through behavioral observations and linked to relationship quality (Gottman, 1994). Scores range from 1 to 7; higher scores indicate more frequent use. In the current sample, Cronbach’s alpha ranged from .70 for compliance to .88 for conflict engagement (M alpha =. 83).
Because interparental aggression is associated with adolescents’ dating aggression (Kinsfogel & Grych 2004), four items from the Conflict Tactics Scale (Straus, 1979) were added to the CRSI to assess adolescents’ physical aggression with romantic partners (“Forcefully pushing or shoving”, “Slapping or hitting”, “Throwing items that could hurt”, and “Kicking, biting, or hair pulling”). Cronbach’s alpha for the physical aggression scale in the current sample was 88.
Adolescents in a current romantic relationship of at least six months were videotaped in a series of discussions with their romantic partner (n = 88). First, partners were separated and asked to generate lists describing current problems outside of their romantic relationship (e.g., “conflict with parents about curfew”), current problems within their relationship (e.g., “partner spends too much time with friends”), and personal goals (e.g., “find a job“). Items were rated for importance. After a warm-up task, couples engaged in five six-minute discussions on the following topics: 1) a problem the participant had outside of the relationship, 2) a problem the partner had outside of the relationship, 3) a goal of the participant, 4) a goal of the partner, and 5) a romantic relationship problem. Couples were asked to discuss the topic they rated as most important. For the relationship problem discussion, the topic rated as most important problem was chosen, irrespective of whether the participant or partner named the problem. The observation ended with a wrap-up discussion of fun times in the relationship.
The five discussions were coded using a version of the Interactional Dimensions Coding System (IDCS; Julien, Markman, Lindahl, Johnson, & Van Widenfelt, 1987) that was adapted for adolescent couples. The warm-up and wrap-up segments were not coded. To minimize halo effects, each segment was coded at a different time. One rater coded all discussions, and a second coder rated a subset (18%) for reliability purposes. Both had been trained to reliability on the original ICDS coding system. Interrater reliabilities all exceeded .78 (mean ICC =.85).
The adolescent version of the IDCS is a global coding system tapping fourteen dimensions of behavior, affect, and relationship quality. The nine dimensions capturing each partner’s behavior and affect were used in the current study (communication skills, problem-solving, denial, withdrawal, support/validation, positive affect, negative affect, conflict behavior, and dominance). For each discussion, raters assigned scores to each person on the nine individual dimensions described above using a nine-point scale with higher scores indicating greater frequency and intensity. For purposes of data reduction, scale scores were calculated by averaging a participant’s scores for the nine individual scales across the five discussion tasks. Internal consistencies for averaged scale scores ranged from .82-.88 (M Cronbach’s alpha = .85).
Results from a principal axis factor analysis using oblique rotation on the nine individual scale scores yielded three factors: a) Communication: communication skills, problem-solving skills, denial (negative loading), and withdrawal (negative loading) scales; b) Conflict: conflict behavior and dominance scales; and c) Positive Affective Expression: support validation, positive affect, and negative affect (negative loading). All scales loaded .50 or greater on their respective primary factors except negative affect (.47); all secondary loadings were lower. The scales with primary loadings on a given factor were averaged to create factor scores.
Associations between demographic variables and the dependent variables were examined. None of the demographic variables describing parents’ marriage (i.e., parents’ age, length of parents’ marriage, or step versus biological parent couple), ethnicity, or family SES were significantly correlated with any of the dependent variables (all ps>.05). Regression analyses indicated that neither adolescent ethnicity nor family SES moderated any of the associations between measures of interparental conflict and adolescents’ romantic relationships.
The 183 participants in this study were compared with the 77 participants from the larger study who were excluded because they did not live with a parent who was married. The two groups did not differ in their romantic relationship conflict or on demographic variables. Similarly, participants who did and did not complete the observation session did not significantly differ on any of the primary study variables or demographic variables.
Means and standard deviations for each measure are presented in Table 1 for the entire sample as well as by gender. Girls reported more interparental conflict than did boys, though the two did not differ in their appraisals. The only significant gender difference in romantic relationship conflict was that girls scored higher than boys in their use of conflict engagement. Correlations among the study variables are presented by gender in Table 2. Significant relations between self-reported and observed conflict frequency offer evidence for the validity of adolescents’ reports of conflict in romantic relationships. Each index was related to greater use of destructive conflict styles. Greater perceived interparental conflict was associated with more negative conflict appraisals for boys and girls. As expected, the patterns of associations among interparental conflict, conflict appraisals, and adolescents’ romantic relationships varied by gender. Hence, gender was examined as a moderator in subsequent multivariate analyses.
The central hypotheses of this study were: 1) adolescents’ perceptions and appraisals of interparental conflict would each be associated with romantic relationships characterized by greater conflict, less positive problem-solving, and more destructive conflict behavior; 2) adolescents’ appraisals of interparental conflict would moderate these associations; and 3) the pattern of moderated effects would vary by gender. To test these hypotheses, we conducted a series of regression analyses that examined the independent and interactive effects of perceived interparental conflict, appraisals of interparental conflict, and gender on adolescents’ conflict in romantic relationships (Baron & Kenny, 1986; Holmbeck, 2002). Prior to the analyses, all predictors were centered to reduce multicollinearity (Aiken & West, 1991). Main effects were entered in the first step, followed by two-way interactions in the second step and the three-way interaction in the third step. Dependent variables included reports of the amount of romantic relationship conflict and the five self-reported conflict styles (i.e., conflict engagement, aggression, compliance, withdrawal, and positive problem-solving) as well as the three observed behavior factors (i.e., communication skills, conflict behavior, and positive affective expression). All significant two and three-way interactions were probed following procedures in which slope estimates are calculated and examined at high (M + 1 SD) and low (M − 1 SD) levels of the moderator (Aiken & West, 1991; Holmbeck, 2002). The post-hoc analyses provided information about how associations between perceived interparental conflict and romantic relationship conflict varied for those who made high versus low negative conflict appraisals and for boys versus girls. Parallel analyses were run using the Self-Blame and Perceived Threat scales in place of the combined negative appraisals scores. The results for the separate analyses were nearly identical, and thus, the findings for the combined appraisal index are presented.
Results of regressions predicting adolescent-reported conflict and conflict styles are presented in Table 3. When predicting the amount of conflict in romantic relationships, only perceived interparental conflict was associated with greater conflict in romantic relationships. Neither main nor moderated effects were found for conflict appraisals or gender.
In contrast, the relation between interparental conflict and adolescents’ conflict styles in romantic relationships largely depended on their appraisals of that conflict. Conflict appraisals and gender exerted both main and moderated effects on adolescents’ conflict styles in romantic relationships. When predicting physical aggression, a significant main effect for conflict appraisals was qualified by a significant 3-way interaction between interparental conflict, appraisals, and gender. Results from the post-hoc tests (Figure 1) indicate that perceived interparental conflict was positively associated with physical aggression for boys who were high on negative conflict appraisals (b = .58, p = .05), but not for boys who were low on negative conflict appraisals (b = −.03, p = .79). Perceived interparental conflict was not associated with girls’ use of physical aggression with romantic partners regardless of whether girls’ were high or low on negative conflict appraisals (b = −.04, p = .77 and b = .18, p = .19, respectively).
A significant main effect for conflict appraisals was also qualified by a significant 3-way interaction among interparental conflict, appraisals, and gender when predicting conflict engagement (see Table 3). Yet the results of post-hoc tests show a slightly different pattern of gendered effects than for physical aggression (Figure 2). As with physical aggression, interparental conflict was associated with greater use of conflict engagement when boys were high on negative conflict appraisals (b = .87, p = .006) but not when they were low on negative conflict appraisals (b = .16, p = .41). Among girls, however, the association between interparental conflict and conflict engagement was marginally significant and negative when conflict appraisals were high (b = −.35, p = .07) but not significant when negative appraisals were low (b = .06, p = .73). Thus, for girls who were high on negative conflict appraisals, perceived interparental conflict was associated with less use of conflict engagement.
In the regression predicting conflict withdrawal, there were no main effects, but the interaction between conflict appraisals and gender was marginally significant (see Table 3). Results of post-hoc tests indicated that negative appraisals were associated with greater withdrawal for boys (b = .48, p = .001) but not girls (b = .02, p = .88). Neither interparental conflict nor appraisals were significant predictors of adolescents’ use of positive problem solving, and these factors did not interact with each other or with gender. When predicting compliance, there was a significant interaction between interparental conflict and negative conflict appraisals (see Table 3). Results of post-hoc tests indicated that interparental conflict was associated with greater use of compliance only for adolescents who were high on negative conflict appraisals (b = .38, p = .01 versus b = −.09, p = .51).
Table 4 presents the results of analyses predicting the quality of observed interactions for the subset of 88 adolescents who participated in the observation session with their romantic partner. In the regression predicting conflict behavior, the interaction between conflict appraisals and gender was significant. Results of post hoc tests indicated that more negative conflict appraisals were associated with more conflictual interactions for boys (b = .49, p = .002) but not for girls (b = −.06, p = .79). When predicting positive affective expression, the interaction between conflict appraisals and gender again emerged as a significant predictor. Results of post-hoc tests indicated that for boys, more negative appraisals were associated with less positive affective expression (b = −.37, p = .05), whereas for girls, more negative appraisals were associated with more positive expression (b = .34, p = .03). There were no significant predictors for adolescents’ observed communication.
The results of this study build on a small but growing body of research indicating that interparental conflict is associated with adolescents’ romantic relationship conflict. Consistent with previous reports, interparental conflict had a direct effect on romantic relationship conflict such that more interparental conflict was associated with more conflict in adolescents’ romantic relationships (Kingsfogel & Grych, 2004; Stocker & Richmond, 2007). Adolescents exposed to high levels of interparental may develop similar expectations for their romantic relationships (Crick & Dodge, 1994; Fosco et al., 2007). Anticipating romantic relationships to be conflictual may, in turn, foster heightened attention to mild or ambiguous conflict cues and negative expectancies about conflict resolution (Crick & Dodge, 1994; Simon et al., 2008).
Whereas interparental conflict was directly associated with amount of romantic relationship conflict, its relation to adolescents’ conflict behavior with romantic partners was largely contingent upon adolescents’ conflict appraisals. In fact, all direct effects for interparental conflict and conflict appraisals on adolescents’ romantic conflict styles were qualified by interactive effects. Greater interparental conflict was associated with greater use of destructive conflict styles only for adolescents who made more negative appraisals of interparental conflict. To our knowledge, these findings are the first to identify appraisals of interparental conflict as moderators of youths’ conflict behavior with peers. The results support the idea that adolescents actively interpret and respond to parents’ marital interactions in ways that shape their behavior in other relational contexts (Cummings & Davies, 2002). They suggest that experiences of threat and self-blame in connection with interparental conflict could function as a vulnerability in romantic relationships for adolescents exposed to high levels of interparental conflict. Consistent with research by El-Sheikh and Harger (2001), when youth are exposed to high levels of marital conflict, they may develop ways of appraising conflict that generalize to other relational contexts. Patterns of self-blaming and threatening conflict appraisals that develop for some adolescents could bias their expectations, attention, and processing of romantic interactions in ways that increase the potential for destructive conflict (Crick & Dodge, 1994).
The idea that appraisals of interparental conflict could generalize to other relational contexts is consistent with research suggesting that appraisals become stable dispositions by late adolescence (Kerig, 2001; Wadsworth et al., 2004). The mechanisms by which interparental conflict appraisals might generalize and whether this process is specific to romantic relationships or also includes other close peer relationships (e.g., same-sex and other-sex friendships) are important issues for future research.
As hypothesized, many of the moderated effects were gender specific. The pattern of findings fits largely with the differential response hypothesis proposed by Davies and Lindsay (2004). Among boys who made negative conflict appraisals, interparental conflict predicted greater use of conflict engagement, physical aggression, and withdrawal and less positive affect. In contrast, interparental conflict was associated with less use of conflict engagement and more initiation of positive exchange among girls who make negative conflict appraisals. Moreover, interparental conflict was not associated with girls’ use of aggression or withdrawal.
As suggested earlier, adolescents’ negative appraisals of interparental conflict could generalize to romantic relationships, fostering heightened attention to conflict cues, negatively biased assessments of situations or partners, and ineffectual conflict tactics. Our findings suggest that gender role socialization is an important factor in shaping adolescents’ choice of conflict strategies. To the extent that boys’ socialization emphasizes agency, assertion, and individual well-being, power assertive acts (e.g., verbal or physical aggression) or self-focused regulation behaviors (e.g., withdrawal) may become favored in the face of negative conflict appraisals. For girls, whose socialization emphasizes interpersonal connection, sensitivity to conflict could trigger efforts to maintain a positive connection (Ruble & Martin, 1998). We had originally hypothesized that this dynamic would be manifested by compliance. Although this pattern emerged, it was not specific to girls. What was particular to girls was a tendency to avoid or “smooth over” potential conflict with romantic partners when interparental conflict was accompanied by negative conflict appraisals. The extent to which such behavior encourages positive outcomes or incurs psychosocial costs (e.g., self-silencing, diminished authenticity) would be an important topic for further investigation (Neff & Harter, 2002).
It is worth noting that the gendered pattern of effects does not appear to reflect base rate differences in boys’ and girls’ use of various conflict behaviors. The only mean level difference in boys’ and girls’ self-reported or observed conflict behavior was for conflict engagement, on which girls scored higher than boys. In our view, the current pattern of results suggests that appraisals of interparental conflict might signify different vulnerabilities for boys and girls in their romantic interactions (Davies & Lindsay, 2004). The role of emotion socialization in shaping these behavioral patterns is an important area for future investigation. For example, boys’ proclivity toward engagement, aggression, and withdrawal strategies may speak to the importance of anger regulation in boys’ responses to interpersonal conflict and negative conflict appraisals (Kingsfogel & Grych, 2004). Some might externalize anger via power assertive behavior whereas others might regulate anger through withdrawal (Gottman & Krokoff, 1989; Markman et al., 1993). Girls’ tendency to smooth over real or potential conflict may be more closely tied to fear. Self-blaming appraisals may lead to fear of retribution, and appraisals of threat may reflect fears of physical or psychological harm. If true, girls with negative conflict appraisals may be more likely associate interparental conflict with danger or defenselessness.
The only moderated effect that was not gender specific was compliance. Interparental conflict was associated with more use of compliance for all adolescents who made more negative conflict appraisals. Although we had hypothesized that this effect would be stronger for girls, it seems that both boys and girls who blame themselves for conflict or anticipate harmful consequences for conflict may readily yield to romantic partners’ preferences as a way to atone for potential wrongdoing or avert harm. Although this style could serve to circumvent some conflict, it may result in less intimate and satisfying relationships (Simon et al., 2008). As a conflict style, compliance requires individuals to subordinate personal needs and has been associated with a lack of authenticity in romantic relationships (Neff & Harter, 2002). To the extent that romantic relationships facilitate identity development, persistent self-subjugation could have important implications for adolescents’ well-being as well the quality of their romantic relationships (Furman & Shaffer, 2003).
The current study extends prior research by examining adolescents’ perceptions and appraisals of interparental conflict as independent and interactive predictors of self-reported and observer-rated conflict in adolescents’ romantic relationships. Yet several limitations need to be considered when interpreting the results. For example, we have discussed the findings in ways that would suggest that parents’ marital behavior affects adolescents’ romantic experiences. Intuitively, this seems most plausible but the data are correlational, and the effects could be in the other direction, bi-directional, or reflect a third unmeasured variable. Although we assessed numerous conflict tactics, our list was not exhaustive, and other conflict behaviors merit attention. For example, many adolescents use relational aggression to manage interpersonal conflicts, and some studies suggest that it is more common among girls than boys (Linder, Crick, & Collins, 2002; Zimmer-Gembeck, Geiger, & Crick, 2005). Hence, it is possible that a greater number of findings emerged for boys than girls in the current study because the conflict dimensions assessed were more salient for boys than girls.
Sample characteristics may limit the generalization of our findings. First, all participants were heterosexual. It is not clear whether heterosexual parents’ relationships are equally salient to heterosexual and sexual minority youth or whether similar findings could be expected for adolescents living with a same-sex couple. In addition, we studied intact families, most of whom were biological parents. Further research is needed to examine adoptive or blended family contexts as well as the effects of parental divorce. Lastly, the recruitment of participants through advertisements and home mailings could have biased sample selection in ways that could limit the external validity of the findings.
Another potential limitation of the current study concerns the use of adolescent report to assess both interparental and romantic relationship conflict, which could have inflated associations among the constructs. Yet it is also possible that adolescents are the most accurate reporters of their conflict exposure, as several studies indicate that their reports of interparental conflict are better concurrent and prospective predictors of adjustment than either parent or researcher ratings (Amato & Booth, 2001; Cummings, Davies, & Simpson, 1994; Emery & O’Leary, 1982; Grych & Fincham, 1990). Furthermore, adolescents’ ratings tend to be more closely related to researchers’ ratings than are parents’ ratings (Cummings, et al., 1994). Similar concerns regarding mono-method informant bias may also limit e interpretation of the moderated effects. However, the correspondence of results across self-reported and observed romantic behavior lends credibility to the validity of adolescents’ self-reports about conflict in romantic relationships and suggests that the significant results were not spurious. The sheer number of correlational analyses increases the possibility of Type I errors; but the pattern of significant results was consistent and readily interpretable.
Overall, the presence of links between interparental conflict and romantic relationship conflict during adolescence is noteworthy. Longitudinal studies are required to determine how these associations unfold over time. Links first seen in adolescents’ relationships could grow stronger, with behavioral patterns repeated and expectations reinforced across emerging adults’ relationships with various romantic partners. The associations might also become stronger over time as adolescents’ relationships mature and take on the centrality and commitment more closely resembling that of marriage. Alternatively, they could become weaker as individuals acquire more direct experience in addressing conflicts in romantic relationships. Whatever the course, the current findings suggest that links between parents’ and their offspring’s romantic relationships are present during adolescence.
Preparation of this manuscript was supported by Grant 535092 from the National Institute of Mental Health (V.A. Simon, P.I.) and Grant 50106 from the National Institute of Mental Health (W. Furman, P.I.) and Grant HD049080 from the National Institute of Child Health and Human Development (W. Furman, P.I.). We would like to thank MaryFrances Porter for her co-ordination of the data collection and Rachel Derrington for her assistance in coding the videotapes. Thanks also go to the participating adolescents, families, and schools.
Valerie A. Simon, Assistant Professor of Psychology at the Merrill Palmer Skillman Institute at Wayne State University, 71 East Ferry Street, Detroit MI, 48202. She can be reached by ; Email: ude.enyaw@nomisv or at 313.871.1790.
Wyndol Furman, John Evans Professor of Psychology at the University of Denver, 2155 Race Street, Denver CO, 80208. He may be reached by ; Email: ude.ud.ysp@namrufw or at 303-871-3688.