Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 1 Delineating the Sequelae of Destructive and Constructive Interparental Conflict for Children Within an Evolutionary Framework Date Submitted: September 13, 2011

We examined the joint role of constructive and destructive interparental conflict in predicting children’s emotional insecurity and psychological problems. In Study 1, 250 early adolescents (M = 12.6 years) and their primary caregivers completed assessments of family and child functioning. In Study 2, 201 mothers and their two-year old children participated in a multi-method, longitudinal design with three annual measurement occasions. Findings from structural equation modeling in both studies revealed that children’s emotional insecurity in the interparental relationship mediated associations between destructive interparental conflict and children’s psychological problems even after including constructive conflict and family and child covariates as predictors. Conversely, emotional insecurity was not a mediator of associations between constructive interparental conflict and children’s psychological problems when destructive interparental conflict was specified as a risk factor in the analyses. The results are consistent with the evolutionary reformulation of emotional security theory and the resulting primacy ascribed to destructive interparental conflict in accounting for individual differences in children’s emotional insecurity and its pathogenic implications (Davies & Sturge-Apple, 2007). Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 3 Delineating the Sequelae of Destructive and Constructive Interparental Conflict for Children Within an Evolutionary Framework Children exposed to high levels of conflict between parents are more likely to exhibit emotional distress in response to interparental conflict and broader patterns of psychological difficulties (e.g., internalizing and externalizing symptoms) than are children from harmonious homes (Cummings & Davies, 2010). In attempting to understand the unfolding developmental pathways underlying children’s vulnerability, several theories have proposed that children’s distress in response to interparental conflict is a key mechanism mediating interparental conflict and later problems (e.g., Davies & Cummings, 1994; Grych & Fincham, 1990). In support of this assumption, studies have consistently identified children’s distress responses to interparental conflict as significant mediating factors underlying their heightened vulnerability to psychological difficulties when exposed to heightened conflict between parents (Buehler, Lange, & Franck, 2007; Cummings, Schermerhorn, Davies, Goeke-Morey, & Cummings, 2006; Davies, Harold, Goeke-Morey, & Cummings, 2002; Fosco & Grych, 2008). However, the treatment of interparental conflict as a homogeneous construct in these studies has not afforded an analysis of whether different forms of conflict have unique implications for children’s reactivity and their adjustment. Couples vary widely in how they approach and manage conflicts, with some resorting to aggression, disengagement, and escalation of distress and others making efforts to work in a collaborative way toward compromise and resolution. Thus, advancing a process-oriented approach to understanding individual variation in the links among interparental discord, children’s distress reactivity to conflict, and their adjustment requires an understanding of the specific stimulus characteristics of interparental conflict that may have implications for individual differences in child reactivity and Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 4 adjustment (Cummings & Davies, 2010). As a first step in addressing this gap, the objective of this study was to explore the relative roles of constructive and destructive conflict between parents in informing pathways between interparental discord, children’s distress, and their psychological difficulties. Because empirical distinctions between conflict tactics in models of child development have largely been guided by emotional security theory (EST; Cummings & Davies, 2010), our aim is to specifically discriminate between these two forms of conflict in advancing an understanding of children’s emotional security. According to EST (Davies & Cummings, 1994), repeated exposure to destructive interparental conflict characterized by elevated hostility and conflict escalation undermine children’s goal of preserving their security and are manifested overtly in their heightened distress, negative appraisals, and elevated involvement in response to interparental conflict. Elevated concerns about security, in turn, are proposed to increase children’s risk for a wide array of psychological problems. Building on the empirical delineation of differential associations between forms of interpartner conflict and adults’ relationship quality and well-being (e.g., Gottman, 1994; Notarius & Markman, 1993), Cummings and colleagues proposed that distinguishing between destructive and constructive interparental interactions from the child’s perspective may help to inform an understanding of children’s adjustment over time (Cummings & Cummings, 1988; Goeke-Morey, Cummings, Harold, & Shelton, 2003). Utilizing both experimental simulations of interadult conflict and parental diary procedures, studies support the classification of parent aggression and hostility as destructive conflict properties evoking negative emotions and behaviors in children (e.g., Goeke-Morey et al., 2003; Goeke-Morey, Cummings, & Papp, 2007). Other research further supports these “destructive” forms of interparental conflict in increasing children’s vulnerability for mental health problems through its association with children’s Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 5 emotional insecurity in the interparental relationship (Davies, Harold et al., 2002; El-Sheikh, Cummings, Kouros, Elmore-Staton, & Buckhalt, 2008; McCoy, Cummings, & Davies, 2009). Conversely, parental support, cooperation, and problem-solving have been designated as “constructive” from the child’s perspective based on their associations with diminished negative emotions (Cummings & Wilson, 1999; Cummings, Goeke-Morey, Papp, & Dukewich, 2002; Cummings, Geoke-Morey, & Papp, 2004; Goeke-Morey et al., 2003; 2007). It is further hypothesized that these constructive tactics have salubrious implications for children’s mental health by reducing children’s concerns about security in the interparental relationship (Cummings, Goeke-Morey, & Papp, 2003; Easterbrooks, Cummings, & Emde, 1994). Although there is less evidence in support of the child sequelae associated with exposure to constructive conflicts, a recent study found that the association between constructive interparental conflict tactics and children’s prosocial behavior were mediated by signs of diminished child insecurity in the interparental relationship (McCoy, Cummings, & Davies, 2009). Analysis of the nature and frequency of conflict behaviors in the home indicates that the vast majority of couples use both constructive and destructive tactics during conflicts (Cummings et al., 2004). Yet, no studies, to our knowledge, have delineated how these two primary classes of conflict tactics operate jointly in understanding how children cope with and adapt to interparental conflict. In rare cases in which the two types of conflict are distinguished from one another, the predominant approach is to isolate constructive and destructive sets of conflict tactics as predictors of child adjustment in separate analytic models (e.g., McCoy et al., 2009; Cummings et al., 2002; 2004; Goeke-Morey et al., 2003; 2007). As a result, little is known about the interplay between exposure to destructive and constructive forms of conflict in advancing an understanding of children’s coping and adjustment to interparental conflict. Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 6 Accordingly, this study is designed to break new ground by examining the additive and interactive roles of these two classes of interparental conflict tactics in predicting signs of children’s emotional insecurity in the interparental relationship and, in turn, their psychological difficulties. Hypotheses regarding the interplay between constructive and destructive conflict vary across two conceptual versions of EST. According to the original version of EST, both conflict types (i.e., destructive, constructive) are assigned relatively equal empirical weight as independent contributors to child reactivity in the face of interparental conflict (Cummings & Davies, 1996; Davies & Cummings, 1994). A primary proposal is that parental conflict tactics can be distinguished in terms of their constructive and destructive properties based on the degree to which they increase or reduce insecurity (Goeke-Morey et al., 2003). Therefore, interparental conflict characterized by aggression, withdrawal, and detachment is defined as destructive because it is posited to increase signs of insecurity. In contrast, parental displays of cooperation, collaboration, and resolution during conflicts are postulated to enhance children’s security, in part, by reducing their signs of distress (McCoy et al., 2009). Taken together, the underlying assumption of the original formulation of EST is that indicators of children’s emotional insecurity are products of their exposure to both constructive and destructive properties of interparental conflict. The evolutionary reformulation of emotional security theory offers an alternative hypothesis (EST-R; Davies & Sturge-Apple, 2007; Davies & Woitach, 2008). Although EST-R shares with the original theory the assumption that security is a primary goal for children, it differs in proposing that this goal is largely organized around the operation of the social defense system, a behavioral system that evolved through our ancestral history to defuse and avoid threat Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 7 from other members of the social group (Gilbert, 1993; Davies & Sturge-Apple, 2007; Ohman & Mineka, 2001). Operating from the assumption that the emotional security system is specifically designed to contend with conspecific danger, interpersonal threat cues (e.g., angry expressions, dominant posturing, loud voices) are theorized to assume greater primacy in organizing children’s distress responses in social contexts compared to expressions of interpersonal cooperation or support. Thus, although EST-R acknowledges that constructive interparental interactions are likely to have important implications for children’s psychological adjustment, insecurity in the interparental relationship is not proposed to be a primary mediator in these pathways. As a derivative hypothesis, antagonism between parents is expected to be more salient than constructive interparental conflict as a predictor of children’s insecurity in the interparental relationship. A primary objective of this paper is to test the relative viability of the original and reformulated accounts of emotional security theory in explicating the role of emotional security as a mediator in pathways between constructive and destructive interparental conflict and children’s psychological adjustment. Whereas EST proposes that constructive and destructive conflict may uniquely increase children’s vulnerability to psychological problems by undermining their security in the interparental relationship, the more selective hypothesis of EST-R is that emotional insecurity will only serve as an explanatory mechanism in associations between destructive interparental conflict and children’s psychological symptoms. To test these alternative hypotheses, the present paper capitalizes on data drawn from two large studies (n > 200) of children and their families. Examination of the joint role of constructive and destructive interparental conflict raises two additional questions about the operation of pathways within a multivariate context. First, Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 8 pathways between one form (i.e., destructive, constructive) of interparental conflict and children’s functioning may vary as a function of the other form of conflict. For example, in the original EST, destructive bouts of interparental conflict are proposed to assume a more benign meaning for children if they are balanced by expressions of harmony, warmth, and cooperation between parents (Cummings & Davies, 2010; Davies, Harold et al., 2002). Thus, it is possible that witnessing constructive forms of interparental conflict serves as a protective factor in the risk posed by destructive conflict for children’s insecurity and psychological adjustment. As a way to examine this possibility, we explored whether the interaction between constructive and destructive interparental conflict uniquely predicts children’s security and psychological adjustment above and beyond the additive effects of each form. Second, although both versions of EST propose that the hypothesized interrelationships among interparental conflict and child functioning will continue to be robust in the context of other family factors, a plausible alternative hypothesis is that the associations are simply artifacts of the influence of co-occurring child, parenting, family-level, or sociodemographic characteristics. Therefore, to increase the rigor of our analyses for each study, we specify factors within each of these four classes of constructs as covariates in the primary analyses. Selection of these factors was based on a careful consideration of research and theory on the potential impact of specific dimensions of child, parent-child, family, and sociodemographic characteristics. First, at the level of child attributes, children’s general tendencies to experience heightened, prolonged distress is consistently identified as a factor that may artificially inflate mediational paths among interparental conflict, child insecurity, and child adjustment (e.g., Davies, Harold et al., 2002; Harold & Conger, 1997). Second, in the parent-child dyad, the quality of the emotional relationship between parents and children has been regarded as a pivotal factor that may explain Running Head: SEQUELAE OF DESTRUCTIVE AND CONSTRUCTIVE CONFLICT 9 or supersede associations between interparental conflict and children’s adjustment (Buehler, Benson, & Gerard, 2006; Levendosky, Leahy, Bogat, Davidson, & von Eye, 2006). Third, at the level of the family unit, prior research has underscored the possibility that difficulties maintaining cohesive, close-knit relationships and experiences with instability and disruptive events (e.g., changes in caregiver intimate relationships) may co-occur with interparental problems and, in turn, threaten children’s security (e.g., Ackerman, Kogos, Youngstrom, Schoff, & Izard, 1999; Winter, Davies, & Cummings, 2010). Finally, we also included indices of socioeconomic status (e.g., parent education, income) as covariates based on their implications for family and child functioning (Conger, Conger, & Martin, 2010)