Introduction

Marital conflict is an inevitable part of a marital system. Conflicts within families and marriages may sometimes need to occur if important issues are to be worked out. Children who are exposed to well-managed and well-resolved conflict between their parents can learn constructive problem solving and strategies for coping with interpersonal conflict in other relationships (Cummings and Davies 1994; Grych and Fincham 1990). When poorly managed and unresolved, however, conflict can be upsetting and detrimental to children who are exposed to, and often involved in, their parents’ fights. Marital discord has been associated with a number of indices of maladjustment in children, including externalizing behavioral problems such as aggression and internalizing behavioral problems such as depression and anxiety (Cummings and Davies 1994; Davies and Cummings 1994; Emery 1982; Gottman and Fainsilber-Katz 1988; Grych and Fincham 1990; Lindahl and Malik 1999). In addition, children of distressed marriages have been found to evidence higher rates of academic, interpersonal, and physical health problems (e.g., Cummings and Davies 1994). Studies provide evidence of relations between children’s processes of emotional arousal and histories of marital conflict (Cummings et al. 1989b). Higher levels of previous exposure to marital conflict leave children primed for higher, and even more negative, emotional responses in later conflict contexts (Cummings et al. 1985, 1989b).

Researchers have proposed possible mechanisms by which poorly managed marital conflict affects children. These can be conceptualized as two general processes (Emery et al. 1992): (a) a direct process whereby witnessing marital conflict is associated with children’s emotional distress and physiological arousal and (b) indirect processes whereby marital conflict affects children through other aspects of child and family functioning. Others have written about child variables that mediate the impact of marital conflict, including children’s perceptions and appraisals of conflict (Grych and Fincham 1990) and disruption in children’s emotional security (Davies and Cummings 1994).

This view is evident in Grych and Fincham’s (1990) cognitive contextual model for understanding the relations between marital discord and child adjustment. Their model hypothesizes that children’s experience with marital conflict involves cognitive and affective appraisal of the significance and implications of marital conflict, as well as the coping behaviors enacted in response to such conflict. Appraisals that are particularly relevant for influencing the impact of conflict on children are perceived threat, coping efficacy, causal attributions, and ascription of blame (Grych and Fincham 1990). In other words, the cognitive-contextual framework posits that children’s cognitive processing and coping behaviors are shaped by the characteristics of marital conflict and contextual factors such as past experiences with conflict.

Grych and Fincham (1990) further suggest that children’s adjustment may be adversely affected when children develop dysfunctional appraisals and maladaptive coping behaviors. In this way, the manner in which children perceive and experience marital conflict may have implications for their responding to interparental conflict, and for their subsequent adjustment within their family. Studies show that children differentiate between child-related and non-child-related conflicts (Grych and Fincham 1993; Jouriles et al. 1991). Moreover, children discriminate among physical and verbal aggression in marital conflicts (Cummings et al. 1989a). Nonverbal expressions of anger and conflict also elicit distress from children (Cummings et al. 1991). In regards to coping, children who expect that conflict will escalate have been found to experience greater sense of fear and threat when they perceive marital conflict and may be more likely to intervene in the conflict (Grych and Cardoza-Fernandes 2001).

Such explorations on the relations between marital conflict and children’s adjustment have made significant contributions to the field of child and family studies. However, much of the research has focused on samples consisting mainly of European Americans. Thus, it is not clear whether previous findings can be generalized to members of ethnic minorities. To this end, the present study sought to investigate the effects of marital conflict on children’s emotional, cognitive and coping responses by exposing Korean children to experimental marital conflict situations varying in intensity and content. Because extant literature conducted in the United States with primarily European American children indicates that children evidence heightened reactivity to more intense conflict situations (Grych and Fincham 1993) and conflict situations that involve the child (Grych and Fincham 1993), the main focus of this study was to examine whether similar results would be found in a sample of Korean children. Furthermore, the study examined whether history of exposure to interparental conflict is related to more indices of maladjustment in children.

As such, the present study explored the associations between Korean children’s perceived marital conflict, their emotional, behavioral, and cognitive responses and their adjustment under the cognitive-contextual framework. It was hypothesized that Korean children’s perceptions of marital conflict would be related to their emotional and behavioral responding to interparental conflict. Specifically, it was anticipated that children who reported higher levels of interparental conflict at home would evidence increased affective distress and heightened behavioral responding to such conflict. Secondly, it was hypothesized that Korean children reporting a high degree of experiential history of interparental conflict would manifest more indices of depression and internalizing and externalizing behavioral problems.

Method

Participants

Based on the responses of 221 children enrolled in 4th, 5th, and 6th grades of an elementary school located in Seoul, Korea on the Children’s Perception of Interparental Conflict Scale (CPIC; Grych et al. 1992), a total of 116 children whose scores belonged in the upper and lower 25th percentile were divided into “high-conflict group” and “low-conflict group”, respectively. Exclusion criteria included children who were not living with their parents or children whose parents have divorced. Hence, 58 children (23 boys, 32 girls) comprised the “low-conflict group” (CPIC M = 13.19, SD = 2.76) and 58 children (36 boys and 22 girls) comprised the “high-conflict group” (CPIC M = 44.86, SD = 11.00). The mean age for the low-conflict group was 10.6 years (SD = .98) and 10.5 years (SD = .96) for the high-conflict group. The two groups differed in father’s education level, t(114) = 2.05, p < .05, and parents’ perceived socioeconomic status, t(114) = 2.29, p < .05, with fathers in the low-conflict group with more years of education and higher reported socioeconomic status. This study selected upper-grade level children because we were interested in obtaining children’s self-reports of coping and appraisals of conflict and these measures are more valid with children in this age range than with younger children.

Procedure

Children participated in the study in their school classrooms under the supervision of their teachers. Children were presented with a booklet containing four cartoon depictions of marital conflict situations with a response record form following each vignette. At the end of each cartoon vignette, a questioning schedule described in detail elsewhere (e.g., Cummings et al. 1989a, 1991) was presented. To control for possible order effects, the order of presentation of the four different conflict vignettes was randomly ordered by classroom. The children also completed several self-report measures and questionnaires. Mothers independently completed standardized measures aimed to examine children’s emotional and behavioral adjustment.

Marital Conflict Stimuli Vignettes

Children were presented with cartoon depictions of marital conflict situations. The conflict vignettes’ intensity was manipulated primarily through the degree of anger expressed by parents. In the low-intensity (nonverbal) conflicts, the parents disagreed and they opted for the “silent treatment” of each other. In contrast, in the high-intensity (verbal/physical aggression) conflict situations, marital dispute was portrayed through loud, angry disagreement and suggested a physical scuffle behind closed doors. The content of the conflicts also varied. In the child-related conflicts, the topic of the disagreement involved the child (e.g., conflict over child rearing differences). In non-child-related conflicts, the topic concerned something unrelated to the child (conflict with relatives). Hence, four cartoon depictions of marital conflict were presented.

Children’s Appraisal Measures

Children’s Appraisal of Marital Conflict Situations

A response record from based on Grych and Fincham’s (1990) cognitive-contextual framework was translated into Korean by a bilingual graduate student and back translated by a graduate-level student. The initial questions were devised to assess children’s emotional (mad, scared, sad, worried, helpless, and ashamed) and cognitive (e.g., perceived threat, child involvement, self-blame) responses to each of the four marital conflict vignettes. Children were instructed to imagine each conflict situation occurring in their homes and base their responses accordingly. A sample item measuring children’s belief that the conflict would escalate was “The disagreement will get worse.” A sample item measuring children’s appraisal of their involvement was “I would have to take sides.” Children were asked to indicate their response by checking “strongly disagree”, “disagree”, “agree”, or “strongly agree.” Perceived threat was assessed by inquiring about two types of beliefs: that the conflict would escalate, and that the child would be drawn into the conflict. Children also rated the degree to which the child was to blame for the conflict on this 4-point scale. Their appraisals of efficacy were their ability to do something about the conflict and its likely effectiveness.

Coding of Coping Strategies

In order to classify the children’s coping strategy, children were asked to write down what they would do in that particular situation. This open-ended question was intended to assess the child’s understanding of how to resolve conflicts. Their first coping strategy was coded and classified by two independent raters into one of nine a priori categories: intervene (e.g., “I’d tell them to stop fighting”), be obedient (e.g., “I’d do what they told me to do”), side with one parent (e.g., “I’d say I agree with Mom”), address the source of the conflict (e.g., “I’d tell them they don’t have to get me anything”), physical withdrawal (e.g., “I’d go to my room”), self-distraction (e.g., “I’d play with games”), seek support (e.g., “I’ll go and talk with grandmother”), emotional response (e.g., “I’d start crying”), and ignore (e.g., “I’d just stay out of it”). Responses that did not fall into one of these categories were coded as “Other.” The nine categories were further classified into direct intervention (intervene and side with one parent), indirect intervention (be obedient and address the source of the conflict), withdrawal (physical withdrawal and seek support elsewhere), and no reaction (self-distraction and ignore). Information regarding age, sex, and group identification was removed for the coding. Inter-rater reliability was assessed by determining the degree of agreement between the two raters. The inter-rater reliability coefficient was found to be .82.

Children’s Perceptions of Interparental Conflict Scale (CPIC; Grych et al. 1992)

This is a 51-item questionnaire designed to assess children’s appraisals of multiple dimensions of marital conflict. The CPIC yields a total score and three subscales that assess perceptions of parental conflict, threat to the self, and self-blame. The Conflict Properties Scale, the subscale used in this study, assesses children’s appraisals of the intensity, frequency, and level of resolution of their parents’ conflict, with higher scores indicating perceptions of conflict that occurs often, involves higher levels of hostility and aggression, and is poorly resolved. This scale was translated into Korean from English by a bilingual graduate student and back translated by another graduate student in clinical psychology. Very good internal consistency was found for this scale (α = .90).

Children’s Emotional and Behavioral Problem Measures

Korean Child Behavior Checklist (K-CBCL; Oh et al. 1997)

Mothers completed the K-CBCL, the Korean standardized version of the Child Behavior Checklist (CBCL), a measure of children’s competence and externalizing and internalizing behaviors (Achenbach and Edelbrock 1983) with well-established reliability and validity. The K-CBCL is a 118-item checklist of problem behaviors and competencies that parents rated on a 3-point scale ranging from not true (0) to very true or often true (2) of their child in the past 6 months. We report only the internalizing (i.e., somatic complaints, withdrawn and anxious-depressed behaviors) and externalizing (i.e., delinquent and aggressive behaviors) syndromes.

Children’s Depression Inventory (CDI; Cho and Lee 1990; Kovacs 1985)

The Korean version of the CDI (Kovacs 1985) was employed in the present study. This is a 27-item self-report measure of depression. Each item on the CDI consists of a set of three statements describing a symptom of depression. Responses are scored from 0 to 2, with larger numbers indicating more severe expression of a symptom. The CDI demonstrates high internal consistency in normal samples (α = .94; Saylor et al. 1984). The internal consistency in this study was found to be .84.

Results

Effects of Conflict Group, Conflict Intensity and Content on Children’s Responses

Children’s affective and cognitive responses were initially analyzed in 2 × 2 × 2 (conflict exposure group × conflict intensity × conflict content) MANOVAs. The first two factors were repeated measures, and the latter factor was between-subjects factor. Tables 1 and 2 list the means and standard deviations for the affective and cognitive variables across the group and intensity and content conditions and indicate which effects were significant.

Table 1 Mean and standard deviation of children’s emotional responses according to Group (G), experimental interparental conflict situations’ intensity (I) and content (C)
Full size table
Table 2 Mean and standard deviation of children’s cognitive response according to group (G), experimental interparental conflict situations’ intensity (I) and content (C)
Full size table

Affective Responses

The effects of children’s perceptions of marital conflict, and intensity and content of hypothetical conflict situations on children’s responses to the 6 emotional reactions to marital conflict scenarios were examined with MANOVA using the 6 affect ratings as dependent variables. With the exception of “sadness”, “anger” and “shame”, children in the high-conflict group reported as experiencing emotions to a greater degree. High-conflict group experienced more fear, F(1, 114) = 12.56, p < .001, worried more, F(1, 114) = 5.72, p < .001, and felt more helpless, F(1, 114) = 18.52, p < .001, than children in the low-conflict group. Analyses also indicated that high-intensity conflicts led to greater anger, F(1, 115) = 9.63, p < .01, fear, F(1, 115) = 34.27, p < .001, sadness, F(1, 115) = 13.89, p < .001, worry, F(1, 115) = 4.21, p < .05, and shame, F(1, 115) = 34.46, p < .001.

However, results pertaining to anger, sadness, and shame need to be qualified by a significant interaction with content, F(1, 114) = 4.89, p < .001, F(1, 114) = 5.60, p < .05, F(1, 114) = 25.27, p < .001, respectively. Furthermore, there was a conflict group × intensity × content interaction effect on the measure of fear, F(1, 114) = 5.54, p < .05. One-way ANOVA and post-hoc analyses (Scheffé) were computed to identify the source of interaction effects. Analyses showed that compared to nonverbal-non-child-related conflict situation (M = 2.18, SD = .99), verbal/physical aggression-child-related conflict situation aroused more anger (M = 2.43, SD = 1.14), F(3, 460) = 3.28, p < .05). In addition, compared to the two nonverbal conflict situations, children reported more shame in the high intensity-child-related conflict situation, F(3, 460) = 6.60, p < .001. The main effect for content was significant for shame, F(1, 115) = 13.38, p < .001, with child-related content leading to greater fear than non-child-related content. Such results suggest that conflict intensity affects children’s emotional reaction more so than content.

Cognitive Responses

2 × 2 × 2 (conflict exposure group × conflict intensity × conflict content) MANOVAs were performed on the 7 cognitive responses of children to the conflict vignettes. A significant main effect for group was found for all of the cognitive indices of perceived threat. Children with experiential histories of exposure to marital conflict evidenced greater self-blame, F(1, 114) = 11.28, p < .001, felt more drawn into the conflict, F(1, 114) = 25.69, p < .001, showed greater likelihood to believe that the conflict will escalate, F(1, 114) = 24.84, p < .001, that they must take sides, F(1, 114) = 14.89, p < .001, and that the conflict will get more serious, F(1, 114) = 36.49, p < .001. Furthermore, Korean children in the low-conflict group reported stronger belief that the parent’s relationship will improve, F(1, 114) = 12.92, p < .001, and that the parents will resolve the conflict, F(1, 114) = 30.20, p < .001.

A main effect for conflict intensity was found. The extent to which the conflict was perceived as the child’s fault was greater when conflict was intense, F(1, 115) = 13.25, p < .001, and child related F(1, 115) = 33.22, p < .001. The expectation that the conflict would escalate, was greater for child-related than non-child-related conflict, F(1, 115) = 8.85, p < .01, but was not related to content. In a second threat measure children reported greater fear of becoming involved in the conflict when conflict concerned the child, F(1, 115) = 22.55, p < .001, but was not related to intensity. In addition, the interaction of content and intensity was significant for children’s self-blame, F(1, 114) = 42.71, p < .001. Univariate F tests showed that self-blame is higher for child-related conflict across levels of intensity, but that it is particularly high when conflict is high in intensity and child related, F(3, 460) = 15.62, p < .001. Intensity × content interaction was also found in children feeling drawn into the conflict, F(1, 114) = 16.81, p < .05. Univariate F tests were significant in the direction hypothesized: children felt more drawn into conflict situations that were more intense and involved them in content, F(3, 460) = 6.77, p < .001. In sum, in relation to non-child-related conflicts, conflicts about children elicited more shame, responsibility, self-blame, and fear of being drawn into the conflict.

Children’s Coping Response to Marital Conflict Stimuli

A 2 × 2 × 2 (conflict exposure group × conflict intensity × conflict content) MANOVA was first performed. First, children’s proposed coping strategy for each conflict situation was coded as follows: Direct Intervention scored 1, Indirect Intervention scored 2, Withdrawal scored 3, and No Behavioral Action scored 4. No main effect was found for conflict intensity. A main effect for conflict content was found. Results indicated that when the conflict concerned the child in content, children tended to endorse more direct intervention coping strategies, F(1, 115) = 12.99, p < .001. A significant conflict intensity × content interaction effect was found in the preferred method of coping, F(1, 114) = 5.56, p < .05. One-way ANOVA and post-hoc analyses revealed that children were particularly likely to endorse direct intervention coping strategies in the nonverbal (less intense) conflict situations that did not involve the child in content, F(3, 460) = 3.99, p < .01.

To determine whether the distribution of the frequency of coping strategies endorsed by the two conflict groups (high versus low experiential history of marital conflict) is significantly different, Mann–Whitney test was performed (Table 3). The results revealed that compared to low-conflict group children, children in the high-conflict group endorsed more indirect coping strategies, z = 2.38, p < .05 and passive-aggressive coping strategies, z = 2.22, p < .05. Low-conflict group children preferred more direct coping strategies, z = 2.29, p < .05.

Table 3 Frequency and percentage of children’s coping strategy by conflict group
Full size table

Experiential History of Marital Conflict and Children’s Adjustment

Children’s internalizing and externalizing behavioral problems were compared across the two conflict groups (Table 4). As expected, children of families with interparental conflict were reported to exhibit more externalizing behavioral problems, t(114) = 4.01, p < .001, internalizing problems, t(114) = 5.56, p < .001, and more depressive symptoms, t(114) = 8.04, p < .001, than children from families not reporting such conflicts between parents.

Table 4 Children’s emotional and behavioral problems according to conflict group
Full size table

Discussion

Guided by Grych and Fincham’s (1990) cognitive-contextual framework, the present study aimed to delineate children’s appraisals of and responses to different types of marital conflict in a sample of Korean children. To this end, children’s affective, cognitive, and coping responses to cartoon depictions of marital conflict situations were examined as a function of (a) children’s report of the degree of experiential history of marital conflict at home, (b) intensity of marital conflict (nonverbal versus verbal/physical), and (c) the content of conflict (non-child-related versus child-related).

The present study extends the findings of existing literature by showing that in addition to creating greater negative affect, conflict characterized by anger and hostility leads to greater self-blame and increases fears that the conflict will escalate and involve the child. We also found that unlike intensity, which had a generalized negative effect on children’s affect and cognitions, the content of conflict affected only certain types of appraisals. Compared to conflict concerning topics unrelated to the child, child-related conflicts led to greater fear of the child being drawn into the conflict, shame, and self-blame. Content was not significantly related to feelings of anger, sadness, or worry. These findings are consonant with previous studies (e.g., Grych and Fincham 1990). Child-related conflicts thus appear to have a greater impact on those variables most closely related to the self and self-evaluation.

The findings also confirm that children’s history of exposure to conflict between parents affects their emotional reactions, cognitive appraisals and coping strategies when confronted with hypothetical marital conflict situations. Children experiencing a high degree of interparental conflict evidenced heightened reactivity and sensitivity in responding to conflict situations. Thus, this study supports previous findings conducted in Western cultures that repeated exposure to hostility sensitizes children to discord (Cummings et al. 1985, 1989b). In particular, conflict characterized by physical aggression and related to child in content led to greater self-blame. Furthermore, experiential history of interparental distress related to more indices of maladjustment in children. Children reporting a high level of interparental conflict manifested more externalizing and internalizing behavior problems in addition to more self-reported depressive symptoms.

Such findings are consistent with recent conceptualizations of children’s adjustment to marital conflict, in which children’s appraisals of conflict are linked to patterns of emotional and behavioral responding to such conflict (Davies and Cummings 1994; Grych and Fincham 1990). Specifically, children’s perceptions of high levels of marital conflict appear to motivate a pattern of affective distress and behavioral dysregulation. Children’s patterns of heightened responding to marital discord may be understood as serving a short-term function within the family, wherein expressions of distress and behavioral disruption may alert parents to their children’s distress, resulting in efforts to shield children from over displays of conflict.

The present study also examined the strategies that children generated for coping with conflict. Children’s preference in coping strategy is of particular interest because it is likely to be an important determinant of whether children become involved in their parents’ conflicts. Our findings suggest that certain types of appraisals and characteristics of conflict, especially its content, increase the probability that children will get drawn into their parents’ arguments. Children reported feeling more able to help resolve the conflict and more willing to directly intervene when the conflict was less intent and child related.

Contrary to studies conducted in the United States with European American children (e.g., Grych and Fincham 1993), however, Korean children reporting a high degree of experiential history of interparental conflict did not prefer direct intervention as their preferred way of coping, indicating the possibility of cross-cultural differences in coping with marital conflict. In particular, a new category of “passive-aggressive” coping strategy (e.g., “I will not eat”, “I will beat up my brother”) had to be newly included in this study because of a relatively large number of such responses. Further analysis also indicated that in comparison to children in the low-conflict group, more children in the high-conflict group chose such strategies. Future research is needed in order to classify whether such difference in coping as a function of experiential history of interparental conflict is a culturally unique phenomenon and the implications it has on children’s adjustment.

One possible explanation may involve the cross-cultural difference in family interaction patterns. In a study of family interaction patterns of Japanese American and European American families (Hsu et al. 1985), European Americans were rated as significantly clearer in self-disclosure than Japanese, who were seen to be somewhat reticent to express individual thoughts and feelings. Making or expressing clear statements about individual feelings and thoughts may risk disagreement and endanger harmonious social relationships. It is thus culturally adaptive for those of East Asian descent to be more indirect and ambiguous. Marital anger could reflect a family tendency to express both positive and negative emotion, a pattern that has been associated with social competence in children (Cassidy et al. 1992). However, there may be cross-cultural differences in the manner in which a family tends to express positive and negative emotions (e.g., Hess et al. 1987) and this should be investigated further within the context of investigating the links between marital conflict and children’s adjustment in non-Western samples.

More research is needed to address unresolved questions about possible gender and age-related differences in children’s involvement in and exposure to parental conflict and gender-related differences in children’s expression of conflict-related distress (Snyder 1998)—especially in the context of culture—in addition to examining developmental change in children’s understanding of conflict, strategies for dealing with conflict, and the nature of conflict-related distress (Grych and Fincham 1990). Continuing research to assess the stability of appraisals across time with Korean and other ethnic minority children may reveal differences in how these children appraise and respond to marital conflict.

This study focused on relations between Korean children’s history of exposure to interparental conflict and their appraisals and responses to conflict situations. What is clear from this study is that Korean children, much like children from Western cultures, respond to interparental conflict with greater psychological distress and exposure to marital conflict is related to indices of child maladjustment. A goal for future work will be to broaden examination of the network of processes that may influence the strategies used by children when they confront interparental conflict situations—not only by broadening the topology of children’s response patterns and conflict situation structural variables, but also by not ignoring to encompass a cross-cultural perspective. A theoretical approach to the question of how ethnic and cultural background might shape the impact of interparental conflict on children is needed. Central to such an approach would be the consideration of the unique histories, experiences, environments and cultural orientations of families of color.