Intra-Family Conflict in Relation to Boys’ Adjustment at School

 


Several models of the development of antisocial behavior emphasize the role of intra-family conflict on children’s later relationships and adjustment (Brody, Stoneman, & Burke, 1987; Patterson, Reid, & Dishion, 1992; MacKinnon-Lewis, Starnes, Volling, & Johnson, 1997).  Although individual models posit different mechanisms of effect, in general, early dyadic relationships within the family, such as interparental, parent-child, and sibling relationships, have been viewed as the training ground for the practice of verbal and physical aggression and other interpersonal conflict strategies and skills.  Both positive and negative strategies learned in the home are hypothesized to carry over into interactions in relationships in other contexts, such as with teachers and peers in school (Deutsch, 1973; Shantz, 1987).  Children who learn to use destructive and aggressive/aversive conflict negotiation skills in the home are hypothesized to then develop highly conflictual relationships with teachers and peers, which in turn has been associated with school failure, peer rejection, and later antisocial outcomes such as involvement with deviant peers and juvenile delinquency (Loeber & Dishion, 1983; Patterson et al., 1992). 

Much of the extant research has focused on demonstrating cross-sectional or short-term longitudinal associations between conflict occurring in two family subsystems (Brody et al., 1987; MacKinnon, 1989), or between conflict in one family dyad and child aggression (Patterson et al., 1992; Stocker & Dunn, 1990; Stormshak, Bellanti, Bierman, & Conduct Problems Prevention Research Group, 1996).  Fewer studies have been conducted to investigate the linkages between conflict in several different early family relationships and later conflict in relationships with teachers and peers.  The present study reports on longitudinal relations among different forms of dyadic conflict in the home (interparental, sibling, and parent-child) during the preschool years in relation to conflict in teacher-child and peer relationships at school entry, in a sample of low-income boys.


In this study, we focus on the effects of destructive conflict (i.e., conflict characterized by the use of coercion and threats to achieve goals; Shantz, 1987) in early relationships on later relationships.  Evidence that frequent, aggressive/coercive destructive conflict impedes the development of positive relationships has come from research from predominantly four theoretical perspectives: social learning (Patterson et al., 1992), family systems (Brody et al., 1987; MacKinnon-Lewis et al., 1997), attachment (Greenberg & Speltz; 1988), and sociological (Hotaling, Straus, & Lincoln, 1989; Straus, 1991a) models.  Patterson’s and colleagues (Patterson, Dishion, & Bank, 1984; Patterson et al., 1992) research has focused on how parents unwittingly train and reinforce aversive child behavior through social learning mechanisms, ultimately leading to the child’s open defiance and use of coercion in other relationships (Patterson, 1982).  Family systems models have posited that conflictual behavior occurring in a subsystem of the family (e.g., the interparental dyad) affects other subsystems in the family (e.g., sibling interactions; Brody et al., 1987; Minuchin, 1974), primarily by affecting the family’s emotional climate through an affective contagion mechanism (i.e., “spillover”; Erel & Burman, 1995).  Sociological models of violence such as Straus’ (1991a) hold that violence experienced in one area of life tends to engender violence in other areas, based on results from large, nationally-represented studies that have shown that the use of physical conflict strategies between parents is correlated with physical conflict in parent-child and child-child relationships (see Hotaling et al., 1989, and Straus, 1991a for reviews).  Finally, attachment theorists (Greenberg & Speltz, 1988; Pianta, 1997; Sroufe, 1983) have described how for insecurely attached children, a history of caregiver unresponsiveness would lead to the child learning to reciprocate with negative behaviors in order to elicit attention from caregivers, which in turn leads to negative and aversive behaviors in later close interpersonal relationships (Cichetti, 1984).  


In support of these models, recent investigations have demonstrated that the presence of high conflict in one family relationship is often correlated with conflict in another.  For example, high levels of marital conflict have been demonstrated to be positively related to parents’ and children’s reports of more frequent verbal and physical aggression toward the child (Osborne & Fincham, 1996; Strassberg, Dodge, Bates, & Pettit, 1992), as well as to harsh, rejecting parenting styles (see Davies & Cummings, 1994, for a review; Erel, Margolin, & John, 1998; Holden & Ritchie, 1991).  Sibling conflict has been shown to be related to interparental conflict and parent-child conflict (Brody, Stoneman, McCoy, & Forehand, 1992; MacKinnon-Lewis et al., 1997).  These studies support the idea of the development of conflictual, coercive processes across multiple relationships in early family life.  Only a few studies have explored the connections among conflict experienced in multiple family relationships and conflict in later relationships in school, particularly in early childhood (MacKinnon-Lewis et al., 1994).  Nonetheless, there is a modicum of studies that has examined relations between certain types of conflict in early family life and measures of school or peer competence that can be drawn upon to hypothesize about the continuity of conflictual processes across home and school.  These investigations will now be reviewed and discussed in relation to the conceptual model and hypotheses that were then tested in the present study.

Intra-Family Conflict and Child Adjustment

Exposure to destructive marital conflict is thought to teach children that aggression is an efficient tool to achieving interpersonal goals.  Moreover, frequent and intense early parental conflict may set an aversive, hostile family climate, providing the “background” in which coercive strategies are modeled, implicitly and explicitly taught, practiced, and then employed in interactions with others.  Studies of destructive marital conflict during the preschool years have shown marital discord to be related to a range of later child behavior problems (see Davies & Cummings, 1994, for a review).  Interparental relationships characterized by overt and hostile conflict has been positively associated with increased use of aggression with siblings (MacKinnon, 1989; Erel et al., 1998) and peers (Gottman & Fainsilber-Katz, 1989), as well as classroom sociometric status (Strassberg et al., 1992).  The range of behavioral outcomes associated with interparental conflict suggest that it may have a diffuse, general effect on children’s conflict style that is exhibited in many different relationships (e.g., with teachers, peers).


Yet when other family relationships are examined in conjunction with marital effects, an indirect and/or moderating effect is often demonstrated.  For example, Bierman and Smoot (1991) examined coercive family processes in relation to child conduct problems measured at ages 5-6.  They found that reported low marital satisfaction was correlated with child conduct problems in the home.  In subsequent path analyses, marital satisfaction and coercive parenting indices were demonstrated to contribute to the prediction of child conduct problems, which in turn, predicted problems with peers.  These findings are consistent with the idea that intra-family conflict plays a role in the genesis of child conduct problems, particularly when occurring in multiple relationships.  However, this study was also cross-sectional, and assessed marital satisfaction versus marital conflict, the latter of which has been more strongly related to child adjustment (Jouriles et al., 1991; Straus, 1991a).  Few if any studies have examined interparental conflict directly or interactively with characteristics of other family relationships in relation to conflict specifically in teacher-child and child-peer relationships.  From the results presented here, it seems possible that conflict in interparental and parent-child dyads might act in a synergistic (i.e., additive or interactive) way to affect the use of aversive conflict in relationships in other settings.


When qualities of the parent-child relationship have been examined in relation to child adjustment in school, the results support the idea that parent-child conflict might also have direct as well as moderating effects on both teacher-child and child-peer relationships.  It is reasonable to expect that children may exhibit a similar style of interaction with their parent and teacher, as both are characterized as hierarchical relationships in which the adult is instrumental in assisting the child in developing emotional and behavioral regulatory strategies (Pianta, 1997).  Most of this research has explored the quality of adult-child relationships from an attachment perspective.  In both preschool (Hamilton & Howes, 1992; Howes, Hamilton, & Matheson, 1994; Howes, Matheson, & Hamilton, 1994; Motti, 1986, cited in Pianta, 1997; Pianta, 1992) and early school-age samples (Cohn, 1990; Pianta, 1994; Toth & Cichetti, 1996), qualities indicative of insecure parent-child relationships (i.e., dependency, nurturance) were related to teacher’s ratings of teacher-child relationships and social competence with peers.  Their work supports a model in which mother-child relationships influence relationships children have with teachers, which in turn, affect children’s interactions with peers in the classroom (Pianta, 1997; Howes, Matheson et al., 1994).  However, it is important to note that "quality of relationship" was defined in terms of attachment security, rather than conflict.  Thus, although consistent with the idea that coercive processes may spread from mother-child to teacher-child to peer relationships, this pathway cannot be inferred from these results.  In an another study, involving low-income kindergartners, mother-child aggression was related to poor teacher-child relationship quality (Strassberg et al., 1992), although teacher-child relationship was assessed with only one item from a questionnaire.  We know of no other studies that have examined teacher-child conflict specifically in relation to parent-child conflict, either concurrently or longitudinally.

Other investigators have reported associations between qualities of the parent-child relationship and qualities of relationships with peers.  Parental power assertion and use of physical discipline have been shown to predict children’s use of power assertive strategies with peers (Putallaz, 1987) and children’s sociometric status with peers, with higher levels of parental coerciveness predicting more negative child outcomes (Bierman & Smoot, 1991; DeKovic & Janssens, 1992; MacKinnon-Lewis et al., 1994; Pettit, Dodge, & Brown, 1988; Strassberg et al., 1992; Strassberg, Dodge, Pettit, & Bates, 1994).  In a recent investigation, Crockenberg and Lourie (1996) reported positive associations between mothers’ and fathers’ use of coercion in conflict management and children’s report of use of manipulation with peers at age 6.


Several other researchers have shown higher rates of children’s aggression with others when their parents are rejecting and harsh (see Maccoby and Martin, 1983, for a review; Pettit et al., 1988) and more frequently use aggressive physical discipline strategies (Strassberg et al., 1994).  The strength of relations between parent-child and child-peer conflict range from modest to moderate, depending on the nature of the sample (e.g., stronger relations are found in families experiencing physical abuse, Fantuzzo et al., 1991; Strassberg et al., 1994).  Taken together, these results suggest that parent-child conflict has an effect on children’s other relationships outside the home.    

Recent research has explored the contributions of sibling relationships to other relationships in and out of the home (Brody et al., 1987; Erel et al., 1998).  Qualities of sibling relationships are often hypothesized to be more strongly related to peer relationships, because both are characterized as close-age relationships with roles that are less hierarchically-organized than those of adult-child relationships, and these dyads tend to engage in similar activities and functions (e.g., play; Shantz, 1987; Stocker & Dunn, 1990; Stormshak et al., 1996).  However, the few studies that have examined the relations between sibling conflict and conflict in peer relationships have generally shown mixed results, with modest correlations between aggression in sibling and peer interaction emerging at times (Abramovitch, Corter, Pepler, & Stanhope, 1986; Berndt & Bulleit, 1985; Mendelsohn, Aboud, & Lanthier, 1994; Stocker & Dunn, 1990).  Observed sibling conflict was shown to be modestly related to teacher reports of externalizing behaviors in early school-age children (Bank & Burraston, 1997).  In general, the theoretical reasoning linking sibling and peer conflict appears sound, but the empirical results have failed to show a consistent effect.  In comparison, little is known about how early sibling conflict relates to the quality of teacher-child dyadic relationships.  Again, we know of no other study that has examined conflict directly in these two settings.


When sibling relations are examined in conjunction with other family relationships, they are often less strongly related to child aggression or peer outcomes than interparental and/or parent-child conflict (Dunn, Slomkowski, Beardsall, & Rende, 1994; MacKinnon-Lewis et al., 1997).  There is some evidence that qualities of sibling relationships may be affected by other relationships in the family in a compensatory way, such that in some families where extreme interparental conflict is common or individual parent-child relations are problematic, positive and strong relationships may develop between siblings as a buffer against the atmosphere of family violence (Jenkins, 1992).  Thus, it is difficult to know whether sibling conflict adds further “skills” in conflictual behavior with others above and beyond the influence of parent-parent and parent-child relations in the preschool and early school-age years.  It is possible that sibling conflict would moderate (i.e., interact) with parent-child and/or interparental conflict as in Jenkins’ work with older children (1992) to either promote or protect against poor relationships, depending on the type and frequency of conflict experienced in other family relationships.

One recent study indicates that sibling aggression may contribute to peer relations above that accounted for by parent-child relationship quality.  In a study of 8 to 10 year old boys, MacKinnon-Lewis and associates (1997) found that significantly more variance in peer outcomes was explained when aggression between sibings was added as a predictor.  They described a pathway in which boys who had rejecting mothers were more aggressive in their sibling interactions, which in turn, increased the likelihood that they were more aggressive and rejected by peers.  These findings are supportive of an additive effect of sibling and parent-child conflict on children’s use of coercion with others.  However, the data were cross-sectional, and unlike the present study, parental conflict and conflict with teachers were not assessed.  It is not clear whether the addition of interparental conflict would add or interact with other indices of family conflict to predict children’s relationships at school.


The studies and results reviewed above could also be drawn upon to support another hypothesis -- that the accumulation of dyadic conflict across multiple family relationships may best predict the development of poor relationships with teachers and peers.  That is, conflict experienced in any one relationship may not place the child at risk for problems in school relationships.  But as the number of family relationships in which conflict occurs increases, the risk for conflict in relationships outside of the home may increase.  There is evidence that the occurrence of multiple risk factors in early childhood has a linear relationship with behavioral outcomes, especially in high-risk groups, such as families from socioeconomically disadvantaged backgrounds (Rutter, 1978; Shaw, Winslow, Owens, & Hood, 1998).  This cumulative risk hypothesis is not inconsistent with the models (i.e., social learning, family systems, attachment, and sociological) presented earlier.  Conflict in individual dyadic relationships could be conceptualized as one of many risk factors.  For this model to be supported, one would expect to see significant additive and/or interactive effects of conflict in relationships on outcomes, and some of the proceeding results are consistent with this interpretation (e.g., MacKinnon-Lewis et al, 1997; Erel et al., 1998).  However, the effects of these particular dyadic relationships (interparental, sibling, and parent-child) have not yet been examined together to test this hypothesis.

The present study investigates how relations among interparental conflict, parent-child conflict, and sibling conflict across the preschool years predict teacher-child and child-peer conflict at school entry.  We examine the question: Do patterns of intra-family conflict in the home increase the probability of children displaying conflictual relationships with teachers and peers at school?  From social-learning, family systems, attachment, and sociological theories, it is hypothesized that conflictual behavior experienced in dyadic family relationships (including parent-parent, parent-child, and sibling) would be at least modestly related to conflict in other settings.  Based on prior research that suggests more modest direct effects for sibling conflict (Erel et al., 1998), we hypothesize that direct relations involving interparental and parent-child conflict will be stronger than sibling effects.  We also expect that experiencing conflict in multiple family relationships will place children at greater risk for displaying conflict with teachers and peers at school than children who experience or observe conflict in only one family relationship.  Therefore, we are particularly interested in examining two- and three-way interactions with respect to children’s relationships with peers and teachers at school, given that it may require experiencing conflictual relationships in multiple relationships at home for such a behavior to generalize to the school setting.  We hypothesize that interparental and mother-child conflict uniquely contribute to the prediction of conflict in school relationships, but based on the work of Brody and colleagues (1987; 1992), MacKinnon (1989), and Jenkins (1992), sibling effects are anticipated only when entered as interaction terms with indices of parent-parent and parent-child conflict, especially for child-peer conflict (MacKinnon-Lewis et al., 1997).


Hypothesized relations will be tested with a sample of 117 boys from low-income families from ages 3-1/2 to 6.  Levels of conflict in five dyadic relationships were examined: parent-parent, parent-child, sibling, teacher-child, and child-peer.  This study seeks to improve upon previous research by using a multi-method, multi-informant approach to assess dyadic conflict across contexts and time in a sample of young children, and by examining direct, additive, and interactive relations among these variables using hierarchical regression analyses.   

Method

Study children and their families participated in data collection procedures when the boys were ages 3-1/2, 5, and 6.  Parents reported on conflict in their marital relationship1 at age 3-1/2 and parent-child relationships at age 5, and destructive conflict between siblings was observed in a one-hour play session.  Teacher reports of teacher-child and child-peer conflict were obtained when the child was age 6.

Subjects


 Subjects were recruited from Women, Infant, and Children Nutritional Supplement Program (WIC) clinics in the metropolitan Pittsburgh area (Shaw et al., 1998).  The WIC program provides financial and nutritional resources to low-income families.  Mothers with boys between 6 and 17 months of age and another sibling living at home were asked to participate in a longitudinal study examining the development of young children.  Three hundred and ten families composed the original sample.  In order to allow for more consistent representation of all measurements under examination, a subset of the original sample were selected for the present study using the following criteria.  All subjects for which there was a close-age sibling (i.e., between one year younger and four years older who were available for the age 5 sibling observation) and had available teacher data at age 6 comprised the subsample, resulting in 117 families meeting selection criteria.  This sibling age criteria was chosen so that siblings would more likely be a playmate of the target child and be at a similar developmental level, and less likely to be engaged in caretaking of the target child (as may happen with much older siblings).  Of these 117 sibling pairs, 17 involved siblings who were 1 year younger, 6 were the same age, 28 were 1 year older, 28 were 2 years older, 27 were three years older, and 11 were 4 years older.  Because of this decision, 63 sibling pairs for whom we collected data were not included in analyses (most were excluded because their sibling was too old or too young, n=36, or teacher data was not available, n=27).  The decision to use the selection criteria was supported theoretically by others’ research showing that more extreme age differences between siblings results in different behaviors and roles than those closer in age (Dunn, 1983).  It was also supported empirically by testing age differences in the frequency of destructive sibling interaction among siblings one year younger to four years older with an ANOVA, F (5, 175) = 1.80, ns.  There were no significant differences in the frequency of sibling conflict across the six sibling ages included in the subsample.  Additionally, there was no significant difference in the frequency of sibling conflict when the selected and unselected groups were examined, t = -.83, ns.  Taken together, these results suggest that the close-in-age criteria between sibling pairs was appropriate.

The subsample primarily consisted of Caucasian and African-American families (58% and 35%, respectively, 7% other).  Mothers’ and fathers’ average educational level was 12.9 years, and the mean per capita income at recruitment for the entire sample was $2,892 (approximately $11,616 per year for a family of four).   

Mean scores on all variables were compared for the subsample and those subjects who did not meet selection criteria.  Interparental conflict at age 3.5 was significantly higher for the selected group (X’s = 25.33 and 19.04, p < .05).  All other means were comparable across the selected and unselected groups.

Procedures 


Families participated in three waves of data collection when the boys were 3.5, 5, and 6 years of age.  The age 3.5 assessment consisted of one laboratory visit, in which mothers completed questionnaires and participated in several structured tasks with their child.  At the age 5 visit, each mother was interviewed in the home about her child’s behavior and her relationship with the child.  The target child and his sibling were videotaped playing with sets of toys, while the examiner and mother completed questionnaires in the same room.  The siblings were videotaped together for one hour and played with up to three sets of toys brought by the examiner.  Sibling conflict was coded from videotapes.  At age 6, data on the quality of the teacher-child relationship and the child’s relationship with peers were collected by mail.  Teacher reports were solicited after school had been in session for at least two months, so that teachers would be familiar with the child’s typical behavior.  Due to differing registration criteria in the various school districts, approximately half of the subsample were in kindergarten (51%), and the rest were in first grade (46%) or second grade (3%) when teacher data were collected.  Mean scores for outcome variables were compared for children in kindergarten versus first grade.  No significant differences were detected.

Of the 117 families who met the selection criteria, data were available for 109 subjects at the age 3.5 visit and 116 subjects at the age 5 visit.  Data for teacher-reported outcomes at age 6 were available for 117 children.  However, due to the varying availability of data for all subjects across the two-and-a-half years, sample sizes range from 98 to 117.

Measures

Self-Report

Hollingshead Four Factor Index of Social Status (Hollingshead, 1975).  Mothers provided information about their own and their partner’s educational level and employment at each assessment.  The Hollingshead index is widely used, and allows for calculation of socioeconomic status scores for both one- and two-parent families.  The index has been shown to be highly correlated with an occupation index designed by the National Opinion Research Center (r = .927).      


Conflict Tactics Scales (CTS-Form N) (Straus, 1979).   The CTS assesses the use of verbal reasoning, verbal aggression, and violence in the family.  This questionnaire was administered at the age 3.5 visit.  The CTS consists of 26 items which measure the frequency of conflict resolution tactics used by partners over the past year.  For each item, the subject also reports how often the child witnessed the behavior.  Although the CTS yields several factors, we chose to composite four factor scores (Frequency and Exposure to Verbal Aggression and Physical Aggression) by summing them, as both constructs are thought to be important to the development of coercive processes in families (Patterson et al., 1992; Fantuzzo et al., 1991).  The CTS is widely used, has been demonstrated to have adequate reliability and validity, and has been shown to be associated with both child behavior problems (Fantuzzo et al., 1991) and child responses to conflict (J.S. Cummings, Pellegrini, Notarius, & E.M. Cummings, 1989).  In large nationally represented samples, the alpha coefficients for the Verbal Aggression and Physical Aggression subscales were high (ranges from .77 - .88; Straus, 1979; Straus, 1991b).  

Student-Teacher Relationship Scale (STRS; Pianta & Steinberg, 1991) and Adult-Child Relationship Scale.  This 30-item questionnaire was originally designed to assess teacher’s perceptions of their relationship with children.  Attachment-related issues are ascertained, as well as their feelings about the child and his behavior.  We adapted it for use with mothers by slightly changing the wording of items to reflect the parent-child relationship, and used it at the age 5 home visit.  In addition, teachers completed the STRS as the children approached age 6.  For the present study, because of our interest in conflictual dyadic relationships, the “conflicted relationship” factor was utilized, comprised of eight items assessing the frequency of conflict on a 5-point Likert scale (e.g., This child and I seem to always struggling with one another, This child feels I am unfair to him).  Coefficient alpha for this measure in a sample of preschool children was .90 (Pianta & Steinberg, 1991).  In our sample, Cronbach’s alpha equaled .89 for teacher-reported data, and .81 for mother-reported data.


Teacher’s Checklist of Peer Relationships (Dodge, 1986).  This questionnaire consists of nine items which assess children’s sociometric status within the classroom, as well as the frequency of specific coercive child behaviors.   For the present study, the hostile-aggressive factor was used, comprised of six items that assess the child’s use of aversive and aggressive conflict management strategies with peers (e.g., “this child gets other kids to gang up on a peer he does not like”).  Responses to these items were shown to be internally consistent (coefficient alphas = .50 and .93, respectively) and reliable across teacher raters in a sample of school-age children (Dodge & Somberg, 1987).  For our sample, Cronbach’s alpha for this measure is .93.

Behavioral Observation


Sibling Conflict Coding System (SCCS, Garcia, 1998).  Sibling conflict constructs were coded from videotapes using the SCCS, an observational coding system for sibling and parent/child interaction based on a live coding system developed by Volling and Belsky (1992).  The system was designed to capture the amount and quality of sibling conflict during a one-hour sibling interaction.  Both molecular and global ratings were coded.  At the molecular level, episodes of conflict were recorded and then given an intensity rating of 1-5 depending on the length of the conflict, level of affect, and whether or not physical aggression was involved.  Global ratings were assessed on various dimensions (e.g., destructive conflict, negative reactivity of each sibling).  Because low levels of certain types of conflict (e.g., constructive conflict, negotiation) between siblings have been shown to be normative and perhaps adaptive (Dunn, 1983), but destructive and hostile aggression between siblings has been linked to maladaptive outcomes (Dunn & Brown, 1994), we used a measure of destructive sibling conflict in the present analyses.  In order to qualitatively capture destructive conflict between siblings and to strengthen the generalizability of the construct, codes were collapsed into one “destructive sibling conflict” score using the following method (Garcia, 1998).  From the molecular codes, scores for frequency (number of conflict sequences), intensity (average of all ratings), and number of seconds spent in the most intense/aggressive conflict (levels 4 and 5) were derived.  These three scores, plus the global rating for destructive conflict, were standardized and then summed to create a composite score, which was used in further analyses.  The correlations among these variables ranged from .71-.88 (p < .001), and the alpha coefficient for the composite was .89.  This composite reflects how many times the siblings engaged in high-intensity aggression and conflict, how prolonged the conflict was, and the level of negative affect of the conflict throughout the play interaction (Garcia, Shaw, Owens, & Yaggi, in press).  Again, because we wanted to focus on effects of destructive conflict in close-age sibling relationships, only those children with siblings between one year younger and four years older were used in analyses.  This measure of destructive sibling conflict has been shown to be correlated with maternal reports of children’s aggression and delinquency at ages 5 and 6 in the present sample (Garcia et al., in press).

Eighteen (or 10%) of the sibling interaction videotapes were coded by all coders in order to determine reliability.  Cohen’s kappa coefficient for the molecular codes across raters was .81.  Global ratings were assessed using 4-point scales.  For these ratings, percent agreement averaged .72.  Because these levels of agreement for the global ratings on reliability tapes were deemed unacceptable, a consensus procedure was utilized for the entire data set.  Approximately twenty minutes of each videotape were viewed by the coding team.  The original coder was required to justify the global ratings, which were then verified by consensus.  Consensus ratings were also used when raters had questions regarding unclear conflict sequences.  All four coders watched the questionable conflict and decided on the appropriate code.  This was done for approximately 1/4 of all codes for each tape (Garcia et el., in press).

Results

Results are presented in four stages.  Initially, descriptive statistics are presented for the independent and dependent measures.  Second, direct correlations among indices of family conflict and school-based conflict are examined.  Third, additive and interactive effects of intra-family conflict were explored in relation to children’s conflictual relationships with teacher and peers by computing two, seven-step hierarchical regressions.

Descriptive Statistics for Study Measures


Descriptive statistics for all independent and dependent variables are presented in Table 1.  Subjects’ scores differed from the normative samples on a few measures that have been used in prior research.  The average scores for the frequency of Verbal Aggression and Physical Aggression factors on the CTS at age 3.5 were at the 80th and 90th percentiles, respectively.  Mean socioeconomic status was low, representative of semi-skilled workers (Hollingshead, 1975).  Normative data on other measures of conflict were unavailable, so it is uncertain whether subjects’ scores on other measures fell within normative ranges, although examination of the mean scores displayed normal distribution.

___________________________________

 Insert Table 1 about here

___________________________________

Relations Among Measures of Intra-Family and Child Problems

Pearson Correlation Coefficients were used to test the hypothesis that interparental conflict, parent-child conflict, and sibling conflict were positively associated with teacher-child and child-peer conflict.  Table 2 summarizes these results.  Correlations presented were computed using all subjects who met the selection criteria and for whom data were available.  Relations between outcome variables and SES also are presented in Table 2 to test for “third variable” effects.  SES was not significantly correlated with teacher-child or child-peer conflict scores.  Thus, SES was not used in further analyses.  

____________________________________

 Insert Table 2 about here

____________________________________


In general, correlations across intra-family conflict factors and among informants were positive, but modest in size.  Mothers’ ratings of mother-child conflict exhibited the strongest correlations with other predictors, ranging from .13 (with age 5 sibling conflict) to .26 (with age 3.5 interparental conflict).  Sibling conflict was not directly related to interparental conflict.  The correlation between teacher-child conflict and child-peer conflict was high (r=.73).  Although the high correlation between these two teacher-rated measures may reflect an informant bias, due to the theoretical difference designated in the literature between these two constructs and the potentially different effects of intra-family conflict on these two relationships, these scores were treated as independent outcomes and remained separate in further analyses.

Longitudinal correlations between predictor and outcome variables were generally positive, and in 5 of 6 cases, statistically significant at at least a trend level.  As hypothesized, earlier interparental conflict and mother-child conflict exhibited the highest and most consistent relations with teacher-child and child-peer conflict (ranging from .13 to .31).  Sibling conflict was not significantly related to teacher-child conflict, and only approached significance with child-peer conflict.  Overall, the data suggest that interparental conflict and mother-child conflict demonstrate modest, but consistent, links to teacher problems rated 1 to 4 years later.

Additive and Interactive Effects of Intra-Family Conflict on Teacher-Rated Outcomes at Age 6

In order to test the effects of intra-family conflict on relationships at school, additive and interactive effects were assessed for teacher-reported outcomes.  We computed two identical, seven-step hierarchical regression equations.  As some of these subjects were missing family conflict data, the final sample sizes for the two regressions were 98 and 106 subjects. 


Predictor variables for regressions included interparental conflict, mother-child conflict, and sibling conflict.  Dependent measures were the two teacher-rated outcome measures: teacher-child conflict and child-peer problems.  All predictors were initially centered to reduce multicollinearity that may occur with product terms.  Predictors were entered in the following order.  Main effects were entered in chronological order:  age 3.5 interparental conflict, age 5 mother-child conflict, and age 5 sibling conflict.  Next, the two-way interaction terms were entered in chronological order (interparental conflict X mother-child conflict, interparental conflict X sibling conflict, and mother-child conflict X sibling conflict).  Lastly, the three-way interaction term (interparental X mother-child X sibling conflict) was entered.  Results are displayed in Table 3.  As predicted, for teacher-child conflict, mother-child conflict entered individually at step two accounted for a moderate portion of the variance (R2 change = .04, p < .05), although the predicted main effect for interparental conflict was not attained.  The two-way interaction of verbal/physical aggression at age 3.5 and sibling conflict also predicted additional significant variance (R2 change = .03, p < .05).  Furthermore, the three-way interaction of interparental, mother-child, and sibling conflict accounted for unique variance (R2 change = .03, p < .05). 

____________________________________

Insert Table 3 about here

____________________________________

To examine the pattern of interactions, follow-up analyses of simple slopes were conducted using the method outlined by Aiken and West (1991).  To examine the two-way interaction between interparental conflict and sibling conflict, controlling for other predictors listed in Table 3, teacher-child conflict was regressed on interparental conflict at age 3.5, at 1/2 standard deviation above and below the mean of sibling conflict at age 5.  These simple slopes were then plotted at 1/2 standard deviation above and 1/2 standard deviation below the mean for interparental conflict at age 3.5 (See Figure 2).  The 1/2 standard deviation cut-point was chosen in order to represent patterns of interaction comparable to a reasonable number of subjects in each “group” (e.g., at least 10) if one were to compute a group-based ANOVA test of interaction effects.  A test of the simple slopes demonstrated that there was a trend for teacher-child conflict scores to be low for those children who experienced low sibling conflict at age 5, t=1.76, p < .10.  At high levels of sibling conflict, changes in teacher-child conflict scores were nonsignificant, t=.53, ns.  The size of the difference in teacher-child conflict scores between low and high parental and sibling conflict was about 1/3 of a standard deviation.

_____________________________________

Insert Figure 1 about here

_____________________________________


We then examined the significant three-way interaction using simple slope analyses.  Again, 1/2 standard deviation above and below mean cut-points on predictor variables were used in order to insure that the patterns depicted would be representative of enough subjects for appropriate interpretation of effects.  Although the use of 106 and 98 subjects in the respective regressions is above the recommended guidelines (a minimum of 10-subjects to each step; Kleinbaum, Kupper, & Muller, 1988), caution is still warranted in interpreting these effects, as three-way interactions in a relatively small sample can be somewhat unstable. 

The patterns of the simple slopes analyses are displayed in Figure 2.  Only one of the simple slopes was significant, although others were generally in the expected direction.  For children experiencing high mother-child conflict and low sibling conflict, as interparental conflict at age 3.5 increased, teacher-child conflict scores increased (t = 2.47, p < .01).  For this configuration of conflict factors, differences in teacher-conflict scores across varying levels of interparental conflict equaled approximately 1/2 standard deviations.  Other slopes were not significantly different from zero.  While the interaction effects are not large, the results generally support our hypotheses.  The three-way interaction primarily shows an amplification effect of what was demonstrated by the main effect for mother-child conflict and the two-way effect for interparental and sibling conflict; that is, that experiencing mother-child conflict was the strongest risk factor for poor teacher-child conflict (3 out of 4 configurations in the regression equations involving high mother-child conflict were elevated), and sibling and interparental conflict interacted such that low levels of both generally predicted low teacher-child conflict.  The patterns support the cumulative risk hypothesis as well.  If high levels of conflict were experienced in more than one relationship (e.g., mother-child and sibling), risk was greater for having a conflictual relationship with the teacher at age 6.  It is significant that the lowest teacher-child conflict score is associated with the pattern of low conflict experienced across all three intra-family dyadic relationships, and the highest teacher-child conflict is associated with experiencing high conflict in all three relationships. 


___________________________________

Insert Figure 2 about here

____________________________________ 

The most robust findings were in predicting teacher-rated peer problems at age 6.  Using the same method of analysis, two of the main effects provided significant variance to this outcome: parental conflict at age 3.5 ( R2 change = .09, p < .01), and mother-child conflict at age 5 (R2 change = .07, p < .05).  As with teacher-child conflict, the three-way interaction also contributed additional variance (R2 change = .05, p < .05).  Moreover, when examined with simple slopes analyses (not presented here), the pattern of the three-way interaction was similar to that of teacher-child conflict.  That is, for children experiencing high levels of mother-child conflict and lower levels of sibling conflict, child-peer conflict scores significantly increased as interparental conflict increased (t = 3.70, p < .01).  Other slopes were not significantly different from zero.  As with teacher-child conflict, overall, when conflict was experienced in more than one family relationship, child-peer conflict scores were higher than for those subjects who experienced conflict in one or no other relationships.  Of note is that the combination of high mother-child and high parental conflict appears to leave boys especially vulnerable to having trouble with peers, with sibling conflict demonstrating significant, but more muted effects. 

Discussion


This study examined the continuity of conflictual processes in relationships across family and school environments in a sample of low-income boys.  Several researchers had previously documented relations between conflict in dyadic relationships within families.  The development and potential “spread” of family conflict to school relationships has been posited by several models, but rarely tested longitudinally in early childhood.  In the present study, we were interested in whether patterns of intra-family conflict affected the quality of relationships children have with teachers and peers at school.  Based on previous work, we had hypothesized that mother-child and interparental conflict would show strong direct and interactive effects on teacher-child and child-peer relationships, but predicted that sibling conflict would have mixed effects (i.e., sibling relationships might contribute more to child-peer than teacher-child conflict, and might show a protective effect when conflict was high in other relationships).  We also thought that intra-family conflict might exhibit multiplicative effects, such that experiencing conflict in more than one relationship would increase vulnerability to poor relationships in school. 

This study both affirms and extends previous research by demonstrating that patterns of intra-family conflict have implications for children’s relationships at school.  Overall, the findings were consistent with our hypotheses that early conflictual relationships within the family during infancy and preschool years are modestly related to later conflictual relationships in school, mirroring the results of other investigations (MacKinnon et al., 1994; Parke & Ladd, 1992; Parke et al., 1989).  The results are consistent with the tenets of social learning, family systems, attachment, and sociological theorists regarding the development of conflict within early family relationships (Brody et al., 1987; Greenberg & Speltz, 1988; Greenberg, Speltz, & DeKlyen, 1993; Hotaling et al., 1989; Patterson et al., 1992).  However, experiencing conflict in any one particular relationship does not appear to greatly enhance risk for conflict in later relationships (i.e., direct correlations were low).  Time of measurement may be a factor in the relatively weak correlations; there was a two-and-a-half year lapse between our assessments of interparental conflict and teacher-rated outcomes, and a one-and-a-half lapse between interparental conflict and other within-family dyadic conflict measurements, while sibling and mother-child conflict were measured concurrently.


Our analyses indicated complex and different patterns of additive and interactive effects for teacher-child and child-peer conflict.  Although the interactive findings must be interpreted cautiously, the patterns were especially consistent with a cumulative risk model; that is, the experience of conflict in multiple dyadic family relationships appeared to put children at greater risk for poor teacher-child and child-peer relationships.  When conflict was experienced in more than one family relationship, risk for poor teacher-child and child-peer relationships was generally increased by about 1/3 to 1/2 standard deviations.  Results from these analyses advance the notion that intra-family conflict has negative effects on children’s relationships with teachers.  Conflictual relationships with teachers at school entry may have important consequences for children’s later adjustment, as children with problematic relationships with teachers have been shown to be at higher risk for academic problems, school failure, and peer rejection (see Pianta, 1992, for a review; Pianta, Steinberg, & Rollins, 1995).

As reflected by both the direct correlations and the two significant three-way interactions, mother-child conflict exerted more of a direct effect on child outcomes than other forms of intra-family conflict.  Main effects accounting for a moderate portion of the variance were exhibited in predicting both teacher-child and child-peer conflict.  This is supported by previous literature on parent-child relations, in which the quality of the mother-child relationship has been found to play a significant role in children’s socialization (Maccoby, 1992, 1996; MacKinnon-Lewis et al., 1994).   


Although the interactive findings demonstrate that high levels of sibling conflict in the presence of other conflictual relationships can amplify risk for poor school relationships, sibling conflict was not directly related to other forms of conflict (except weakly to mother-child conflict).  This is not necessarily surprising, as earlier studies have shown mixed results.  The lack of findings may be due to one or more of several factors2.  It may be that sibling conflict is quite common among children around school entry age, and so the presence of conflict in the sibling relationship may not strongly differentiate effects on children’s behavior with others, even when conflict is highly aggressive.  Alternatively, in homes in which family relationships are characterized by coercion and conflict, perhaps siblings restricted or controlled their behavior, because they have learned that they need to “behave” in front of others when they are home (i.e., they may be hypervigilant and over-controlled due to the increased risk for parental anger and violence).  Thus, the observation would not necessarily capture the “true” sibling interaction.  The significant three-way interaction effects suggesting that it is at lower levels of sibling conflict and high levels of mother-child conflict that the increase in interparental conflict predicts greater peer- and teacher-child conflict is consistent with this conclusion -- that is, when children are out of their home environment, perhaps they no longer feel the need to keep tight control over their conflict strategies and will thus exhibit their “true” coercive and/or aggressive tendencies in the school environment.

Relatedly, perhaps the measurement strategy (i.e., by observation) was not effective in capturing the enduring quality of the sibling relationship.  For example, Brody and colleagues (Brody et al., 1987, 1992; Brody, Stoneman, & McCoy, 1994) found direct relations between sibling- and parent-reported sibling conflict and interparental conflict, but when observed sibling conflict was considered, relations were weak or nonsignificant.  They point to two potential reasons.  First, reported data may allow for a better measurement of the subtle, more affective quality of relationships between siblings.  Second, reported data may better reflect the long-term history of the quality of the relationship than behaviors observed in a one-hour observation.  Our other measures of intrafamily conflict were self-reported data.  Had we also obtained self-report data on sibling conflict, we may have found direct associations with sibling relationship quality to be stronger.  Nonetheless, even though the results need to be interpreted cautiously, significant interactive effects with sibling conflict were found.  This leads us to believe that our observational measure of sibling conflict provides informative data about conflict in family relationships.  These results and MacKinnon-Lewis’ et al. findings (1997) highlight the importance of the sibling relationship as a context for learning and practicing conflict strategies.  Furthermore, the results support Patterson’s hypothesis that sibling conflict contributes to coercive behavior in the school setting (Parke et al., 1989; Patterson, 1986; Patterson et al., 1992).

The Development of Conflictual Processes: Links Across Family and School


Our data are supportive of a developmental pathway in which interparental conflict provides a backdrop of emotionally charged negative interactions, which have lasting effects on children’s beliefs and learning about power and control in interpersonal relationships.  One of the most important tasks in the preschool years is to learn how to regulate emotions and to control behavior, and dyadic relationships within the family are thought to play an important role in facilitating and teaching these skills (Thompson & Calkins, 1996).  For children in maritally-stressed and/or violent homes, this early interparental conflict may have an impact on the development of parent-child relationships, perhaps by disrupting parenting practices, as suggested by Fauber, Forehand, Thomas, and Wierson (1990), or by affording less exposure to positive conflict resolution strategies (Davies & Cummings, 1994).  Both aversive interparental conflict and parent-child coerciveness affect the quality of sibling relationships, as children may practice the coercive strategies witnessed and/or experienced in other family relationships (Patterson, 1984).  By age 5, children who have a lot of experience with aversive conflict may have poorly developed affective/emotional and behavioral control, or may have learned that coercive conflict strategies are effective with others (Patterson et al., 1992).  Both of these mechanisms could result in the early school-age child developing a “style” of interacting with others that is characterized by hostile, aversive/aggressive behavior.  In turn, as the child then calls upon these negative strategies with peers or teachers in the classroom, the likelihood of being rejected or disliked increases.  Over time, academic problems and poor peer relationships place these children at high risk for a variety of maladaptive outcomes.  Finally, although not examined in the present study, problems with teachers and peers in school may then lead to increased risk for continuing problems in relationships in the home during middle childhood and adolescence.     


While the present data are consistent with the proposed pathway, there are some issues which limit our interpretation of the effects.  First, although associations were evident over time and the results support the idea of conflict synergystically “accumulating” over relationships, the analyses were correlational in nature.  We did not measure each construct at every time point (i.e., a cross-lag design), and thus, cannot determine the temporal relations and the order of the emergence of conflict in these dyads.  There may be alternate directions of effects (e.g., parent-child conflict leading to interparental conflict) than those explored in the present design.  Second, father-child relationships have also been found to be important to the development of children’s social behavior, and to be affected by intra-family conflict (DeKlyen, Speltz, & Greenberg, 1998; Coiro & Emery, 1998; MacKinnon-Lewis et al., 1994).  Unfortunately, due to the relatively small number of fathers living in the home in our selected sample, we were unable to examine the potential effects of father-child conflict on the development of conflict processes. 

Moreover, we did not directly measure intra-family coercive processes at the microsocial level described by Patterson and others (1992), although our results are generally consistent with his early starter model.  Fourth, we chose to focus on the effects of dyadic conflict in these relationships for theoretical reasons, but there may be effects of other constructs (e.g., temperament, maternal depression) on these predictors and outcomes that were not considered.  For example, several studies suggest that sibling relationship quality is affected by the age, gender, and temperamental style of the siblings (Dunn, 1983; Brody et al., 1987; Brody et al., 1994; Brody, Stoneman, & Gauger, 1996; Stoneman, Brody, & Burke, 1989).  Although we chose to examine close-in-age sibling pairs for comparability to previous studies and to limit the effects of sibling age differences on our outcomes, examining the quality of relationships among multiple sibling relationships within the family would likely yield important information relevant to the spread of conflict to school.  Fifth, due to sample characteristics (i.e., low-income boys), the results may be of limited generalizability to other populations.  While relations were comparable with other findings for young boys (MacKinnon-Lewis et al., 1994; Patterson et al., 1992; Strassberg et al., 1994), there is some evidence that pathways to antisocial behavior may be different for girls (Crockenberg & Lourie, 1996; Osborne & Fincham, 1996; Patterson et al., 1992).  Lastly, because many of our measures were self-report questionnaire data, there is the possibility that the correlations between these constructs are inflated due to shared-method variance.  However, the possibility of shared method effects is tempered by the longitudinal design and by the use of different reporters and scales. 


Despite these limitations, this study represents an important first step in exploring the longitudinal associations among these five types of dyadic conflict.  Several questions remain regarding the development and conflictual and coercive processes in young children.  For example, it is likely that while some characteristics of and experiences in specific relationships may facilitate risk, others may protect against poor outcomes.  Children may compensate for poor relationships by seeking out and creating positive relationships with others (e.g., grandparents, daycare workers; Jenkins, 1992; Rutter, 1990; Stolba & Amato, 1993).  The present data did not allow for a clear interpretation of potential protective effects across all of the relationships children experience in early childhood.  Future research will hopefully shed light on the effects of “good” relationships in the context of multiple negative relationships.  Lastly, although the present study supported the hypothesis that as conflict experiences accumulated across relationships, risk for problematic school relationship increased, we did not examine how these experiences contribute to the outcomes.  We would encourage other researchers to engage in more detailed analyses of the precise mechanisms by which these processes develop and spread into other contexts.

 

 

References

Abramovitch, R., Corter, C., Pepler, D., & Stanhope, L.  (1986).  Sibling and peer interaction: A final follow-up and a comparison.  Child Development, 57, 217-229.

Aiken, L.S, & West, S. G.  (1991).  Multiple regression: Testing and interpreting interactions.  Newbury Park, CA: Sage Publications.

Bank, L. & Burraston, B. O.  (1997).  Parent and sibling indicators of adolescent adjustment outcomes.  Paper presented at the meeting of the Society for Research in Child Development, April, Washington, D.C.

Berndt, T. & Bulleit, T.  (1985).  Effects of sibling relationships on preschoolers’ behavior at home and at school.  Developmental Psychology, 22, 640-648.

Bierman, K. L. & Smoot, D. L.  (1991).  Linking family characteristics with poor peer relations: The mediating role of conduct problems.  Journal of Abnormal Child Psychology, 19, 341-356.

 Brody, G. H., Stoneman, Z., & Burke, M.  (1987).  Family systems and individual child correlates of sibling behavior.  American Journal of Orthopsychiatry, 57, 561-569.

Brody, G. H., Stoneman, Z., & Gauger, K.  (1996).  Parent-child relationships, family problem-solving behavior, and sibling relationship quality: The moderating role of sibling temperaments.  Child Development, 67, 1289-1300.

Brody, G. H., Stoneman, Z., & McCoy, J. K. (1994).  Contributions of family relationships and child temperaments to longitudinal variations in sibling relationship quality and sibling relationship styles.  Journal of Family Psychology, 8, 274-286.

Brody, G. H., Stoneman, Z., McCoy, J. K., & Forehand, R.  (1992).  Contemporaneous and longitudinal associations of sibling conflict with family relationship assessments and family discussion about sibling problems.  Child Development, 63, 391-400.

 

Cichetti, D.  (1984).  The emergence of developmental psychopathology.  Child Development, 55, 1-8.

Cohn, D.  (1990).  Child-mother attachment of six-year-olds and social competence at school.  Child Development, 61, 152-162.

Coiro, M. J. & Emery, R. E. (1998).  Do marriage problems affect fathering more than mothering?  A quantitative and qualitative review.  Clinical Child and Family Psychology Review, 1, 23-40.

Crockenberg, S. & Lourie, A.  (1996).  Parents’ conflict strategies with children and children’s conflict strategies with peers.  Merrill-Palmer Quarterly, 42, 495-518.

Cummings, J.S., Pelligrini, D.S., Notarius, C.I., & Cummings, E.M.  (1989).  Children’s responses to angry adult behavior as a function of marital distress and history of interparent hostility.  Child Development, 60, 1035-1043.

Davies, P. T. & Cummings, E. M. (1994).  Marital conflict and child adjustment: An emotional security hypothesis.  Psychological Bulletin, 116, 387-411.

DeKlyen, M., Speltz, M. L., & Greenberg, M. T.  (1998).  Fathering and early onset conduct problems: Positive and negative parenting, father-son attachment, and the marital context.  Clinical Child and Family Psychology Review, 1, 3-22.

DeKovic, M. & Janssens, J.  (1992).  Parents’ child-rearing style and child’s sociometric status.  Developmental Psychology, 28, 925-932.

Deutsch, (1973).  The resolution of conflict.  New Haven: Yale University Press.

Dodge, K. A.  (1986).  A social information processing model of social competence in children.  In M. Perlmutter (Ed.), Eighteenth Annual Minnesota Symposium of Child Psychology (pp. 77-125).  Hillsdale, NJ: Erlbaum.

Dodge, K. A. & Somberg, D. R.  (1987).  Hostile attributional biases among aggressive boys are exacerbated under conditions of threats to the self.  Child Development, 58, 213-224.

 

Dunn, J.  (1983).  Sibling relationships in early childhood.  Child Development, 54, 787-811.

Dunn, J., & Brown, J. R.  (1994).  Affect expression in the family, children’s understanding of emotions, and their interactions with others.  Merrill-Palmer Quarterly, 40, 120-137.

Dunn, J., Slomkowski, C., Beardsall, L., & Rende, R.  (1994).  Adjustment in middle childhood and early adolescence: Links with earlier and contemporary sibling relationships.  Journal of Child Psychology and Psychiatry, 35, 491-504.

Erel, O., & Burman, B.  (1995).  Interrelatedness of marital relations and parent-child relations: A meta-analytic review.  Psychological Bulletin, 118, 108-132.

Erel, O., Margolin, G., & John, R. S.  (1998).  Observed sibling interaction: Links with the marital and mother-child relationship.  Developmental Psychology, 34, 288-298.

Fantuzzo, J.W., DePaola, L.M., Lambert, L., Martino, T., Anderson, G., & Sutton, S.  (1991). Effects of interparental violence on the psychological adjustment and competencies of young children.  Journal of Consulting and Clinical Psychology, 39, 258-265.

Fauber, R., Forehand, R., Thomas, A. M., & Wierson, M.  (1990).  A mediational model of the impact of marital conflict on adolescent adjustment in intact and divorced families: The role of disrupted parenting.  Child Development, 61, 1112-1123.

Garcia, M. M.  (1998).  Sibling conflict and the development of behavior problems.  Unpublished master’s thesis, University of Pittsburgh, Pittsburgh, PA.

Garcia, M. M., Shaw, D. S., Owens, E. B., & Yaggi, K. E.  (In press).  Destructive sibling conflict and the development of conduct problems in young boys.  Developmental Psychology.

Gottman, J. M., & Katz, L. F.  (1989).  Effects of marital discord on young children’s peer interaction and health.  Developmental Psychology, 25, 373-381.

Greenberg, M. T.  & Speltz, M. L.  (1988).  Attachment and the ontogeny of conduct problems.  In J. Belsky & T. Nezworski (Eds.), Clinical implications of attachment (pp. 177-218).  Hillsdale, NJ: Erlbaum.

 

Greenberg, M. T., Speltz, M. L., & DeKlyen, M.  (1993).  The role of attachment in the early development of disruptive behavior problems.  Development and Psychopathology, 5, 191-213.

Hamilton, C. E. & Howes, C.   (1992).  A comparison of young children’s relationships with mothers and teachers.  In R. C. Pianta (Ed.), Relationships between children and non-parental adults: New directions in child development  (pp. 41-60).  San Francisco: Jossey-Bass.

Holden, G. W.  & Ritchie, K. L.  (1991).  Linking extreme marital discord, child rearing, and child behavior problems: Evidence from battered women.  Child Development, 55, 794-327.

Hollingshead, A.B.  (1975).  Four Factor Index of Social Status.  Unpublished manuscript, Yale University, Department of Sociology, New Haven.

Hotaling, G. T., Straus, M. A., & Lincoln, A. J.  (1989).  In L. E. Ohlin & M. H. Tonry (Eds.), Family Violence (pp 315-375).  Chicago: University of Chicago Press.

Howes, C., Hamilton, C. E.,  & Matheson, C. C.  (1994).  Children’s relationships with peers: Differential associations with aspects of the teacher-child relationship.  Child Development, 65, 253-263.

Howes, C., Matheson, C. C., & Hamilton, C. E.  (1994).  Maternal, teacher, and childcare history correlates of children’s relationships with peers.  Child Development, 65, 264-273.

Jenkins, J.  (1992).  Sibling relationships in disharmonious homes: Potential difficulties and protective effects.  In F. Boer & J. Dunn (Eds.), Children’s sibling relationship: Developmental and clinical issues (pp. 125-138).  Hillsdale, NJ: Erlbaum.

Jouriles, E.N., Murphy, C.M., Farris, A.N., Smith, D.A., Richters, J.E., & Waters, E.  (1991).  Marital adjustment, parental disagreements about child rearing, and behavior problems in boys: Increasing the specificity of the marital assessment.  Child Development, 62, 1424-1433.

Kleinbaum, D. G., Kupper, L. L., & Muller, K. E.  (1988).  Applied regression analysis and other multivariable methods.  Belmont, CA: Duxbury Press.

Loeber, R. & Dishion, T. J.  (1983).  Early predictors of male delinquency: A review.  Psychological Bulletin, 94, 68-99.

 

Maccoby, E. E.  (1992).  The role of parents in the socialization of children: An historical overview.  Developmental Psychology, 28, 1006-1017.

Maccoby, E. E.  (1996).  Peer conflict and intrafamily conflict: Are there conceptual bridges?  Merrill-Palmer Quarterly, 42, 165-176.

Maccoby, E. E., & Martin, J. A.  (1983).  Socialization in the context of the family: Parent-child interaction.  In P.H. Mussen (Ed.), Handbook of child psychology, Vol. IV, Socialization, personality, and social development (pp. 1-102).  New York: Wiley.

MacKinnon, C. E.  (1989).  An observational investigation of sibling interactions in married and divorced families.  Developmental Psychology, 25, 36-44.

MacKinnon-Lewis, C., Starnes, R., Volling, B., & Johnson, S.  (1997).  Perceptions of parenting as predictors of boys’s sibling and peer relations.  Developmental Psychology, 33, 1024-1031.

MacKinnon-Lewis, C., Volling, B. L., Lamb, M. E., Dechman, K., Rabiner, D., & Curtner, M. E.  (1994).  A cross-contextual analysis of boys’ social competence: From family to school.  Developmental Psychology, 30, 325-333.

Mendelsohn, M. J., Aboud, F. E., & Lanthier, R. P.  (1994).  Kindergartners’ relationships with siblings, peers, and friends.  Merrill-Palmer Quarterly, 40, 416-435.

Minuchin, S.  (1974).  Families and family therapy.  Cambridge, MA: Harvard University Press.

Motti, F.  (1986).  Relationships of preschool teachers with children of varying developmental histories.  Unpublished doctoral dissertation, University of Minnesota.

Osborne, L. N. & Fincham, F. D.  (1996).  Marital conflict, parent-child relationships, and child adjustment: Does gender matter?  Merrill-Palmer Quarterly, 42, 48-75.

Parke, R. D. & Ladd, G. W.  (1992).  Family-peer relationships: Modes of linkage.  Hillsdale, NJ: Erlbaum.

 

Parke, R. D, MacDonald, K., Burks, V., Carson, J., Bhavnagri, N, Barth, J., & Beitel, A.  (1989).  Family and peer systems: In search of the linkages.  In K. Kreppner & R. Lerner (Eds.), Family systems and life-span develpment (pp. 65-92).  Hillsdale, NJ: Erlbaum.

Patterson, G. R.  (1982).  Coercive family processes.  Eugene, OR: Castalia.

Patterson, G. R.  (1984).  Siblings: Fellow travelers in coercive family processes.  In R.J. Blanchard (Ed.), Advances in the study of aggression (pp. 174-213).  New York: Academic Press.

Patterson, G. R.  (1986).  The contribution of siblings to training for fighting: A microsocial analysis.  In D. Olweus, J. Block, M. Radke-Yarrow (Eds.), Development of antisocial and prosocial behavior: Research, theories, and issues (pp. 235-261).  New York: Academic Press.

Patterson, G. R., Dishion, T. J., & Bank, L.  (1984).  Family interaction: A process model of deviancy training.  Aggressive Behavior, 10, 253-267.

Patterson, G. R., Reid, J. B., & Dishion, T. J.  (1992).  Antisocial boys.  Eugene, OR: Castalia.

Pettit, G., Dodge, K, & Brown, M.  (1988).  Early family experience, social problem solving patterns, and children’s social competence.  Child Development, 59, 107-120.

Pianta, R. C.  (1992).  Beyond the parent: The role of other adults in children’s lives.  New Directions in Child Development, 57.  San Francisco: Jossey-Bass.

Pianta, R. C.  (1994).  Patterns of relationships between children and kindergarten teachers.  Journal of School Psychology, 32, 15-31. 

Pianta, R. C.  (1997).  Adult-child relationship processes and early schooling.  Early Education and Development, 8, 11-26.

Pianta, R. C., & Steinberg, M.  (1991).  Relationship between children and kindergarten teachers: Associations with home and classroom behavior.  Paper presented at the meeting for the Society for Research in Child Development, Seattle, WA.

Pianta, R. C., Steinberg, M., & Rollins, K.  (1995).  The first two years of school: teacher-child relationships and deflections in children’s classroom adjustment.  Development and Psychopathology, 7, 295-312.

 

Putallaz, M.  (1987).  Maternal behavior and children’s sociometric status.  Child Development, 58, 324-240.

Rutter, M.  (1978).  Family, area, and school influences in the genesis of conduct disorders.  In L.A. Hersov & D. Schaffer (Eds.), Aggression and antisocial behavior in childhood and adolescence (pp. 95-114).  Oxford: Pergamon Press.

Rutter, M.  (1990).  Psychosocial resilience and protective mechanisms.   In R. J. Rolf, A. S. Masten, D. Cichetti, K. H. Neuchterlein, & S. Weintraub (Eds.), Risk and protective factors in the development of psychopathology (pp. 181-214).  Cambridge, England: Cambridge University Press.

Shantz, C. U.  (1987).  Conflicts between children.  Child Development, 58, 283-305.

Shaw, D. S., Winslow, E. B., Owens, E. B., & Hood, N.  (1998).  Young children’s adjustment to chronic family adversity: A longitudinal study of low-income families.  Journal of the American Academy of Child and Adolescent Psychiatry, 37, 545-553.

Sroufe, L. A.   (1983).  Infant-caregiver attachment and patterns of adaptation in preschool: The roots of maladaption and competence.  In M. Perlmutter (Ed.), Minnesota Symposium on Child Psychology, Vol. 16.  Hillsdale, NJ: Erlbaum.

Stocker, C. & Dunn, J.  (1990).  Sibling relationships in childhood: Links with friendships and peer relationships.  British Journal of Developmental Psychology, 8, 227-244.

Stolba, A. & Amato, P. R.  (1993).  Extended single-parent households and children’s behavior.  Sociological Quarterly, 34, 543-549.

Stoneman, Z., Brody, G. H., & Burke, M.  (1989).  Sibling temperaments and maternal and paternal perceptions of marital, family, and personal functioning.  Journal of Marriage and the Family, 41, 75-88.

Stormshack, E. A., Bellanti, C. J., Bierman, K. L., and the Conduct Problems Prevention Research Group  (1996).  The quality of sibling relationships and the development of social competence and behavioral control in aggressive children.  Developmental Psychology, 32, 79-89.

 

Strassberg, Z., Dodge, K. A., Bates, J. E., & Pettit, G .S.  (1992).  The longitudinal relation between parent conflict strategies and children’s sociometric standing in kindergarten.  Merrill-Palmer Quarterly, 38, 477-493.

 Strassberg, Z., Dodge, K. A., Pettit, G. S., & Bates, J. E.  (1994).  Spanking in the home and children’s subsequent aggression toward kindergarten peers.  Development and Psychopathology, 6, 445-461. 

Straus, M. A.  (1979).  Measuring intrafamily conflict and violence: The conflict tactics scales. Journal of Marriage and the Family, 75-88.

Straus, M. A.  (1991a).  Discipline and deviance: Physical punishment of children and violence and other crime in adulthood.  Social Problems, 38, 133-152.

Straus, M. A.  (1991b).  Family violence in american families: Incidence rates, causes, and trends.  In D.D. Knudsen & J.L. Miller (Eds).  Abused and battered: Social and legal responses of family violence (pp.17-34).  New York: Aldine de Gruyter.

Thompson, R. A., & Calkins, S. D.  (1996).  The double-edged sword: Emotional regulation for children at risk.  Development and Psychopathology, 8, 163-182.

Toth, S. & Cichetti, D. (1996).  The impact of relatedness with mother on school functioning.  Journal of School Psychology, 34, 247-266.

Volling, B. L.  & Belsky, J.  (1992).  The contribution of mother-child and father-child relationships to the quality of sibling interaction: A longitudinal study.  Child Development, 63, 1209-1222.

 

Footnotes

1 Although the term interparental typically refers to mother-father dyads, the term as used in the present study reflects conflict between the mother and the caregiver whom she identified as most involved in caretaking of the target child.  Many of the families in this sample show diverse family structure and caretaking arrangements.  Within the subsample examined here, at the age 2 assessment, caregivers were represented by husbands (56%), boyfriends (33%), grandmothers (8%), aunts/uncles (1%), parents’ friends (1%), and others (1%).

 

2 The authors are grateful to the anonymous reviewers for their ideas regarding possible interpretations of sibling effects.

 

 


 

Table 1

Mean Scores of Independent and Dependent Variables

 

Age of Child

 

Variable                                     Age 3.5                                   Age 5                               Age 6

____________________________________________________________________________________

 

M         (SD)       n              M      (SD)      n             M     (SD)      n

 

Socioeconomic Statusa                3.87      (.93)       116                                    

   

Interparental Conflictb                 24.58   (21.47)   109                 

 

Parent-Child Relationshipc

   Mother-rated Conflict              20.90  (6.85)   114             

 

Sibling Conflictd                         -.006   (.94)     117

 

Teacher-Child Relationshipc

   Teacher-rated Conflict                                                                                          13.32  (6.14) 117

 

Child-Peer Relationshipe

   Teacher-rated Hostile-                                                                                           11.76  (5.29)     108

   Aggressiveness Scale Score

____________________________________________________________________________________

 

a Hollingshead Four Factor Index (scores range from 1-5, higher scores indicate lower SES)

b Conflict Tactics Scale - Sum of Frequency and Exposure to Verbal Aggression and Physical Aggression score (scores range from 0-160)

c Adult-Child and Teacher-Child Relationship Scale - Conflicted factor score (scores range from 8-40)

d Sibling Conflict Coding System Destructiveness Standardized Score  

e Teacher’s Checklist of Children’s Peer Relationships - Hostile-Aggressive scale score (scores range from 6-30)


 

 

 

 

 

Table 2

Relations Among Dependent and Independent Variables

 

 

Age 6 School Outcomes

Teacher-Child Conflict           Child-Peer Conflict

 

SES - Age 2                                                                       .11+                         .14+

(1)                              (107)

 

Interparental Conflict

Age 3.5 Verb. & Phys.                                                     .13+                              .31***

Aggression Comp.                                                            (109)                             (101)

 

Parent-Child Relationship - Age 5

Mother-rated Conflict                                                       .21***                           .31**

(2)                                (108)

 

Sibling Conflict - Age 5

Destructiveness Standardized Score                                  .07                                 .15+

(3)                              (108)

 

 

+p< .10, *p< .05, **p< .01, ***p< .001

NOTE: n’s are in parentheses. 

 

 

 

 

 

 


Table 3

The Interactive Effects of Conflict in Different Dyadic Relationships --

Interparental, Mother-Child, and Sibling Conflict -- on Predicting Teacher-Rated Problems from Age 3.5 to Age 6

 

                    Teacher-Child Conflict                                                   Child-Peer Conflict

_____________________________________________________________________________________________________________________

 

Independent Variables                                          Mult.R.      R2     R2cha.     Fchange.      Mult.R.     R2     R2cha.  Fchange

____________________________________________________________________________________________________________________ 

Step 1: Verbal + Phys. Aggr. at Age 3.5                  .14         .02      .02            2.15            31            .09        .09     10.54**

 

Step 2:  Mother-Child Conflict                                 .25         .06       .04            4.89*         .41            .17        .07        8.24**

at Age 5

 

Step 3:  Sibling Conflict at Age 5                             .26         .06        .00          .43              .42            .18        .01        1.30

 

Step 4:  Verbal + Phys. Aggr.                                 .28         .08      .01            1.46             .44            .19        .01        1.73

X Mother-Child Conflict

 

Step 5: Verbal + Phys. Aggr.                                  .34         .11      .03            4.00*            .45            .21        .01        1.79

X Sibling Conflict

 

Step 6:  Mother-Child Conflict                                .34          .11      .00            .11                .45            .21        .00        .02

X Sibling Conflict

 

Step 7:  Verbal + Phys. Aggr.                                 .39          .15      .03            3.94*            .49            .24        .03     4.10*

X Mother-Child Conflict

X Sibling Conflict

 

Overall F = 2.51, p < .01        Overall F = 4.21, p < .001

                   df = (7, 99)                            df = (7, 91)        

_____________________________________________________________________________________________________________________

 

+p < .10, *p < .05, **p < .01, ***p < .001

 

 


 

Figure Captions

 

Figure 1.  The interactive effects of interparental conflict at age 3.5 and sibling conflict at age 5 in predicting teacher-child conflict at age 6.

 

Figure 2.  The interactive effects of interparental conflict at age 3.5, mother-child conflict at age 5, and sibling conflict at age 5 in predicting teacher-child peer conflict at age 6.