How Social Skills Are Influenced by Siblings

This is a study about determining what, if any factors of social skills are affected or influenced by siblings. If so, what can be done to help these factors become majority or all positive influences. This will help us understand the relationship of sibling interactions and the social adjustment of children with behavior problems while determining the negative and positive affects siblings have on social skills.

With parents and siblings, a total of 53 first and second-grade students were interviewed about conditions of the sibling relationship. When considering conflict and feeling together, 3 types of sibling dyads surfaced: conflictual (high levels of conflict, low levels of feelings), involved (moderate levels of conflict and feelings), and supportive (low levels of conflict, high levels of feelings). On most levels of social adjustment at school, students in involved sibling relationships showed better adjustment than did students in conflictual relationships. Results are discussed in terms of a developmental model for at-risk children in which some sibling relationships may foster the development of social skills in addition to providing emotional support, which may enhance adjustment at school.

Along with others, most of Patterson’s research (Loeber & Tengs, 1986; Patterson, 1986) is about the negative impact of siblings. It shows how negative behavior from siblings can create antisocial behavior in children. But when guided properly, research suggests relationships between siblings can and will foster positive socialization skills. In general, when children are younger, they spend majority of their time with siblings, which is how the relationship foundation is built (Bank & Kahn, 1975; Crouter & McHaJe, 1989). Siblings, especially older, give the child entertainment, togetherness, and support. This, in turn, will help develop the child emotionally. Negative and positive characteristics of sibling relationships are associated with language, social understanding, and perspective taking (Brown & Dunn, 1992; Howe, 1991; Howe&Ross, 1990; Stewart & Marvin, 1984), recommending that sibling interactions can provide opportunities to learn negotiation skills, affect regulation, and behavioral control. In sibling research, normative samples, conflicts have been positively related to sharing, helping, and cooperation in a preschool sample (Dunn & Munn, 1986). Howe and Ross (1990) identified a constellation of behaviors that they used to define friendly sibling relationships.

These behaviors included both conflictual negative exchanges as well as positive discussion of feelings and internal states. In this sample, both negative and positive interactions were related positively to outcome measures such as caretaking and perspective taking. Similarly, Stocker and Dunn (1990) found that competitive, controlling, older siblings were rated by mothers as having more positive best friendships. They postulated that these negative aspects of the relationships (control and competitiveness) are skills that reflect a greater social understanding and thus the potential for more positive friendships outside the sibling relationship. Perhaps the skills learned in the sibling relationship (rather than just behaviors displayed) may be important in the development of relationships outside the family.

Fifty-three behaviorally disruptive, White, target children ranging in age from 6 to 8 years old, their mothers, and their siblings were recruited as participants. The children were concurrently participating in a longitudinal intervention project designed to examine the developmental progression of conduct problems and the effectiveness of a preventive intervention program (Fast Track; Conduct Problems Prevention Research Group, 1992). A two-step screening process was used to identify the children. In the spring of their kindergarten year, 936 children were screened with an abbreviated version of a teacher rating form (the Teacher Observation of Classroom Adaptation—Revised [TOCA-R]; Kellam, 1989), which assesses a range of conduct problems at school. Consenting parents of children who scored in the top 30% on the sum scores of the TOCA-R were then called, and phone interviews were conducted with these parents. A 24-item scale of conduct problems drawn from the Child Behavior Checklist (Achenbach & Edelbrock, 1981) and the Revised Problem Behavior Checklist (Quay & Peterson, 1987) was used at this screen.

Based on the average teacher and parent problem scores, 80 children were then chosen for the intervention. Children were chosen for the high-risk sample only if they displayed several problem behaviors across multiple settings (home and school); the selected sample represented the top 10% of the sampling population in terms of their cross-situational problem behavior ratings. The gender distribution of the sample reflected epidemiological data suggesting more disruptive behavior problems in boys than in girls. Family characteristics of the high-risk sample were indicative of additional risk factors. For example, 23% of families were unemployed, 25% of mothers were high school dropouts, 35% of families were in the lowest socioeconomic status category, and 36% of parents were single. The 53 target children with siblings ranging in age from 4-12 years old were eligible to participate.

When target children had several siblings in this age range, the one closest in age was chosen as the designated sibling. In 40 of the dyads, the target child was the younger sibling, whereas in 12 dyads the target child was older (one child was a twin). Twenty-seven of the dyads were male pairs, 9 were female pairs, and 17 were mixed. Children were interviewed individually while attending the intervention program, outside of regular school time. Mothers were interviewed at home by a trained interviewer who was known to the family. The child interview was based on the sibling interview scales used by Stocker and McHale (1992). Two items were added to expand the Conflict and Warmth scales and to achieve comparability in the interview items used for older and younger children. The resulting scale included 26 items, each presented in a 4-point Likert scale. For the target children and younger siblings, a picture presentation was used that was modeled after the Harter and Pike (1984) Pictorial Scale of Perceived Competence and Acceptance for Young Children. For older children, interview questions were identical in content, but pictures were not used.

Based on prior research (Stocker & McHale, 1992), at least two factors were expected, reflecting the positive and negative dimensions of the sibling relationship. Although the clear positive and negative content of items suggested these two factors would emerge, the at-risk nature of this sample and rewording of items warranted factor analysis to confirm these dimensions. In the present study, factor analyses were conducted separately for items reflecting initiated versus received behavior. Positive items included mutual play, sharing, affection, sharing secrets, and comforting the sibling. Negative items included stealing, mean behavior, fighting, physical aggression, and anger.

The Maternal Interview About Children's Sibling Relationships (Stocker, Dunn, & Plomin, 1989) was administered to mothers. Mothers rated the sibling dyad and the behaviors of both children in the dyad by answering questions addressing specific qualities of the sibling relationship and children's behavior toward each sibling, for example, arguing, fighting, and playing. Questions were open-ended and were rated on a 6-point scale ranging from 0 (almost never/rarely) to 5 (regularly/every day). The Conflict scale contained items that assessed physical fighting, competition, and jealousy-rivalry with the mother. The Warmth scale included items that assessed desire to be with the sibling, affection toward the sibling, joint play, caretaking, and nurturance.

Sociometric interviews were conducted individually with the classmates of the target children. At least 75% of the peers in each class participated in the interviews. Each child was presented with a list of the classmates in his or her classroom. The interviewer read through the roster once, then asked for a series of nominations. Children were asked to pick the three classmates they liked the best and three classmates they liked the least. Initially, these scores were standardized within classrooms and were combined to identify sociometric status groups, following the Coie and Dodge (1983) procedure. The standardized like-least and like-most scores were retained as continuous measures of positive and negative dimensions of peer relations, as they provided a better range of scores to assess children's functioning in this domain.

The Social Health Profile (SHP) provided a teacher-rated measure of social skills and behavioral control at school. This 41-item measure includes items from the TOCA-R (Werthamer-Larsson & Kellam, 1989) as well as 9 items assessing social competence. Three scales from this measure were used in analyses with item content reflecting Social Competence, Inattention/School Performance, and Emotional Control. These scales were constructed based on face validity and intercorrelations of items to reflect constructs of interest to this study.

First, analyses of variance (ANOVs) were conducted on each of the factor scales to examine the impact of the cohort (Cohort 1 vs. Cohort 2), gender composition, age of sibling, and marital status on the sibling relationship. Few of these analyses revealed significant differences. However, children in Cohort 1 had siblings who reported more initiated conflict compared with children in Cohort 2, whereas mothers in Cohort 2 reported more sibling warmth. In addition, younger siblings reported that they initiated more warmth toward target children than did older siblings. There were no other effects of gender composition, age, or marital status on target child, sibling, or maternal ratings of the sibling relationship. Thus, in general, older and younger children reported similar sibling experiences.

In addition, mothers and children from married and single-parent families reported similar sibling relationships. We also conducted ANCAAs on each measure of school adjustment and sociometric nominations to examine cohort, gender, and age effects. Children in mixed-gender sibling relationships tended to have more like-most nominations. No other effects were significant. Thus, neither intervention, gender composition, nor sibling age appeared to have a major effect on the dependent variables studied here. Hence, participants were combined across the cohort, gender, and age of sibling for further analyses. Maternal reports of the sibling relationship were collected on all participants as were teacher and parent ratings. Two children were missing sociometric data.

We computed correlations to explore relations among target child, sibling, and maternal ratings of conflict and warmth in the sibling relationship. To construct an index of the quality of the sibling relationship that would include the opinions of both target child and sibling, we combined the ratings of members of each sibling dyad. Because it was possible that children's ratings of their own behaviors (particularly conflict behaviors) would be diminished by defensiveness, we examined the constructs of initiated conflict and initiated warmth separately from the constructs of received warmth and received conflict.

Next, to test the hypothesis that sibling conflict and sibling warmth would be associated with school social adjustment, we conducted correlations between dyadic measures of initiated and received conflict and warmth and target child school adjustment measures of aggression, emotional control, inattention, social competence, and positive and negative sociometric nominations. The correlations suggest that received conflict was associated with teacher ratings of aggression and social competence. In addition, teacher ratings of emotional control, inattention, and peer like-least nominations also tended to vary as a function of received conflict. That is, children in sibling dyads characterized by high rates of received conflict appeared to be more aggressive and less socially competent at school. In addition, ratings of initiated warmth were significantly correlated with higher levels of emotional control. Maternal ratings of the sibling relationship and ratings of initiated conflict and received warmth were less useful in terms of predicting target child behaviors and social-emotional competencies at school. This pattern of correlations is consistent with the hypothesis that sibling conflict can provide a training ground for child aggression (Loeber & Tengs, 1986; Patterson, 1986).

Consistent with previous studies, factor analyses of child and maternal perceptions of the sibling relationship revealed two dimensions, one reflecting warmth and support and the other reflecting conflict (Stocker & McHale, 1992; Stocker et al.,1989). Of these dimensions, children and mothers tended to agree more on their ratings of warmth than conflict. In addition, target children, siblings, and mothers may have different perspectives on what constitutes high conflict within the sibling relationship. Despite the difficulty in collecting reliable ratings, children's ratings of the conflict they received from their siblings were correlated with some aspects of their school social adjustment. consistent with the models proposed by Patterson and his colleagues (Loeber & Tengs, 1986; Loeber, Weismann, & Reid, 1983; Patterson, 1982), sibling conflict was related to aggressive social difficulties with peers, poor peer relations, and behavioral problems at school. In addition, some research suggests that aggressive children minimize their own aggression and perceive peers as more aggressive than themselves (Lochman, 1987). When correlations between warmth in the sibling relationship and school social adjustment were considered, the level of support that children and siblings reported initiating toward each other was significantly correlated with teacher ratings of emotional control and moderately correlated with social competence.

Several limitations of this study warrant discussion. Furthermore, the direction of results is subject to interpretation. Although research has consistently focused on the relationship between childhood relationships with siblings and subsequent child adjustment at school, the reverse effect may also be true. For example, Kramer and Gottman (1992) found that peer interactions before the birth of a sibling were predictive of sibling interactions 6 to 14 months later. They suggested that children learn skills within their peer relationships that enable them to interact effectively with siblings. Observational data may be useful in future research to further clarify the relationships between sibling interactions and social skills in high-risk samples.

Future research should also extend the present results by focusing on potential mediators between the sibling relationship and measures of adjustment in a longitudinal design. For example, if aggressive children are learning social-cognitive skills within sibling relationships, they would likely have higher scores on problem-solving and emotional identification measures. In addition, measures of self-esteem and coping would provide information on the potential protective function of sibling support in childhood. The study of mechanisms through which sibling relationships affect later development will be an important direction of future research.