Preschoolers’ aggression and parent–child conflict: A multiinformant and multimethod study
Introduction
Recently, researchers have begun to focus their attention on the use of relational aggression during early childhood (e.g., Crick et al., 1997, Hawley, 2003, Nelson et al., 2005, Ostrov and Keating, 2004). Previous research typically has found that relational aggression is more common in preschool girls than in preschool boys, whereas preschool boys display more physical aggression than their female peers (e.g., Bonica et al., 2003, Crick et al., 2006; cf. Estrem, 2005, Hart et al., 1998). Physical aggression involves the intent to hurt, harm, or injure another person (e.g., hitting, kicking, punching), whereas relational aggression involves behaviors that damage or threaten damage to relationships (e.g., malicious gossiping, peer exclusion) (Crick & Grotpeter, 1995). Relational aggression often is direct in nature during early childhood (e.g., “You can’t play with us” [Ostrov & Keating, 2004]) and is associated with adjustment difficulties such as peer rejection and loneliness for the perpetrators and their victims (Crick et al., 1997, Crick et al., 2006, Ostrov et al., 2004).
Past theory has focused on two dimensions of parenting practices that have been identified as (a) behavioral or psychological “control” and (b) acceptance, sensitivity, or an “emotional relationship” (Cummings, Davies, & Campbell, 2000, p. 162). Parent–child hostility, defined as harsh and inconsistent parental control combined with low levels of warmth or acceptance, tends to be associated with parent–child conflict (Cummings et al., 2000). Moreover, in past research, parent–child hostility has been found to predict externalizing problems such as physical aggression (Cummings et al., 2000). Studies exploring these theoretical dimensions have found that both parent–parent and parent–child relationships are linked to aggressive behavior in children during early childhood. For example, a cross-cultural study conducted with Russian children found less paternal responsiveness and more maternal coercion to be positively associated with physical and relational aggression (Hart et al., 1998). Hart and colleagues (1998) also found that marital conflict was associated with relational aggression, but only for boys. In a study conducted in the United States and Australia, children were found to model the relationship they had with their parents and to use a similar approach when interacting with peers (Russell, Hart, Robinson, & Olsen, 2003). Nelson, Hart, Yang, Olsen, and Jin (2006) found in a Chinese sample that psychological control (i.e., use of parenting techniques such as love withdrawal designed to limit children’s self-expression) was linked to aggression (both physical and relational) in girls, whereas physical coercion was associated with physically aggressive behaviors in boys. In agreement with Nelson and colleagues, Casas and colleagues (2006) found that psychological control (e.g., guilt induction) was associated with teacher-reported relational aggression in girls. Finally, a recent study using a diverse sample of 5- to 8-year-olds found that negative maternal affect was positively associated with teacher-reported relational aggression (Brown, Arnold, Dobbs, & Doctoroff, 2007). These past studies point to the complexity of this growing literature and suggest important differences based on the gender of the child.
The current study was conducted to investigate several empirical goals, including testing whether the intercorrelations between physical and relational aggression depend on the source of data and whether the association between parent–child conflict and children’s physical and relational aggression depends on the informant type. More specifically, this study was designed to build on past findings by exploring the role of parent–child conflict, a construct that is related to parenting practices such as psychological control or coercion, but rather than being an individual characteristic of the parent, parent–child conflict is a product of the dynamic bidirectional relationship between the parent and the child. Thus, similar to the past literature (e.g., Casas et al., 2006), we explored the impact of the family system and the potential moderating role of gender in aggressive behavior. To our knowledge, no study has yet tested the association between parent–child conflict and relational aggression using various methods during any developmental period. Given the documented positive association between socioeconomic status (SES) and relational aggression (Bonica et al., 2003, McNeilly-Choque et al., 1996), it is important to explore whether it is necessary to statistically control for these influences. Unlike some past studies (cf. Prinstein, Boergers, & Vernberg, 2001), we controlled for the variance associated with the alternative subtype of aggression. This practice permits an exploration of the unique amount of variance associated with each subtype of aggression.
This study had three central empirical goals. First, we explored the intercorrelations between physical and relational aggression (i.e., the association between physical and relational aggression) for the three informant types (parents, teachers, and observers). Conceivably, shared method variance will artificially inflate the intercorrelations for parent and teacher reports, but observations are based on multiple observers over several sessions, likely reducing this risk (Casas et al., 2006, Crick et al., 2006). Thus, we hypothesize moderate to high levels of correlation between physical and relational aggression constructs for teacher and parent reports, and we hypothesize low levels of association between the aggression subtypes for observations of these aggression subtypes. Second, we explored the interinformant agreement (i.e., the association between the informants) for physical and relational aggression across three informant types. Casas and colleagues (2006) examined the concordance between parents and teachers and found significant amounts of agreement among mothers and teachers for physical aggression but not relational aggression. Fathers and teachers significantly agreed for physical aggression and tended to agree for relational aggression (Casas et al., 2006). To our knowledge, no other studies have examined interinformant agreement across parents, teachers, and observers.
Pellegrini and Bartini (2000) made the distinction between private (ipsative) and public (normative) assessments of victimization and aggression. They found that direct observation measures of social behavior were related to teacher ratings and some peer ratings but not to private assessments such as diaries and self-reports. The authors further argued that some instances of aggression are secretive and quickly displayed, and they suggested that future observational techniques be designed for rare events to capture infrequently occurring behaviors (Pellegrini & Bartini, 2000). Pellegrini (2001) further called for additional studies with multiple methods and informants, including direct observation, to examine correlations across sources of data. In keeping with this past work, we anticipated moderate levels of agreement for both aggression subtypes between teachers and observers (e.g., Crick et al., 2006, Ostrov, 2006, Ostrov and Keating, 2004) and low but significant amounts of concordance between parents (i.e., privy to more private displays of aggression) and teachers (Casas et al., 2006) and between parents and observers. Third, we investigated how parent–child conflict affects concurrently observed physical and relational aggression with peers at school during early childhood when controlling for the alternative aggression construct and gender (Brown et al., 2007). We hypothesize that parent–child conflict will be positively and uniquely associated with both observed and teacher-reported relational and physical aggression.
To address these three empirical goals, we conducted time-intensive naturalistic observations of aggression subtypes at school and obtained both teacher and parent reports of physical and relational aggression during early childhood. In addition, we obtained parent reports of parent–child conflict, and for validity purposes we collected teacher reports of student–teacher conflict.
Section snippets
Participants
All of the parents who provided written consent for their children to participate in the larger ongoing study (e.g., Ostrov, in press) were invited to have one parent or legal guardian complete a parent questionnaire packet to be distributed and returned via mail. The participating preschools, all located in a large city in the northeastern United States, were four nationally accredited, university/college-affiliated child care centers. Parents who completed the packet were given a $10 gift
Results
The study was designed to test three objectives. Collectively, these objectives concern the role of the source of data and potential differences between informants concerning physical and relational aggression during early childhood. The first goal was to replicate and extend prior findings on the intercorrelations between physical and relational aggression across the three types of measures and informants. The second goal was to replicate and extend the past literature concerning correlations
Discussion
The current study had three empirical objectives. These goals broadly address the role of the source of data and potential differences between informants concerning physical and relational aggression during early childhood. The first objective was to test the intercorrelations between physical and relational aggression using three different informant types (parents, teachers, and observers). The second goal was to test the interinformant agreement for both physical and relational aggression.
Acknowledgments
The contributions of Emily E. Ries, Jamie L. Guzzo, Kirstin Grös, Stephanie A. Godleski, Adam Mullins, and the entire Early Childhood Play Project staff (University at Buffalo) are greatly appreciated. Portions of this study were completed in partial fulfillment of the degree of bachelor of arts with high honors by the second author under the direction of the first author. We thank the families, teachers, and directors of participating schools.
References (34)
- et al.
Parenting predictors of relational aggression among Puerto Rican and European American school-age children
Early Childhood Research Quarterly
(2007) - et al.
Early parenting and children’s relational and physical aggression in the preschool and home contexts
Journal of Applied Developmental Psychology
(2006) - et al.
A longitudinal study of physical and relational aggression during early childhood
Journal of Applied Developmental Psychology
(2006) - et al.
A normative study of family media habits
Journal of Applied Developmental Psychology
(2002) Strategies of control, aggression, and morality in preschoolers: An evolutionary perspective
Journal of Experimental Child Psychology
(2003)Deception and subtypes of aggression in early childhood
Journal of Experimental Child Psychology
(2006)- et al.
An observational study of delivered and received aggression, gender, and social–psychological adjustment in preschool: “This white crayon doesn’t work …”
Early Childhood Research Quarterly
(2004) - et al.
Preschoolers’ aggression, affiliation, and social dominance relationships: An observational, longitudinal study
Journal of School Psychology
(2007) - et al.
Relational aggression, relational victimization, and language development in preschoolers
Social Development
(2003) Ecology of the family as a context for human development research perspectives
Developmental Psychology
(1986)
Relational and overt aggression in preschool
Developmental Psychology
Relational aggression, gender, and social–psychological adjustment
Child Development
Developmental psychopathology and family process: Theory, research, and clinical implications
Relational and physical aggression among preschoolers: The effect of language skills and gender
Early Education and Development
Overt and relational aggression in Russian nursery-school-age children: Parenting style and marital linkages
Developmental Psychology
Cited by (35)
Using the social information processing model to understand gender differences in the manifestation and frequency of aggression
2022, Aggression and Violent BehaviorMeasurement of aggressive behavior in early childhood: A critical analysis using five informants
2021, Journal of Experimental Child PsychologyCitation Excerpt :However, parents might not be as objective as teachers and other observers when reporting on their own children’s behavior with peers. In previous studies, parent measures have shown acceptable internal consistency and, in some work, moderate correlations with teacher reports of physical and relational aggression (Ostrov & Bishop, 2008). However, in other studies, parent report was not significantly associated with teacher or observer report, potentially indicating that parents are reporting on aggression outside of the classroom setting (Ostrov, Gentile, & Mullins, 2013).
Relational aggression during early childhood: A systematic review
2021, Aggression and Violent BehaviorThe Relational aggression scale (RAS): Psychometric properties of a newly developed measure of relational aggression
2018, Revue Europeenne de Psychologie AppliqueeEvaluation of Positive Parenting Programme (Triple P) in Singapore: Improving parenting practices and preventing risks for recurrence of maltreatment
2017, Children and Youth Services ReviewThe inter-rater reliability of observing aggression: A systematic literature review
2017, Aggression and Violent BehaviorCitation Excerpt :Observers were asked to describe events in enough detail to make it possible to verify accuracy in placing the event under the designated category. Ostrov and co-authors employed this coding procedure afterwards in a series of studies using sometimes overlapping samples of children between 39 and 50 months (Ostrov, 2010; Ostrov & Bishop, 2008; Ostrov & Crick, 2007; Ostrov, Crick, & Stauffacher, 2006; Ostrov, Pilat, & Crick, 2006). IRR proved to be good to excellent in all Ostrov's studies, as it was in the study of Gower, Lingras, Mathieson, Kawabata, and Crick (2014) who used an adapted version of this observation procedure.