An observational study of delivered and received aggression, gender, and social-psychological adjustment in preschool: “This White Crayon Doesn’t Work …”

https://doi.org/10.1016/j.ecresq.2004.04.009Get rights and content

Abstract

A semi-structured observational study investigated gender differences in delivered and received relational, physical, verbal, and nonverbal aggression in a young preschool sample (N=60). Findings revealed that gender differences in subtypes of aggression may be apparent as early as 3 years of age. Specifically, girls were found to deliver and receive more relational aggression than males, whereas boys tended, although not significantly, to deliver and significantly received more physical aggression than females. Relational and physical subtypes of delivered and received aggression were differentially associated with preschoolers’ social-psychological adjustment.

Introduction

The study of gender linked peer-directed aggression and victimization has recently generated a plethora of findings that have important implications for researchers, clinicians, and educators who are concerned with accurately accounting for the developmental trajectories of both boys and girls (Crick et al., 1999, Crick et al., 1999; Zahn-Waxler, 1993). In defining the gender-linked hypothesis of aggression, Crick and her colleagues describe two subtypes of aggression. Physical aggression is defined as those behaviors that involve intent to hurt or to harm others through physical acts such as pushing and pinching as well as verbal acts such as threatening physical force (Crick et al., 1999, Crick et al., 1999), and is typically perpetrated by boys (for a review see Coie & Dodge, 1998). Relational aggression is conceptualized as the intent to harm others by removing or threatening to damage a relationship or feelings of social acceptance and inclusion in social groups and is relatively more common among girls in early and middle childhood (for a review see Crick et al., 1999, Crick et al., 1999).

Researchers have also begun to evaluate a gender-balanced model of peer victimization. Specifically researchers studying preschoolers (Crick et al., 1999, Crick et al., 1999), school-aged children (Crick et al., 2001; Schafer, Werner, & Crick, 2002), and adolescents (see Crick et al., 2001) note that hostile female–female interactions, typically consist of relational victimization, which has been operationalized as being the frequent target of relationally aggressive acts from peers, friends, or significant others (Crick et al., 2001). In contrast, physical victimization involves being the frequent target of physically aggressive behaviors from others and is more typical of hostile male interactions (Crick et al., 1999, Crick et al., 1999; Schwartz, Dodge, Pettit, & Bates, 1997). The fact that young girls often play in gender-segregated female groups (Maccoby, 1990), suggests that they may be at increased risk for experiencing relational victimization from their peers. Moreover, given that girls have been found to have an increased desire to achieve close interpersonal social bonds as well as an increased tendency to focus on relational features of interactions (Block, 1983; Crick & Grotpeter, 1995; Cross & Madsen, 1997), they may be at increased risk for experiencing negative consequences from relational victimization. In contrast, boys may be differentially vulnerable to physical acts that subvert their social dominance relationships given their increased propensity for attending to group structural changes and instrumental/dominance interactions (Bugental, 2000).

A gender-based approach to the study of aggression and victimization in early childhood is important in that it may significantly increase our understanding of the social development of young children and may elucidate specific factors that are relevant for deciphering maladaptive developmental pathways specific to each gender (Crick & Rose, 2000; Zahn-Waxler, 1993). Despite the possible scientific and practical benefits such studies would provide, few researchers have responded to existing calls for research in this area (Crick et al., 1999, Crick et al., 1999; Pellegrini, 1998) resulting in surprisingly little empirical work conducted with early childhood samples (for reviews see Archer, in press; Crick et al., 1999, Crick et al., 1999). Despite progress made in this field to date, there remain several limitations and the present study is designed to address some of these existing limitations.

Researchers utilizing teacher-and peer-report methods have generally found evidence of gender differences in relational and physical aggression in preschoolers in the United States (Bonica, Yershova, Arnold, Fisher, & Zeljo, 2003; Crick et al., 1997) and in early childhood studies conducted abroad (Russell, Hart, Robinson & Olsen, 2003; cf. Hart, Nelson, Robinson, Olsen, & McNeilly-Choque, 1998). Initial evidence from recent research has also documented gender differences in relational and physical victimization in preschool samples based on teacher reports (Crick et al., 1999, Crick et al., 1999; cf. Bonica et al., 2003). Although these findings are consistent with the gender-linked model, they are based on a rather small body of literature as the majority of studies of relational aggression and victimization have focused on school-age children and adolescents (for reviews see Crick et al., 1999, Crick et al., 1999; Crick et al., 2001). Moreover, in past studies the overlap between physical and relational aggression has been moderate to high (see Crick et al., 1997) and further research testing the orthogonal nature of these two constructs is needed during early childhood. One major limitation of the extant developmental literature is the lack of research focusing on young preschool children in empirical studies of relational aggression and relational victimization (cf. Crick et al., 1997, Crick et al., 1999, Crick et al., 1999, Hart et al., 1998; McNeilly-Choque, Hart, Robinson, Nelson, & Olsen, 1996). A second major limitation of past research is that the study of childhood aggression and victimization subtypes has rarely utilized direct observational methods. Past research has instead relied almost exclusively on peer nominations, self-reports, teacher-report paradigms, semi-structured interviews, or on children’s reactions to hypothetical social interactions (for reviews see Archer, in press; Crick et al., 1999, Crick et al., 1999; Crick et al., 2001).

Although findings are mixed, the majority of studies have demonstrated that girls deliver more relational aggression than boys, who are more likely to display more physical aggression than girls (Crick & Grotpeter, 1995; Rys & Bear, 1997; cf. Tomada & Schneider, 1997). These general developmental findings and gender differences however remain largely untested with observational methods for samples of young preschool children (see Crick et al., 1999, Crick et al., 1999). In addition, those few preschool studies that have been conducted have often yielded mixed findings regarding gender differences in relational aggression during early childhood, perhaps due to cultural differences in the samples investigated (Hart et al., 1998; but see Crick et al., 1997, Crick et al., 1999, Crick et al., 1999, McNeilly-Choque et al., 1996; Ostrov & Keating, 2004; Russell et al., 2003). Given these discrepant findings, verification of past results with observational techniques is essential.

The developmental study of aggression during early childhood has a rich history of utilizing observational methods for the study of physical and verbal aggression (for reviews see Coie & Dodge, 1998; Reid, Baldwin, Patterson, & Dishion, 1988). Observational studies during early childhood have utilized a diverse array of sampling techniques and methodologies including naturalistic and systematic free play observations in the classroom and on the playground (e.g., Arnold, Homrok, Ortiz, & Stowe, 1999; Fagot & Hagan, 1985; Laursen & Hartup, 1989; McNeilly-Choque et al., 1996), structured observations utilizing play group designs (e.g., Coie, Dodge, Terry, & Wright, 1991; Olson, 1992;), as well as home and structured laboratory assessments (e.g., Cummings, Iannotti, Zahn-Wazler, 1989).

Recently studies utilizing observational techniques with preschoolers have investigated both physical and relational aggression, offering some behavioral confirmation of past hypotheses and findings. In the first known observational attempt to study relational and overt aggression during early childhood (head start sample: M=58 months, S.D.=3.70; university sample: M=60 months, S.D.=4.30), McNeilly-Choque et al. (1996) reported weak to moderate correlations between teachers, peer reports and actual observed behavior on preschool playgrounds. Ostrov and Keating (2004) observed various aggression subtypes displayed and received by preschool children (M=64 months; S.D.=6.77) during naturalistic free play sessions and in a semi-structured coloring task. In general, these authors found that boys delivered and received more overt aggression (i.e., physical and verbal aggression) than girls, whereas girls delivered and received more relational aggression than boys. Taken together these studies offer the impetus for the present research, which is designed to replicate and extend these past findings with a younger sample, while addressing some fundamental unanswered research questions. Despite these extensive past and current efforts to observe aggressive (delivered and received) and antisocial behavior in young children, the field has only recently begun to develop observational systems to capture relational aggression for young boys and girls.

In order to rectify the two major limitations, the lack of observational studies of relational aggression and the paucity of research during early childhood, the present observational study was designed to test the gender-linked hypothesis of delivered and received aggression with the youngest observational sample to date.1 In the present study, we employed a substantially revised version of the coloring task paradigm developed by Ostrov and Keating (2004) to assess subtypes of aggression and prosocial behavior. The semi-structured coloring task was designed to be similar to those utilized in past developmental studies during early childhood (e.g., Camras, 1984; Charlesworth & Dzur, 1987). The coloring task consisted of a 9 min mild conflict provoking semi-structured task designed to investigate aggressive and prosocial tactics with same-sex peers by providing preschool triads with limited resources (i.e., only one functional crayon) in a coloring task. Although, 9 min is a brief observational period, based on past studies in which a 5–10 min time intervals were used to assess aggression, (Goldstein, Arnold, Rosenberg, Stowe, & Ortiz, 2001; Pepler, Craig, & Roberts, 1998) we were confident that sufficient levels of aggression would be displayed.

In the present study, we used triads (dyads were used by Ostrov & Keating, 2004) in order to increase the ecological validity of the observational task, based on evidence that preschoolers’ aggressive acts often occur in groups of three or more children (Farver, 1996, Trawick-Smith, 1992). Triads were also utilized in order to remove potential group preference biases for each gender, given findings that suggest that male children prefer larger groups and female children prefer dyadic interactions (Benenson, 1993). We also restricted the triads to same-sex groups, given evidence that preschoolers’ interactions are most likely to occur naturally in this constellation (Maccoby, 1990) thus increasing the ecological validity of the context. In addition, the triads consisted of only same-sex children because researchers have found distinct behavioral styles (i.e., girls used more verbal behaviors and boys used more physical behaviors) in preschool same-sex groups in a laboratory-based resource utilization study (Charlesworth & Dzur, 1987). To decrease children’s reactivity we seated the experimenter behind a large divider during the experimental portions of the coloring task in contrast to the approach used by Ostrov and Keating (2004) in which the experimenter and camera were in clear view.

In the present study, four types of aggressive behaviors were assessed: relational aggression, physical aggression, verbal aggression (e.g., verbal insults and mean names), and nonverbal aggression (e.g., mean faces, intrusive pointing, and chin thrusts, which are hostile movements of the chin to signal threat to others, Keating & Heltman, 1994; etc.), all of which have been shown to be important indicators of hostile, mean behavior in past studies (see Coie & Dodge, 1998; Crick et al., 1999, Crick et al., 1999; Galen & Underwood, 1997). Additionally, delivered and received prosocial behavior was also examined to provide a relatively comprehensive and valence-balanced (i.e., positive and negative) assessment of children’s social behavior.

In using this revised method, we hypothesized that physically aggressive tactics would be utilized more by boys, whereas, relationally aggressive strategies would be employed more frequently by girls. Similarly, we predicted that boys would be more likely than girls to be the recipients of physical aggression, whereas girls would be more likely to be the targets of relational aggression. Further, we hypothesized that, in keeping with past findings, verbal aggression would be observed as a frequent tactic used both by boys and girls (Ostrov & Keating, 2004). Finally, based on initial evidence that nonverbal aggression may play a particularly important role in school-aged girls’ groups (Galen & Underwood, 1997), we predicted that nonverbal aggression would be more prevalent among the interactions of preschool girls, relative to those of boys.

The second goal of this study was to examine the association between observed delivered and received aggression subtypes and children’s social-psychological adjustment problems as reported by the focal children’s teachers. In order to avoid shared method variance and to provide an independent assessment of the focal children’s concurrent adjustment, we asked head teachers to complete several established instruments assessing social-psychological adjustment problems. To address this issue in a relatively thorough manner, we included a broader range of social-psychological adjustment indices (e.g., depressed affect, peer rejection, asocial behavior, prosocial behavior) than those utilized in previous observational studies of relational aggression (e.g., peer acceptance). Based on results from past studies in which peer and teacher assessments of aggression and victimization have been used, we hypothesized that all subtypes of observed aggression (delivered and received) would be associated with concurrent social-psychological adjustment problems for both boys and girls (e.g., Crick, 1996).

In order to address our various research goals we used teacher reports of social-psychological adjustment and a revised semi-structured observational paradigm (i.e., coloring task) in which 60 young preschool children were observed interacting with their peers.

Section snippets

Participants

Participants included 60 children (31 boys and 29 girls) ranging in age from 44 to 66 months (M=54.86 months; S.D.=6.06) and their head teachers, selected from two preschool sites (five mixed aged classrooms, with 12–16 children per classroom) affiliated with a large public university located in a large Midwestern city. Children were from diverse socioeconomic backgrounds representing most spectrums of yearly income (US$ 5000 to >90,000), but families were primarily middle class. The

Results

First, a series of ANOVA analyses were conducted that tested the hypothesis that gender differences would emerge for delivered and received subtypes of aggression and prosocial behavior. Next, a series of regression analyses and partial correlations (controlling for age) were conducted that tested the hypothesis that subtypes of aggression would be associated with concurrent social-psychological adjustment problems.

Discussion

Researchers have repeatedly recognized the limitations of teacher and peer informants and have acknowledged the need for observational confirmation of aggression and victimization subtypes of young children’s behavior (Crick et al., 1999, Crick et al., 1999; Crick et al., 1999, Crick et al., 1999). The results of this study provide important observational evidence to support the gender-linked model of aggression and victimization in early childhood. Specifically, boys tended, although not

Acknowledgements

This research and preparation of this manuscript was supported by a NIMH Traineeship (MH-15755) to the Institute of Child Development and by an Eva O. Miller Fellowship in support of the first author, by John E. Anderson Research Funds to the project team, and from NSF (BCS-0126521) and NIMH (MH63684) grants to the fifth author. We wish to thank Brielle Wacker, Ivan Liceaga, Deanna Spanjers and Maran Wolston for their assistance with coding. We thank Craig H. Hart, Crystal Cullerton-Sen, and

References (59)

  • R.S. Atlas et al.

    Observations of bullying in the classroom

    Journal of Educational Research

    (1998)
  • J.F. Benenson

    Greater preference among females than males for dyadic interaction in early childhood

    Child Development

    (1993)
  • J.H. Block

    Differential premises arising from differential socialization of the sexes: Some conjectures

    Child Development

    (1983)
  • C. Bonica et al.

    Relational aggression and language development in preschoolers

    Social Development

    (2003)
  • D.B. Bugental

    Acquisition of the algorithms of social life: A domain-based Approach

    Psychological Bulletin

    (2000)
  • W.R. Charlesworth et al.

    Gender comparisons of preschoolers’ behavior and resource utilization in group problem solving

    Child Development

    (1987)
  • Coie, J. D., & Dodge, K. A. (1998). The development of aggression and antisocial behavior. In W. Damon (Editor-in...
  • J.D. Coie et al.

    The role of aggression in peer relations: An analysis of aggression episodes in boys’ play groups

    Child Development

    (1991)
  • N.R. Crick

    The role of overt aggression, relational aggression, and prosocial behavior in children’s future social adjustment

    Child Development

    (1996)
  • N.R. Crick

    Engagement in gender normative versus non-normative forms of aggression: Links to social-psychological adjustment

    Developmental Psychology

    (1997)
  • N.R. Crick et al.

    Physical and relational peer victimization in preschool

    Developmental Psychology

    (1999)
  • N.R. Crick et al.

    Relational and overt aggression in preschool

    Developmental Psychology

    (1997)
  • Crick, N. R., Nelson, D. A., Morales, J. R., Cullerton-Sen, C., Casas, J. F., & Hickman, S. E. (2001). Relational...
  • Crick, N. R., & Rose, A. J. (2000). Toward a gender balanced approach to the study of social-emotional development: A...
  • Crick, N. R., Werner, N. E., Casas, J. F., O’Brien, K. M., Nelson, D. A., Grotpeter, J. K., et al. (1999). Childhood...
  • S.E. Cross et al.

    Models of the self: Self-construals and gender

    Psychological Bulletin

    (1997)
  • E.M. Cummings et al.

    Aggression between peers in early childhood: Individual continuity and developmental change

    Child Development

    (1989)
  • B.T. Fagot et al.

    Aggression in toddlers: Responses to the assertive acts of boys and girls

    Sex Roles

    (1985)
  • B.R. Galen et al.

    A developmental investigation of social aggression among children

    Developmental Psychology

    (1997)
  • Cited by (0)

    View full text