Early Childhood Research Quarterly An intervention for relational and physical aggression in early childhood: A preliminary study

A preventive intervention for reducing physical and relational aggression, peer victimization, and increasing prosocial behavior was developed for use in early childhood classrooms. Nine classrooms were randomly assigned to be intervention rooms (N = 202 children) and nine classrooms were control rooms (N = 201 children). Classroom was the unit of analysis and both observations and teacher-reports were obtained at pre and post-test. Focus groups were used to develop the initial program. The 6-week program consisted of developmentally appropriate puppet shows, active participatory sessions, passive concept activities and in vivo reinforcement periods. Preliminary findings suggest that the “Early Childhood Friendship Project” tended to reduce physical and relational aggression, as well as physical and relational victimization and tended to increase prosocial behavior more for intervention than control classrooms. Teachers and interventionists provided positive evaluations of the program and there is evidence for appropriate program implementation. © 2008 Elsevier Inc. All rights reserved. Peer relationships serve as a salient context for children’s cognitive, social, emotional, and physiological development across the lifespan (Hartup, 1996; Rubin, Bukowski, & Parker, 2006). Developmental psychology and psychopathology literatures indicate that the skills acquired within peer relationships (e.g., conflict resolution, emotion regulation, perspective taking, friendship formation, and social competence) during children’s early school years significantly impact children’s interpersonal relationships in subsequent developmental periods (for review, see Sroufe, Egeland, & Carlson, 1999). Significant problems with peer relations may also lead to peer rejection, victimization, social withdrawal, internalizing problems, or externalizing problems, including aggressive behavior, all of which may negatively affect academic competence and the transition to school (e.g., Belsky & MacKinnon, 1994; Buhs & Ladd, 2001; Rimm-Kaufman & Pianta, 2000). Arguably the earlier we intervene for aggression, the greater the probability of children improving (Sroufe, 1997), which was the impetus for designing a classroom-based intervention for early childhood in the present study. Children from high-risk environments (e.g., high crime and low SES neighborhoods) may begin school with problems in impulse and/or emotional control and social problem solving (Lochman, Lampron, & Rabiner, 1989; Shaw, Keenan, & Vondra, 1997). These children are subsequently more likely to engage in aggressive social behavior and to experience peer rejection (e.g., Dodge, Pettit, & Bates, 1995). These children’s difficulties are compounded by academic problems that are intertwined with and exacerbated by behavior problems (Hinshaw, 1994). To impact such pervasive problems, interventions must therefore target educational systems to address multiple contextual factors contributing to the development and ∗ Corresponding author at: 214 Park Hall, Department of Psychology, Buffalo, NY 14260-4110, USA. Fax: +1 716 645 3801. E-mail address: [email protected] (J.M. Ostrov). 1 The author is now at the Center for Disease Control and Prevention, Atlanta, GA. 2 These authors made equivalent contributions and are listed alphabetically. 0885-2006/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ecresq.2008.08.002 Author's personal copy 16 J.M. Ostrov et al. / Early Childhood Research Quarterly 24 (2009) 15–28 maintenance of such problems (Greenberg, Domitrovich, Graczyk, & Zins, 2001). To prevent or reduce peer problems in school, interventions must be in settings where aggression occurs (Coie, 1996; Offord, 1996). 1. Aggressive behavior Researchers have continued to generate and refine models of the development of aggressive behavior as well as evidencebased preventive interventions. Physical aggression is defined as the intent to hurt, harm, or injure with physical force or the threat of physical force, including kicking, hitting, pushing, and forcibly taking objects (Dodge et al., 2006). These behaviors have been associated with a host of social–psychological adjustment factors (e.g., peer rejection, loneliness; Miller-Johnson et al., 2002) and disruptive behavioral disorders (e.g., Attention Deficit Hyperactivity Disorder, Conduct Disorder, Oppositional Defiant Disorder; Waschbusch, 2002) that are problems in school settings and a continued challenge for teachers to address in the classroom. Moreover, aggressive behavior is a major educational and public health concern and is associated with problematic school transitions in early childhood (Ladd & Price, 1987) as well as academic difficulties throughout the school years (Leff, Power, Manz, Costigan, & Nabors, 2001). In addition, victimization has been found to be associated with lowquality friendships, characterized by low levels of warmth and intimacy (Goldbaum, Craig, Pepler, & Connolly, 2003). Finally, physical aggression has been found to be more prominent in boys relative to girls (Dodge et al., 2006). Physical and relational aggression (i.e., using the removal or threat of the removal of the relationship to harm, including social exclusion, friendship withdrawal threats, ignoring, spreading malicious rumors, gossip, secrets, and lies; Crick & Grotpeter, 1995) have been theorized and demonstrated to be unique constructs in early childhood and later developmental periods (see Crick, Ostrov, & Kawabata, 2007). During early childhood, these constructs are often correlated when using teacher-report, peer-report, or parent-report (e.g., Crick et al., 2007; Ostrov & Bishop, 2008), but those correlations are not as high in magnitude or sometimes not even significantly associated when observational methods are used (e.g., Crick et al., 2006; Ostrov, 2006). Moreover, teacher-, peer-, self-, and parent-reports of relational and physical aggression have unique associations with adjustment, social–cognitive correlates, physiological markers, and psychopathology (e.g., Bailey & Ostrov, 2008; Crick et al., 2006; Cullerton-Sen et al., in press; Marsee & Frick, 2007; Miller & Lynam, 2003; Murray-Close, Han, Cicchetti, Crick, & Rogosch, 2008; Pellegrini & Long, 2003; Prinstein, Boergers, & Vernberg, 2001). Relational aggression has been found to be uniquely associated with significant social–psychological adjustment problems across development (e.g., peer rejection) and is associated with symptoms of psychological disorders (e.g., ADHD, borderline personality features, eating disorders, substance use and abuse, delinquency, and internalizing problems; Crick, MurrayClose, & Woods, 2005; Crick et al., 2006; Murray-Close, Ostrov, & Crick, 2007; Werner & Crick, 2004; Zalecki & Hinshaw, 2004). In general, relational aggression is believed to be more prevalent in girls relative to boys, but this finding may depend on the developmental period, culture, or assessment strategy used (Crick et al., 2007). During early childhood, however, girls are generally more relationally aggressive with peers than are boys, display more relational aggression to female peers than they do to male peers and display relational aggression more than physical aggression (e.g., Bonica, Arnold, Fisher, Zeljo, & Yershova, 2003; Ostrov & Keating, 2004; Russell, Hart, Robinson, & Olsen, 2003; cf. Estrem, 2005; Hart, Nelson, Robinson, Olsen, & McNeilly-Choque, 1998). 2. Interventions for physical aggression To date, hundreds of studies have focused on prevention or intervention for physically aggressive behavior (see Leff et al., 2001). One of the largest efforts was “Fast Track,” a multi-year, multi-component intensive intervention for young children at-risk for early-onset conduct problems. The Fast Track program was found to have small positive effects on children’s aggressive behavior, social cognitions, prosocial behavior, and social skills during elementary school (CPPRG, 2004). These researchers further documented that their program was cost-effective for children at highest risk despite the considerable expense of the program (Foster et al., 2006). A second program, the Coping Power Program, a preventive intervention for preadolescent physically aggressive boys and their parents has been found to have significant intervention effects with high-risk boys from diverse backgrounds displaying lower rates of antisocial behavior at 12-month follow-ups (Lochman & Wells, 2004). Additional preventive interventions have been developed and found to be effective with children during early childhood and under the age of five. The most widely known intervention of this type is the “Incredible Years” (IYS), an evidence-based program that relies on the use of developmentally appropriate child-size puppets to role-play and communicate various social skills modules for a child training group, as well as parent and teacher training components (see Reid & Webster-Stratton, 2001). The evidence-based IYS programs target externalizing behaviors but were specifically designed for children presenting with Oppositional Defiant Disorder and Conduct Disorder (Reid & Webster-Stratton, 2001). The program generally focuses on teaching parents and teachers child-directed play techniques, consistent and developmentally appropriate parenting strategies, and strategies for improving children’s social interactions with peers (Reid, Webster-Stratton, & Baydar, 2004). Child training is reinforced with activities as well as a token economy system (Reid & Webster-Stratton, 2001). This work has been primarily conducted in individual treatment groups with clinic-referred children (e.g., Webster-Stratton, Reid, & Hammond, 2001). The clinic-based parent and child-training has revealed robust findings. For example, a 2-year follow-up study found that approximately 75% of children that participated in the project had returned to adaptive levels of functioning according to parentand teacher-report (Reid, Webster-Stratton, & Hammond, 2003). An additional independent examination Author's personal copy J.M. Ostrov et al. / Early Childhood Research Quarterly 24 (2009) 15–28 17 of a psychosocial family-based adapted version of the IYS program with preschool children at very high risk for conduct and antisocial behavior has shown marked improvement in a variety of domains including social competence with peers for preschool-age children who receive the program relative to no-intervention control groups (Brotman et al., 2005). In fact, control children were five times higher in levels of observed physical aggression relative to the children that received the intervention at a 16-month follow-up (Brotman et al., 2008). The more recent Dina Dinosaur (“Dinosaur School”) classroom-based component of the IYS program includes 30 lesson plans comprised of videotape modeling, puppets, activities, homework, and role-plays (Webster-Stratton, Reid, & Stoolmiller, 2008; Webster-Stratton & Taylor, 2001). In the past, teachers followed lesson plans for the specific content areas given during circle time or class meetings for two times per week and approximately 15 min per day. Teachers were also encouraged to promote the skills taught in circle time lessons during the day in less structured and free play times both inside and outside the classroom (Webster-Stratton et al., 2008). The classroom-based IYS program has been primarily developed for use with Head Start classrooms and elementary schools serving culturally diverse as well as lower SES families (Webster-Stratton et al., 2008). The IYS universal classroom-based intervention has also demonstrated, via independent blinded observations of classroom behavior, greater effects in reducing externalizing problems (i.e., aggression and disruptive behaviors), and increasing social–emotional competence relative to control classrooms (Webster-Stratton et al., 2008). The program had the greatest effect on classrooms with children at high levels of initial conduct problems and low levels of initial school readiness (Webster-Stratton et al., 2008). Collectively, these programs seem effective for reducing physical aggression and fostering social competence skills, but to date, few intervention programs have been designed to specifically address relational aggression in young children. For example, some relationally aggressive preschool children have surgent or extraverted personality styles (Russell et al., 2003) as well as advanced social and communication skills (e.g., Bonica et al., 2003) that are often targeted in the traditional programs, but they may have greater problems with making and keeping friendships (Johnson & Foster, 2005) that are less emphasized in the current evidence-based programs. 3. Interventions for relational aggression Despite the multiple calls for intervention efforts to address a variety of subtypes of aggressive behavior in schools (Geiger, Zimmer-Gembeck, & Crick, 2004; Leff et al., 2001; Young, Boye, & Nelson, 2006), few studies have been conducted. Given that different developmental factors may be involved in relational and physical aggression (see Zahn-Waxler, Crick, Shirtcliff, & Woods, 2006), programs developed to reduce or prevent physical aggression may not be appropriate for relational aggression. Yoon and colleagues recently wrote: “. . .there is an urgent need to promote a greater understanding of relational aggression and to develop effective, innovative approaches in schools” (Yoon, Barton, & Taiariol, 2004, p. 310). Leff et al. (2001) stated that: “. . .it is important that the public understand the impact that all forms of aggression have on children over time. . ..Building public awareness for gender differences in the expression of aggression will lead to more sensitive and effective interventions for high-risk young females” (p. 358). Only three known empirically supported intervention studies have been conducted to reduce relational aggression. The first known empirically based intervention for relational aggression is the “Friend to Friend” (F2F) Program, which is a comprehensive school-based initiative that targets small groups of relationally aggressive girls in middle childhood. This program, administered primarily within the context of small friendship groups, attempts to decrease aggression and hostile attribution biases (i.e., inferring malicious intent in ambiguous situations), while increasing prosocial and social competence skills (Leff, Goldstein, Angelucci, Cardaciotto, & Grossman, 2007; Leff et al., in press). Initial research suggests this program is highly acceptable to teachers and children and is effective at reducing relational and physical aggression as well as hostile attribution biases among highly relationally aggressive girls (Leff et al., 2007, in press). The second program, the “Second Step, Violence Prevention Program” also focuses on changing attitudes toward physical and relational aggression and enhancing social skills. Specifically, this school-based curriculum attempts to teach social and emotional skills (e.g., empathy, behavioral control, problem-solving and anger management). With younger children, puppets are used and the length of lessons is significantly reduced. Overall, these studies have used a variety of methods, such as behavioral observations (Grossman et al., 1997) and structured conflict paradigms (Frey, Nolen, Edstrom, & Hirschstein, 2005) across a variety of developmental periods, including middle childhood (Grossman et al., 1997; Orpinas, Parcel, McAlister, & Frankowski, 1995) and adolescence (Van Schoiack-Edstrom, Frey, & Beland, 2002). Across these studies, Second Step has demonstrated effectiveness at decreasing physical aggression (Frey et al., 2005; Grossman et al., 1997) and reducing tolerance for physical and relational aggression (Van Schoiack-Edstrom et al., 2002), as well as increasing prosocial behavior (Grossman et al., 1997), social competence (Frey et al., 2005), and perceived use of social skills (Van Schoiack-Edstrom et al., 2002). Further, Grossman et al. (1997) demonstrated retention of these skills over a 6-month period. The third empirically based intervention to reduce relational aggression was developed by Harrist and Bradley (2003). These authors implemented Vivian Paley’s (1992) “You can’t say you can’t play” program in six randomly assigned kindergarten classrooms and four control classrooms. The “you can’t say you can’t play” rule was reinforced with fairy tales addressing social exclusion and inclusion, discussions of emotions, and the use of group role-plays. Children in the intervention rooms reported that they liked each other more at the end of the year than did children in the control rooms, yet they also reported higher levels of social dissatisfaction. However, observations and teacher-reports did not reveal any significant Author's personal copy 18 J.M. Ostrov et al. / Early Childhood Research Quarterly 24 (2009) 15–28 changes at the classroom level for social exclusion. In addition, the authors reported that children generally did not like the rule and followed the “letter of the law but not the spirit of the law” (Harrist & Bradley, 2003, p. 201), in that children would not explicitly tell their peers they could not play but they would treat them unfairly such as assigning them undesirable roles in socio-dramatic play. To date, only one known empirically based program has been developed for specifically addressing relational victimization (i.e., the frequent or chronic receipt of relational aggression from peers; Crick & Grotpeter, 1996). The Walk away, Ignore, Talk, Seek help (WITS) program, created by Leadbeater, Hoglund, & Woods (2003), was a school-based initiative (children were 6-years-old at the start of the 3-year project) to reduce school and classroom levels of relational victimization through a book-based curriculum that permitted teachers to implement the program through literature and activities within the preexisting lesson plans. This program appears to have moderate effects at decreasing victimization and increasing social competence (Leadbeater et al., 2003; Leadbeater & Hoglund, 2006). Collectively these programs demonstrate the efficacy of various types of intervention strategies with older children (i.e., kindergarten and older), but as yet do not target children 5-years-old and younger during early childhood. Our intervention improves on these existing programs by explicitly developing a brief, classroom-wide program for young children in early childhood centers that targets relational and physical aggression (and victimization), prosocial behavior (i.e., relational inclusion), and friendship formation skills. 4. Developing an intervention for subtypes of aggression in early childhood We have developed a preliminary version of a classroom-based preventive intervention (the “Early Childhood Friendship Project”) inspired by and loosely based on some elements (e.g., brief circle time format, use of puppets, participatory activities, targeted reinforcement, focus on both positive and negative behavior) of the “Incredible Years Dina Dinosaur Classroom” (Webster-Stratton et al., 2008) program, while focusing the curriculum more specifically on friendship-making skills (e.g., entering and joining playgroups, inviting peers to play) and issues related to relational aggression. Three hypotheses guided our study. First, we hypothesized the intervention classrooms would show a greater decrease in physical and relational aggression, relative to the control classrooms. Second, we hypothesized that intervention classrooms would show a greater decrease from preto post-test for both physical and relational victimization, compared to control rooms. Third, we hypothesized that intervention classrooms, compared to control classrooms, would show a larger increase over time in prosocial behavior. To test these hypotheses a 6-week curriculum was developed for a sample of young children.