Applied Developmental Psychology 27 (2006) 254 – 268 A longitudinal study of relational and physical aggression in preschool Nicki R. Crick a,⁎, Jamie M. Ostrov b , Jean E. Burr a , Crystal Cullerton-Sen a , Elizabeth Jansen-Yeh a , Peter Ralston a a Institute of Child Development, University of Minnesota, Twin-Cities Campus, 51 East River Road, Minneapolis, MN 55455, United States b Department of Psychology, University at Buffalo, The State University of New York, United States Available online 6 March 2006 Abstract To understand the development of relational aggression during early childhood, 91 girls and boys (M age = 39.0; SD = 7.6 months) and their teachers participated in an 18-month longitudinal study. Children were observed for relational and physical aggression during free play in four time periods. Individually administered interviews were conducted to provide peer reports of relational and physical aggression. Teachers completed measures of relational and physical aggression and peer rejection. Findings support the psychometric properties of the observational methods for use during early childhood. Results suggest that girls are more relationally aggressive than male peers and boys are more physically aggressive than female peers. Moreover, children primarily direct their aggressive behavior at same-sex peers. Finally, relational aggression was found to be moderately stable during early childhood and was associated with future peer rejection problems. Results are discussed in terms of the importance of developing methods to investigate behavior patterns for understanding the early development of and future social–psychological risks that may be associated with relational aggression. © 2006 Elsevier Inc. All rights reserved. Keywords: Relational aggression; Physical aggression; Gender; Sex differences; Early childhood; Longitudinal patterns 1. Introduction Due to its deleterious effects on children's development, peer-directed aggression has been one of the most widely studied adjustment problems in the past several decades. Although hundreds of studies, books, and journals have been dedicated to the topic, past investigations of aggressive behavior have been limited in two crucial ways: (1) aggressive boys have received most of the empirical attention, whereas aggressive girls have often been excluded from relevant studies; and (2) forms of aggression that are salient to boys have been emphasized (e.g., physical aggression) whereas forms that are salient to girls have been neglected (e.g., relational aggression; Bjorkqvist & Niemela, 1992; Crick & Grotpeter, 1995; Robins, 1986). Compounding these limitations, current prevailing theories of the development of aggression depict the behavioral problems of girls as virtually nonexistent until the onset of adolescence (Keenan & Shaw, 1997; Silverthorn & Frick, 1999). Due to these empirical and theoretical shortcomings, we currently know much less about aggressive girls than aggressive boys, and we particularly lack knowledge of young girls' behavior ⁎ Corresponding author. Tel.: +1 10 612 624 3347. E-mail address: crick001@umn.edu (N.R. Crick). 0193-3973/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.appdev.2006.02.006 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 255 problems. Given the negative risk status typically associated with aggression and given the numerous advantages afforded by early intervention (Bloomquist & Schnell, 2002), this lack of knowledge is significant. In one attempt to “suspend our acceptance of the mythology of more benign childhoods for girls” (Zahn-Waxler, 1993, p. 84), a relational form of aggression has been recently identified that has been shown to be more characteristic of girls than the physical forms that have captured the majority of previous empirical and theoretical efforts (for a review see Crick et al., 1999). In contrast to physical aggression, in which physical damage or the threat of physical damage serves as the means of harm, relational aggression includes behaviors in which damage to relationships or the threat of damage to relationships serves as the vehicle of harm (Crick & Grotpeter, 1995). Relational aggression includes both direct and indirect acts such as threatening to end a friendship unless the friend complies with a request, using social exclusion or the “silent treatment” as retaliatory behaviors, and spreading false rumors to encourage peers to reject a classmate. One fundamental question that remains unanswered in studies of relational aggression concerns the early developmental course of relational aggression for boys versus girls. This information is crucial for the generation of theories regarding the etiology of relationally aggressive behavior patterns. Unfortunately, however, investigators interested in these issues have been limited by a lack of reliable and valid observational methods for assessing relational aggression. Researchers have primarily relied upon teacher and peer assessment tools that, although quite useful for addressing a number of empirical questions, may also be subject to a variety of significant biases (e.g., they may be influenced by explicit or implicit stereotypes about males and females, see Pellegrini, 2001b; Susser & Keating, 1990). Recently, however, a naturalistic observational approach for assessing relational aggression among preschoolers has been developed that appears both reliable (e.g., inter-rater reliability of .82) and valid (e.g., observed relational aggression scores and teacher-based scores are associated r = .54; Ostrov & Keating, 2004). 1.1. Observations of relational aggression Our first goal for the present study was to evaluate further the psychometric properties of the observational approach developed by Ostrov and Keating (2004) by employing it in the first prospective study of relational and physical aggression in young children. Specifically, we examined the predictive validity of children's observationally based relational aggression scores derived from this scheme as well as correspondence with other informants (i.e., teachers and peers), and the association between relational aggression and physical aggression. This information is necessary for establishing the appropriateness and generalizability of the Ostrov and Keating approach for use with a variety of different types of samples and in prospective investigations. After establishing the favorable psychometric properties of the Ostrov and Keating observational scheme for the present sample, the second goal of this research was to use the scheme to evaluate sex differences in relational and physical aggression during early childhood. Some but not all recent studies have demonstrated that, in sharp contrast to current theories and empirical investigations of aggression that largely depict girls as lacking in behavioral problems prior to adolescence, a significant proportion of girls can be identified as highly aggressive as early as the preschool years if relational aggression is assessed in addition to physical aggression (e.g., Bonica, Yershova, Arnold, Fisher, & Zeljo, 2003; Crick, Casas, & Mosher, 1997; McNeilly-Choque, Hart, Robinson, Nelson, & Olsen, 1996; Ostrov & Keating, 2004). In addition to studies conducted in the United States, cross-cultural research has demonstrated the particular importance of relational aggression for identifying preschool age aggressive girls in other countries including Russia, China, and Australia (see studies by Hart and co-workers, such as Hart, Nelson, Robinson, Olsen, & McNeillyChoque, 1998; Russell, Hart, Robinson, & Olsen, 2003). Studies of middle childhood have yielded similar results. For example, in one study, 4.2% of the participating girls were identified as aggressive when only physical aggression was taken into account whereas 21.6% were classified as aggressive when both relational and physical aggression were considered (Crick & Grotpeter, 1995). In this instance, a focus solely on physical aggression would have failed to identify over 80% of aggressive girls but would have failed to identify only 7% of the aggressive boys. These studies provide robust evidence to counter the stereotypical view that girls, in general, are not aggressive and that young girls do not exhibit or experience behavioral problems. 1.2. Sex differences in relational and physical aggression Although it is relatively clear that the study of relational aggression adds significantly to our ability to identify and understand aggressive females, less clarity has been achieved regarding sex differences in children's propensity to use 256 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 relational aggression. Mixed findings particularly abound for studies of early childhood. Specifically, studies in which teacher reports of relational aggression have been utilized have tended to show that preschool-age girls are more relationally aggressive than preschool age boys (cf. Hart et al., 1998), whereas studies that have relied on peer reports have often failed to find significant sex differences (e.g., Bonica et al., 2003; Sebanc, 2003; for a review see Crick, Ostrov, & Kawabata, in press-a). In the one study to utilize a reliable and valid naturalistic observational approach, preschool girls were shown to be significantly more relationally aggressive than preschool boys (Ostrov & Keating, 2004). These discrepant findings may be due to a number of factors including reporter bias (for teacher and peer reports; for further discussion of this issue see Pellegrini, 1996, 2001b) or possible changes in sex differences during the preschool years that have not been taken into account in existing studies due to their cross-sectional designs. These issues were addressed in the current study through the use of an assessment method (i.e., naturalistic observation) that is less prone to bias than those employed in the majority of most previous studies (i.e., teacher and peer reports) and through the use of a longitudinal design in which preschoolers were followed for approximately 18 months. We tested the association between physical and relational aggression and based on past findings (e.g., Bonica et al., 2003; Crick et al., 1997; Ostrov & Keating, 2004) we predicted that these constructs would be only slightly associated at each time point using observational methods that are based on information from multiple independent observers. We also predicted that observational assessments would be moderately associated with teacher report methods. We hypothesized that preschool age girls would be significantly more relationally aggressive than preschool age boys. We further predicted, in keeping with the gender-segregated nature of early childhood play settings (Maccoby, 1998) and past findings (Ostrov & Keating, 2004) that relational aggression would be more frequently directed to female peers than to male peers. 1.3. Stability of aggression subtypes Our third objective was to provide the first examination of the stability of relational aggression in early childhood. Existing evidence indicates that individual differences in physical aggression may be relatively stable beginning at about two years of age (e.g., Fagot, 1984; Rose, Rose, & Feldman, 1989; Zahn-Waxler, Ianotti, Cummings, & Denham, 1990). This stability tends to persist into childhood (Cairns, Cairns, Neckerman, Ferguson, & Gariepy, 1989; Coie & Dodge, 1998; Huesmann, Eron, Lefkowitz, & Walder, 1984; Olweus, 1979; Pulkkinen, 1992; Yoshikawa, 1994) and is most apparent from early childhood to third grade for a small group of extremely aggressive children (NICHD ECCRN, 2004). Thus, children who exhibit relatively high levels of physically aggressive behavior during the early preschool years are at risk for continued behavior problems, a trajectory that has been shown to be associated with a host of serious developmental difficulties. Studies of relational aggression indicate a relatively high degree of stability for individual differences for older children (i.e., 9–12 year olds; Cillessen & Mayeux, 2004; Crick, 1996; Crick, Ostrov, and Werner, in pressb; Tomada & Schneider, 1997; Zimmer-Gembeck, Geiger, & Crick, 2005); however, no research has yet been conducted on the stability of relational aggression among young children using naturalistic observations.1 Given the importance of early detection of aggressive behavior patterns in the prevention and treatment of future, as well as concurrent, adjustment difficulties (Levy-Shiff & Hoffman, 1989; Wasik, 1987), this lack of information has significant negative implications for our ability to adequately identify young children at risk. Given initial evidence regarding the greater prevalence of relational aggression among preschool girls (Ostrov & Keating, 2004), this lack of knowledge about the stability of relational aggression during the preschool period has likely had the greatest impact on our understanding of and our ability to appropriately serve young females. In this investigation, we predicted that individual differences in relational, as well as physical, aggression would be relatively stable during the preschool years. 1.4. Peer rejection and aggression subtypes Our fourth and final objective for this study was to examine the concurrent and future associations among relational aggression, physical aggression, and peer rejection during early childhood. Evidence is mounting that relational aggression is associated with social–psychological adjustment problems including social maladjustment, internalizing problems, and externalizing difficulties (for a review see Crick et al., 1999). This association has been demonstrated for 1 Preliminary data with a small preschool sample indicated that relational aggression scores obtained in a semi-structured task were significantly predictive of future naturalistically observed relationally aggressive behavior (Ostrov et al., 2004). N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 257 preschoolers (e.g., Crick et al., 1997; McNeilly-Choque et al., 1996; Ostrov, Woods, Jansen, Casas, & Crick, 2004), for children in middle childhood (e.g., Blachman & Hinshaw, 2002; Crick, 1996, 1997; Crick et al., in press-a, b; Rys & Bear, 1997; Zalecki & Hinshaw, 2004), for adolescents (e.g., MacDonald & O'Laughlin, 1997; Prinstein, Boergers, & Vernberg, 2001), and for adults (e.g., Linder, Crick, & Collins, 2002; Werner & Crick, 1999). Further, this relation has been shown to hold for non-U.S. as well as U.S. samples including Italian, Russian, Chinese, and German samples (e.g., Hart et al., 1998; Tomada & Schneider, 1997; for a review see Crick et al., 1999). Although longitudinal studies are few in number, available evidence indicates that relational aggression predicts future as well as concurrent social–psychological adjustment difficulties (Crick, 1996; Crick et al., in press-a, b; Werner & Crick, 2004; Zimmer-Gembeck et al., 2005). For example, relational aggression in third grade significantly predicts negative changes in peer rejection (i.e., becoming more rejected) three years later in sixth grade, even after controlling statistically for physical aggression, peer status, and prosocial behavior (Zimmer-Gembeck et al., 2005). Taken together, findings from these studies generally demonstrate that relationally aggressive behavior patterns may place children at risk for serious future difficulties. Further, they indicate that a singular focus on physical aggression may seriously hamper our comprehensive understanding of these difficulties. Thus far, prospective studies of relational aggression and adjustment have focused exclusively on children in middle childhood and early adolescence. Consequently, the utility of assessing relational aggression for identifying young children “at risk” for future social–psychological problems is currently unknown. We addressed this gap in the present research. Specifically, we hypothesized that relational aggression would significantly predict future peer rejection for preschool-age boys and girls. Although future associations have been evaluated for school age children (e.g., Crick, 1996; Crick et al., in press-a, b; Zimmer-Gembeck et al., 2005), studies of preschoolers have been limited to concurrent associations (e.g., Crick, Casas, & Mosher, 1997). In this research we hypothesized that relational aggression would provide information about future peer rejection primarily for preschool-age girls. To address these objectives, we followed prospectively a sample of preschool age children for approximately one and a half years. Assessments of aggression and peer rejection were conducted at four time periods for each child (i.e., beginning of school year 1, end of school year 1, beginning of school year 2, and end of school year 2). 2. Method 2.1. Participants A total of 91 children (52 girls; 39 boys) between the ages of 30 and 52 months (M = 39.0; SD = 7.6) who were recruited from two nationally accredited university affiliated preschools in a large Midwestern city participated in this research. Written parental consent and verbal child assent were required for each child's participation. During the first round of recruitment, the overall consent rate between the two sites was 73%. All subsequent recruitment occurred under blanket consent procedures and therefore was 100%. Demographic data on the socioeconomic status and ethnicity of participants was collected at a single time point from one site when the consent rate reached 100%. Given the nature of the sample, these figures are representative of both sites and all data collection time points. Sixty-five percent of participants were European American, 21% were Asian American, 10% were African American, and 4% were of other ethnicity. Twenty-nine percent of children's families were known to be living 150% below the federal poverty level. Twenty-three percent of children were learning English as a second language. Children were recruited in two cohorts (Cohort 1, n = 37; Cohort 2, n = 54). There were no differences in demographic information or other characteristics relevant to the present study between the cohorts, which were recruited from the same schools and classrooms one year apart. For all analyses the two cohorts were combined and treated as one sample. 2.2. Procedures Each child was evaluated twice over the course of two consecutive school years (i.e., four time points). The first assessment was conducted during the fall semester, the second during the spring semester, the third during the following fall, and the fourth took place during the following spring. Assessments occurred approximately 4–6 months apart. During each of the four time periods (Time 1, 2, 3, and 4) children's relational and physical aggression and peer rejection were assessed using naturalistic observations (assessment of aggression), individually administered child interviews (assessment of aggression), and teacher-reported rating scales (assessment of children's aggression and peer rejection). Each assessment method is described in further detail in subsequent sections of the method. 258 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 Typically, children remained in the same classroom throughout each school year. At the start of time 3, children typically transitioned to a new classroom and/or a new teacher. At site 2, in keeping with school policy that strongly discouraged direct research contact with participants, children only participated in the observational and teacher report components and did not receive an individual interview. Each participating school received an honorarium to be used within the school for academic and/or staff development purposes. 2.3. Measures 2.3.1. Observations of relational and physical aggression Prior to data collection at each time point, observers spent considerable time in each participating classroom and on the playground to facilitate children's acclimation to their presence (Pellegrini, 1996). Many of the observations in the classrooms were conducted from a visually shielded observation booth attached to the room in order to diminish children's reactivity to the presence of the observers. However, observers also conducted observations in the classroom in order to hear and see the full peer interactions. When observers were in the presence of the children, in the classroom and always on the playground, observers used a “minimally responsive manner” (Pellegrini, 1996) and were specifically trained in controlling their movements, body language and nonverbal behavior to further diminish possible reactivity. Assessments of reactivity (frequency of looks, comments, questions to the observer) noted by observers when in the presence of children were low to nonexistent over the course of the study (less than 5% of the time; Atlas & Pepler, 1998). Order of children observed was determined randomly within each day of observation and no child was observed more than one time per day. Naturalistic observations were collected during each assessment period by a total of six trained graduate students/professional staff and 18 undergraduates. The majority of observers changed at each time period and rotated through different classrooms throughout the observation period. Both male and female graduate and undergraduate observers participated at each assessment period. Naturalistic observations of children's relational and physical aggression were conducted during free-play using an adaptation of procedures developed by Ostrov and Keating (2004). Using a focal child approach, each child was observed for 10 continuous minutes per assessment by a trained observer who was located in an unobtrusive position that was close enough to hear children's conversations. Over an eight-week period, each child was observed eight times (a total of 80 min per child per time point). In sum, 320 min of observation (5.33 h) was collected for each focal child at each time period, which equates to a total of 29,120 min (485.33 h) of observation time for the present sample across the study. As in prior research, the manner in which the observations were collected was not amendable to kappa coefficient indices of reliability because observers did not specifically record intervals when the behaviors of interest were not present. The use of Intra-Class Correlations (ICC) has been suggested for such cases (Bartko, 1976; McGraw & Wong, 1996) and has been used in past studies (e.g., Goldstein, Arnold, Rosenberg, Stowe, & Ortiz, 2001; NICHD ECCRN, 2004; Ostrov & Keating, 2004; Ostrov et al., 2004). Evidence for favorable interrater reliability of this observational measure has been demonstrated in past research (i.e., ICCs of .75 for physical aggression, and .82 for relational aggression; Ostrov & Keating, 2004). In the present investigation, training consisted of readings, discussions and coding of videotapes of children's aggressive behavior from past studies (Ostrov & Keating, 2004; Ostrov et al., 2004). Observers were required to reach an acceptable level of inter-rater reliability with the videotapes (ICCs N .80), pass a written multiple choice/matching exam (with discussion for any incorrect responses), and conduct several live practice reliability observations in the classroom and on the playground with the trainer. Assessments of reliability were conducted throughout the study to avoid observer drift problems (see Pellegrini, 1996). During every 10-min measurement interval, observers assessed the focal child's display of the following behaviors (including a full description of what occurred and the sex of all children involved) in each incident: a) physical aggression (e.g., hitting, shoving, pulling, taking objects); and b) relational aggression (e.g., excluding from an activity, using friendship withdrawal as a threat; covering ears to signal ignoring). Separate behaviors were recorded based on temporal breaks in the interactions during the observation. Observers, for purposes of a different study, collected several additional social behaviors (e.g., prosocial behavior, play behavior, victimization) not reported here. Each child was observed on eight separate occasions and behaviors noted during observations were summed to yield total behavior scores for each time period. The average number of instances of relational aggression and physical aggression per ten-minute session was calculated for each child at each time point. Observations were also summed separately by the sex of the recipient peer (i.e., relational aggression to a male or to a female peer). Interrater reliability was assessed N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 259 at each time point on 10–15% of observations, spread across each eight-week observation period. Interrater reliability was assessed by using two observers recording behavior on the focal child at the same time. The observers stood as far apart as possible and did not discuss their observations. Each child was observed for reliability and each observer was tested for reliability across the entire study. Reliability was acceptable at each time point for relational aggression (ICCs ranged from .77 to .91) and for physical aggression (ICCs ranged from .86 to .92). 2.3.2. Peer assessment of aggression Trained research assistants (graduate students or professional staff members) interviewed children individually for approximately 15 min in a quiet room/area in the preschool. Children provided verbal assent to participate prior to beginning the interviews. During the interviews, a peer report measure of children's relational and physical aggression was administered in addition to several other measures that were part of a larger project. At the close of the interview, children were thanked for their participation and escorted back to their classrooms. An amended version of a previously developed peer-nomination instrument was used to assess peer reports of children's use of relational and physical aggression (Preschool Social Behavior Scale-Peer Form; PSBS-P, Crick et al., 1997). This modified instrument consisted of 3 items, one depicting the use of physical aggression, a second for relational aggression, and a third for prosocial behavior. The prosocial items served as positively-toned filler items for the present study. In the original version of the measure, children are asked to nominate up to three of their peers who fit the behavioral description presented in each item. Preschoolers' responses to the PSBS-P items have been shown to be reliable with Cronbach's alphas greater than .70 for both the relational and physical aggression scales. Factor analyses have also confirmed the existence of distinct factors for relational and physical aggression (Crick et al., 1997). The measure was modified from a nomination to a rating format in the present investigation based on research suggesting that rating approaches may be more reliable (i.e., higher test–retest reliability) than nomination procedures with young children, as rating procedures require that each respondent provides data on all peers in the classroom (Asher & Dodge, 1996; Asher & Hymel, 1986; Asher, Singleton, Tinsley, & Hymel, 1979; Hymel, 1983; Olson & Lifgren, 1988). In addition, there is empirical evidence suggesting that the relation between sociometric nomination and rating methods is quite strong (see Bukowski & Hoza, 1989). During each interview, children began with several practice items, designed to help children learn the response format for the peer-rating items. Specifically, children were presented with three age-appropriate pictures of food or toys (a teddy bear, hammer, and camera; shown in past studies to generate a range of responses; Bauer, 1993). Children were asked to consider each item, one at a time, and indicate acceptance of it. For example, the interviewer first held up a small teddy bear and asked, “Do you like to play with the teddy bear, yes or no?” If children responded affirmatively, they were further prompted with a follow-up question, “yes — a little or yes—a lot?” Children were taught to sort the items into three boxes labeled with appropriate pictures, “no” with a frown face (0), “yes — a little” with a picture of a neutral face (scored 1), or “yes — a lot” with a picture of a smiling face (scored 2). Once children were familiar with the format of the ratings, they were asked to correctly identify each neutral expression, head and shoulder photo of their classmates (a method that has been commonly employed to elicit peer reports of social behavior and adjustment from young children, see Asher & Dodge, 1996; Asher et al., 1979; Denham et al., 2000; Hart et al., 2000; Hymel, 1983; Olson & Lifgren, 1988). Once children correctly identified each of their classmates, interviewers began to administer the ratings. Interviewers presented children with a photo of each classmate, one at a time, and asked if the child in the photo was physically aggressive (i.e., “Does Bobby hit, push or pinch other kids, yes or no?”), relationally aggressive (i.e., “Do you ever hear Suzie say, ‘You can't come to my birthday party, you can't play’, yes or no?”), and prosocial (i.e., “Does Pat share and help other kids, yes or no?”). If children responded affirmatively, they were further prompted with a follow-up question, “yes — a little or yes — a lot?” Children then placed the photos into appropriately labeled boxes, (i.e., a box with a red stop-sign indicating “no” and scored 0, a box with a yellow sign indicating “yes — a little bit” and scored 1, and a box with a green sign indicating “yes — a lot” scored 2).2 The average rating that each child received from his/her peers for each item on the PSBS-P instrument was calculated and then standardized within each classroom at each time point. 2 Traffic light colors were used here in order to avoid pairing a high amount of aggressive behavior with a smiling face. To make sure that the children understood the change in response colors we first used toy cars and a traffic light to demonstrate the differences between the colors and their associated meaning in the rating task (e.g., “green means I can move my car a lot, yellow means I can only move my car a little bit, red means stop, no go.”). 260 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 2.3.3. Teacher ratings Teachers completed several social behavior and peer rejection measures for each of their participating students at each time point. Only those measures assessing aggression and peer rejection) were used in the present investigation. Both written and verbal instructions for completing these instruments were given to teachers. Once completed, teachers were thanked and given an honorarium for their participation. 2.3.4. Teacher ratings of aggression The Preschool Social Behavior Scale-Teacher Form (PSBS-TF) developed in prior studies was used to assess teacher reports of children's aggression and prosocial behavior (Crick et al., 1997; Crick, Casas, & Ku, 1999). This instrument consists of 16 items, six of which assess relational aggression (e.g., “This child tries to get others to dislike a peer,” “This child tells a peer they won't be invited to their birthday party unless s/he does what the child wants”); six of which assess physical aggression (e.g., “This child kicks or hits others”); and four of which assess prosocial behavior (e.g., “This child is helpful to peers”). The prosocial items were included in this instrument for ethical and methodological reasons (e.g., to avoid asking teachers negative questions only; to avoid negative response sets) and served as positively-toned filler items in the present research. Teachers rated the degree to which children exhibited relational and physical aggression against their peers using a 5-point rating scale (1 = never or almost never true to 5 = always or almost always true). Evidence supports the favorable psychometric properties of the PSBS-TF (e.g., Bonica et al., 2003; Crick et al., 1997; Hart et al., 1998; Ostrov & Keating, 2004). Specifically, teachers' responses to this measure have been shown to be reliable, with Cronbach's alpha above .90 for both the relational and physical aggression scales. Also, factor analyses have confirmed the existence of distinct factors for relational and physical forms of aggression for several preschool samples. Similarly, in the present investigation, all sub-scales of this measure were reliable with Cronbach's alphas N .70. 2.3.5. Teacher ratings of peer rejection Teacher reports of children's peer rejection using the two-item peer rejection scale from the previously described PSBS-TF were used to assess children's rejection by peers (Crick et al., 1997; Crick et al., 1999). Teachers indicated the degree to which children were (1) rejected by same-sex peers (e.g., “This child is rejected by same-sex peers”) and (2) rejected by opposite-sex peers (e.g., “This child is rejected by opposite-sex peers”) using the same five-point response format described for the PSBS-TF behavioral items. Favorable psychometric properties for this subscale have been demonstrated in prior research (e.g., Ostrov et al., 2004). In the present study, alphas were above .70 at all time points. 3. Results In order to examine the study objectives, analyses were conducted to: a) evaluate the association between relational and physical aggression for each informant; b) examine the psychometric properties of the observational scheme used in this study; c) evaluate sex differences in relational aggression during early childhood; d) examine the stability of relational aggression in early childhood; and e) examine the associations among relational aggression, physical aggression, and future peer rejection during early childhood. 3.1. Association between relational and physical aggression Correlation coefficients were computed in order to examine the concurrent associations between relational and physical aggression for each informant. For girls, the correlations between observed physical and relational aggression were not significant at any of the four time points (r's ranged from .02 to .14). For boys, these correlations were significant at time 1(r = .39, p b .01) and time 2(r = .61, p b .001) but not at time 3(r =− .05, ns) or time 4(r = .26, ns). For girls, teacher reports of physical and relational aggression were significantly correlated at time 1(r = .68, p b .001), time 2(r = .36, p b .01), time 3(r = .43, p b .01) and time 4(r = .45, p b .001). For boys this association was also significant at time 1(r = .61, p b .001), time 2(r = .58, p b .001), time 3(r = .58, p b .001), and time 4(r = .69, p b .001). For girls, peer reports of physical and relational aggression were not significantly associated with one another at time 1(r = .13, ns), time 2(r = .14, ns), and time 3(r = .27, ns) but were significantly correlated at time 4(r = .67, p b .01). For boys, peer reports of physical and relational aggression were significantly associated at time 1(r = .68, p b .01) but were not at time 2(r = .28, ns), time 3(r = .58, ns), and time 4(r = .59, ns). N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 261 3.2. Psychometric properties of the observational measure of aggression 3.2.1. Concurrent validity We examined the concurrent validity of the observational scheme used in this study by conducting correlational analyses between observations of aggression and the teachers' ratings of relational and physical aggression (see Table 1). In general, for girls at each time point, observed relational aggression were significantly correlated with teacher-rated relational aggression with the exception that this association was not significant at time 2(p b .06). For boys, observers and teachers agreed significantly in their assessment of relational aggression at time 2. For girls' physical aggression observers' and teachers' assessments were significantly correlated at all time points except the last. For boys' physical aggression, observers and teachers agreed at all time points. At all time periods, correlations between peer reports of aggression and those of other informants did not significantly converge for either form of aggression (see Table 1). Thus, peer reports were not considered further in subsequent analyses. 3.2.2. Predictive validity To explore the predictive validity of the observation scheme the association between observed aggression and future teacher reports was evaluated. For girls, observed relational aggression at time 1 did not significantly predict teacher reports of relational aggression at time 4, r = .26, p b .08. For girls, time 2(r = .36, p b .05) and time 3(r = .42, p b .01) observed relational aggression significantly predicted teacher reported relational aggression at the end of the study (time 4). For boys, observed relational aggression at time 1(r = −.09, ns), 2(r = .18, ns), and 3(r = .05, ns) did not predict teacher reports of relational aggression at time 4. For girls, time 3(r = .53, p b .001) observed physical aggression significantly predicted time 4 teacher reports of physical aggression whereas time 1(r = .15, ns) and time 2(r = .19, ns) observations did not. For boys, time 1(r = .32, p b .05), time 2(r = .35, p b .01), and time 3(r = .45, p b .01) observed physical aggression significantly predicted time 4 teacher reports of physical aggression. 3.3. Sex differences in aggression To address our second objective, a series of repeated measures ANCOVAs using the children's observed aggression as the dependent variable were conducted. Age (in months) served as our covariate because of the relatively large range Table 1 Interrater agreement of relational and physical aggression for boys and girls across four time points Raters Boys Girls Physical Relational Physical Relational Peer and teacher Time 1 Time 2 Time 3 Time 4 .44 .32 .76⁎ −.05 .02 .29 .06 −.51 .47⁎ .69⁎ .17 .17 .10 .14 −.43 −.31 Peer and observerb Time 1 Time 2 Time 3 Time 4 .79⁎⁎⁎ −.05 .28 −.16 .35 .11 .51 .05 .30 .67⁎⁎⁎ .37 .18 .10 .30 −.09 .01 .21 .44⁎⁎ .24 .13 .55⁎⁎⁎ .51⁎⁎⁎ .56⁎⁎⁎ .13 a Observer and teacherc Time 1 Time 2 Time 3 Time 4 .44⁎⁎ .35⁎ .50⁎⁎⁎ .44⁎⁎ Sample sizes (n's) ranged from 7 – 25; bsample sizes (n's) ranged from 8 – 25; csample sizes (n's) ranged from 34 – 52. ⁎p b .05. ⁎⁎p b .01. ⁎⁎⁎p b .001. a .55⁎⁎⁎ .27 .44⁎⁎ .41⁎⁎ 262 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 of ages of the children in our sample. Given our a priori predictions and the relative independence of the two forms of aggression, we analyzed relational and physical aggression in separate models. 3.3.1. Relational aggression Mean (and SD) scores for relational aggression are presented in the upper portion of Table 2 for boys and girls and sex of peer recipients of aggression. A 2 (Focal child sex: boys and girls) × 2 (Peer recipient sex: male and female peers) × 4 (Time points) ANCOVA was conducted with the last 2 variables serving as within-participant factors and relational aggression serving as the dependent variable. Child's age in months at time 1 served as the covariate and thus the reported means are adjusted. A main effect for focal child sex emerged, F(1, 86) = 5.92, p b .05, η2 = .06. Girls (M = .12; SD = .01) were more relationally aggressive than boys (M = .08; SD = .02). A significant peer recipient sex × focal child sex interaction was obtained, F(1, 86) = 13.22, p b .001, η2 = .13. Simple effects tests revealed that girls were more relationally aggressive to female peers than were boys, F(1, 86) = 11.74, p b .001, η2 = .12 (see Table 2). No other significant effects emerged. 3.3.2. Physical aggression The mean (and SD) physical aggression scores are presented in the lower portion of Table 2 by time period and focal child sex. A 2 (Focal child sex: boys and girls) × 2 (Peer recipient child sex: male and female peers) × 4 (Time points) ANCOVA with repeated measures on the last 2 factors was conducted on observed physical aggression scores. Age (in months) at the first time point served as the covariate. A main effect for focal child sex was revealed, F(1, 85) = 22.05, p b .001, η2 = .20. Specifically, boys (M = .18; SD = .02) were more physically aggressive than girls (M = .07; SD = .02). A time × focal child sex interaction emerged, F(3, 86) = 2.65, p b .05, η2 = .03. Simple effect tests revealed that boys were more physically aggressive than girls at time 1, F(1, 87) = 14.13, p b .001, η2 = .14, at time 2, F(1, 87) = 17.10, p b .001, η2 = .16, and at time 3, F(1, 88) = 7.97, p b .001, η2 = .08 At time 4, boys were not significantly more physically aggressive than girls, F(1, 88) = 3.14, p b .08, η2 = .03. (see means in Table 2). A peer recipient sex × focal child sex interaction was revealed, F(1, 86) = 25.39, p b .001, η2 = .25. Simple effects tests revealed that boys (M = .95; SD = .10) were more physically aggressive to male peers than were girls (M = .22; SD = .09), F(1, 86) = 30.81, p b .001, η2 = .26. No other significant effects emerged. We next examined the extent to which aggressive girls would have been unidentified if relational aggression had not been assessed in the current study. Specifically, we identified extreme groups of aggressive girls (i.e., those who were physically aggressive only, relationally aggressive only, and physically plus relationally aggressive) using a criterion of one standard deviation above the sample mean, a commonly used criterion in previous research (e.g., Crick, 1997; Crick & Grotpeter, 1995). The findings indicate that without the inclusion of relational aggression we would have failed to identify 42% (n = 5) of the more highly aggressive girls at time 1, 36% (n = 4) of the more highly aggressive girls at time 2, 50% (n = 5) of the more highly aggressive girls at time 3, and 46% (n = 6) of the more highly aggressive girls at time 4. Table 2 Descriptive statistics for aggression subtypes Peer recipient sex Relational aggression Girls Male Female Boys Mean SD Mean SD 1.05 2.12 .21 .25 1.28 .83 .24 .28 Physical aggression Girls Time 1 2 3 4 Boys Mean SD Mean SD .15 .11 .19 .14 .05 .04 .05 .03 .45 .35 .41 .22 .15 .05 .19 .14 Note. Means are adjusted (age at time 1 served as the covariate). Means for relational aggression are composites across time. N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 263 3.4. Stability of aggression To address our third objective, examination of the stability of relational and physical aggression across the four time points, a series of bivariate correlations were computed using observed and teacher reports of aggression at times 1 and 4 (computed separately by focal child sex). 3.4.1. Relational aggression For girls, observed relational aggression was stable from time 1 to 4, r = .39, p b .01. Teachers' ratings of girls' relational aggression were not significantly associated, from time 1 to 4, r = .28, p b .07. For boys, observed relational aggression was not stable across the study, r = .07, ns. In addition, teachers' ratings of boys' relational aggression was not significantly stable across the study, r = .22, ns. 3.4.2. Physical aggression For girls, observed physical aggression was stable from time 1 to time 4, r = .42, p b .01. For girls, teacher reported physical aggression was not stable from time 1 to 4, r = .17, ns. For boys, observed physical aggression was not significantly stable from time 1 to 4, r = .24, ns. For boys, teacher reported physical aggression was stable across the study, r = .46, p b .01. 3.5. Association between aggression and peer rejection To address the fourth objective, we examined the association between relational and physical aggression and peer rejection using a composite (average across time) measure of teacher reported peer rejection. Teacher reports were used here because the peer reports were not as reliable at each of the time points and teachers have been shown to be accurate informants concerning sociometric status during early childhood (see Wu, Hart, Draper, & Olsen, 2001). An average score was calculated based on each child's score on the peer rejection items of the PSBS-TF at times 2, 3, and 4. This procedure yielded an acceptable internal consistency alpha for this composite (α = .77). Teachers' ratings of time 1 rejection were not included because the main question was concerned with the prediction of future peer rejection. Given our a priori predictions, analyses were conducted separately for girls and boys. 3.5.1. Relational aggression For girls, observed relational aggression at time 1 significantly predicted future teacher reported peer rejection, r = .25, p b .05. For boys, observed relational aggression at time 1 did not significantly predict future peer rejection, r = .24, p b .07. 3.5.2. Physical aggression For girls, observed physical aggression at time 1 did not predict future peer rejection, r = .12, ns. For boys, observed physical aggression at time 1 did significantly predict future teacher reported peer rejection, r = .46, p b .01. 3.5.3. Unique association between relational aggression and peer rejection Finally, to test the effects of relational aggression at time 1 in predicting peer rejection above and beyond the effects of physical aggression at time 1, a hierarchical multiple regression was computed with age entered at step 1, observed physical aggression at time 1 entered at step 2, and observed relational aggression at time 1 entered at step 3. Teacher reported peer rejection at time 4 served as the dependent variable because time 4 peer rejection is a more conservative estimate of future peer rejection. In keeping with prior analyses and our a priori predictions, the equations were computed separately for girls and boys. The model results approached significance only for girls, ΔF(1, 47) = 3.75, p b .06, Δr2 = .07, indicating that it might be fruitful to investigate relational aggression (β = .29) as potentially accounting for variance in the prediction of future peer rejection for girls even after controlling for age and physical aggression. 4. Discussion This research significantly extends our knowledge of aggression by providing the first prospective look at relational aggression during early childhood. Findings indicated general support for our hypotheses. First, findings from this study 264 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 provided additional evidence for the reliability and concurrent validity of the observational scheme utilized (from Ostrov & Keating, 2004) and, more importantly, provided the first evidence for the favorable predictive validity of this scheme (this was particularly true for relational aggression for girls and physical aggression for boys). This observational scheme had never been used before with participants as young as 2-1 / 2 years of age and the present results indicate that observers can reliably observe both physical and relational aggression among very young preschoolers. To date, findings regarding sex differences in relational aggression during early childhood have been mixed with some studies demonstrating that girls are more relationally aggressive than boys and other studies finding no sex differences (for a review see Crick, Ostrov, Appleyard, Jansen, & Casas, 2004). Typically, obtained sex differences have varied as a function of informant type. The most consistent sex differences have been obtained with observational methods, (McNeilly-Choque et al., 1996; Ostrov & Keating, 2004; Ostrov et al., 2004; cf. McEvoy et al., 2003), an assessment approach that is less likely to be influenced by gender stereotypes and other types of informant biases than teacher- or peerreport methods. As a result, observational methods are commonly considered to be the “gold standard” for objective measurement of social behavior. This study indicates agreement across the informants but suggests that peer informants may not be as valid as observer or even teacher reports of young children's aggressive behavior. Future research is needed to clarify the utility of these methods. The present results add to the small, yet growing, body of evidence that girls are indeed significantly more relationally aggressive with peers than boys when relatively objective assessments are employed. Findings regarding sex differences also indicated that girls were more relationally aggressive with female peers than were boys whereas boys were more physically aggressive to male peers than were girls. These findings are not surprising given the gender-segregated nature of the early childhood play milieu and are consistent with sexual selection theory (Maccoby, 1998; Ostrov & Keating, 2004; Pellegrini & Long, 2003). That is, given that young girls are more likely to be relationally aggressive than boys in general and also given that young girls' social interaction partners are most likely to be girls, it stands to reason that girls would be more likely than boys to target their relationally aggressive behaviors toward other girls. Similarly, given that boys are more likely to be physically aggressive during the preschool years relative to girls and that preschool boys' interaction partners are most likely to be boys, it is not surprising that boys would be more likely than girls to target their physically aggressive acts toward other boys. Although rarely pursued in past studies, the present results highlight the importance of assessing the sex of the aggressor as well as that of the target in future research, which often can only be studied using systematic observations (Pellegrini, 2001a). Such information may be particularly informative for understanding the nature and development of bully-victim dyads and interactions. These findings further suggest that relationally aggressive behaviors directed at male peers and physically aggressive behaviors directed at female peers are gender atypical during early childhood and may be associated with greater adjustment problems. This point should be addressed in future research. Evaluation of the utility of physical versus relational aggression for identifying aggressive girls showed that we would have failed to identify between 36% and 50% of aggressive young girls (depending on the assessment period) if we had not included assessments of relational aggression in this study. These results are consistent with those obtained for older, elementary-school children (e.g., Crick & Grotpeter, 1995) and they add to existing studies by pointing out the significance of the study of relational aggression for identifying aggressive girls at much younger ages. The present findings provide the first evidence that individual differences in observed and teacher-rated relational aggression are moderately stable across an 18-month time period in early childhood. These findings are impressive given that different independent observers and teachers contributed data at these different time periods. Interestingly, the stability of relational aggression in this sample was obtained for girls, but not for boys. This suggests that relationally aggressive behavior patterns may begin to crystallize at relatively young ages for girls. If so, it may be particularly important for the healthy development of females for prevention and intervention efforts targeting relational aggression to be initiated during early childhood. Future studies are needed that evaluate when (or if ) individual differences in boys' relational aggression becomes stable over time and also whether individual differences in girls' relationally aggressive behaviors, apparent during preschool, remain stable during the early school years (e.g., during the transition to kindergarten). In contrast to relational aggression, findings for individual differences in physical aggression, depending on the rater, indicated that both girls and boys demonstrated some stability in these behaviors during the preschool years. The future social–psychological risk associated with young children's engagement in relational versus physical aggression was examined for the first time in the present research. Results showed that observed relational aggression at the beginning of the study predicted peer rejection 1-1 / 2 years later for girls and tended to for boys. In contrast, observed physical aggression at the beginning of the study predicted peer rejection 1-1 / 2 years later for boys, but not N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 265 for girls. Further, the findings suggest relational aggression should be further examined as a predictor of future peer rejection above and beyond physical aggression for girls. This is the first evidence that study of relational aggression may be particularly important for understanding social–psychological risk for young girls whereas physical aggression tells most of the story for young boys, findings that are consistent with prospective results obtained with older children (e.g., Crick, 1996; Crick et al., in press-a, b) and concurrent results obtained with young children (Crick et al., 1997). Future studies are needed that expand the range of social–psychological adjustment outcomes assessed during early childhood in keeping with research conducted with older children showing that relational aggression is associated with a multitude of adjustment problems including ADHD (e.g., Blachman & Hinshaw, 2002; Zalecki & Hinshaw, 2004), borderline personality disorder features (e.g., Crick, Murray-Close, & Woods, 2006), disruptive, oppositional behavior (Hipwell, Loeber, Stouthamer-Loeber, Keenan, White, & Krone, 2002; Prinstein et al., 2001) and internalizing problems (Murray-Close, Ostrov, & Crick, in press). In addition, future theory and research is needed to explore if there are specific and unique adjustment outcomes for relational aggression. Future research is also needed to explore the processes by which relational aggression leads to psychopathology. It will also be important to investigate what level of relational aggression is considered harmful across development and what other developmental processes (e.g., theory of mind, delay of gratification, impulsivity) may mediate or moderate these effects. Finally, the present findings further indicate the important role that physical aggression plays in the lives of young children, particularly boys (see Campbell, Shaw, & Gilliom, 2000; NICHD ECCRN, 2004). Clearly, it will be important to assess both relational and physical subtypes of aggression in future studies targeting the developmental outcomes of early childhood aggression (for ideas consistent with this see NICHD ECCRN, 2004). Given the recent literature on bistrategic children (i.e., children that use both prosocial and coercive behaviors, see Hawley, 2003) future research should evaluate the moderating role that prosocial behavior may play in the association between relational aggression and peer rejection and other social–psychological adjustment problems. 4.1. Limitations Despite the innovative and time intensive nature of our research design, several limitations exist. First, our sample size, although large relative to past observational studies (e.g., McEvoy et al., 2003; Ostrov & Keating, 2004), may have been insufficient for testing various effects. Future research with larger samples will be needed to replicate and extend the present longitudinal findings. In addition, despite having multiple informants (i.e., teacher, peer, and observer), we do not have any indication of how the participants in our study behaved at home with children in the neighborhood, with their siblings or parents. Recent studies have documented the importance of the sibling (Stauffacher and DeHart, 2005) and friendship (Burr, Ostrov, Jansen, Cullerton-Sen, & Crick, 2005; Johnson & Foster, 2005; Sebanc, 2003) relationship for the development of relational aggression during early childhood. Clearly, additional research is needed to explore the role of relational aggression within these additional close relationship contexts. Third, our findings with this university-based sample may not generalize to children from low SES backgrounds. However, in contrast to the physical aggression literature (see Dubow & Ippolito, 1994) past research has provided evidence that children from higher SES backgrounds and presumably some social–cognitive advantages (e.g., language capacity) may be at greater risk for displaying relationally aggressive behavior (see Bonica et al., 2003; McNeilly-Choque et al., 1996). Therefore, it is important to continue to study the role of SES in future studies during early childhood. Fourth, the age range in the present study was large and did not permit a more specific investigation of the potential qualitative differences between a 3-year-old and a 5-year-olds display of relational aggression. Future research should include a more restricted age range to overcome this issue. A final limitation is that our study, although longitudinal in nature, may not have started prior to the onset of the relationally aggressive behavior; future research will be needed to explore the specific timing and/or onset of these behaviors during toddlerhood. Moreover, in the present study we do not report on the children's development beyond age 5 and more information is needed to understand how relational and physical aggression exhibited during early childhood are predictive of academic and social functioning during the important transition into kindergarten (Rimm-Kaufman, Pianta, & Cox, 2000). 4.2. Conclusion In conclusion, based on over 29,000 min of observation on 91 preschoolers we have documented that relational aggression between peers may be reliably and validly observed in early childhood free play contexts. This study helps 266 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 to clarify past mixed findings in the developmental literature, by documenting sex differences in relationally aggressive behavior using a relatively objective assessment tool (naturalistic observations). Additional replication of these sex differences are needed using multiple informants and methods across time. Future research is needed to explore how older preschoolers may model and reinforce relationally aggressive behaviors for younger children in multi-age preschool classrooms and in home contexts (i.e., siblings). In the first test of stability during the early childhood period, relational aggression was found to be stable for girls. Relational aggression in peer relationships tended to be predictive of future peer rejection even when controlling for the effects of physical aggression, but only for girls. Developmental researchers must continue to study the role of relational aggression during early childhood, paying particular attention to the mechanisms of action and possible developmental antecedents of relational aggression for young children. Acknowledgment This study was funded by a grant (BCS-0126521) from the NSF to the first author. Preparation of this manuscript was facilitated in part by grants from NIMH (MH-63684) and NICHD (HD-046629) to the first author. The second author was funded in part by a NIMH (MH-15755) traineeship to the Institute of Child Development. The third author is now at Colby College. We thank the participating students, teachers, directors and children. We thank the entire Preschool PALS Project staff and special thanks to Alison Eudeikis for her assistance with the coordination of this project. References Asher, S. R., & Dodge, K. A. (1996). Identifying children who are rejected by their peers. Developmental Psychology, 22, 444−449. Asher, S. R., & Hymel, S. (1986). Coaching in social skills for children who lack friends in school. Social Work in Education, 8, 203−218. Asher, S. R., Singleton, L. C., Tinsley, B. R., & Hymel, S. (1979). A reliable sociometric measure for preschool children. Developmental Psychology, 15, 443−444. Atlas, R. S., & Pepler, D. J. (1998). Observations of bullying in the classroom. Journal of Educational Research, 92, 86−99. Bartko, J. J. (1976). On various intraclass correlation reliability coefficients. Psychological Bulletin, 83, 762−765. Bauer, P. J. (1993). Memory for gender-consistent and gender-inconsistent event sequences by twenty-five month old children. Child Development, 64, 285−297. Bjorkqvist, K., & Niemela, P. (1992). Of mice and women: Aspects of female aggression. San Diego, CA: Academic Press, Inc. Blachman, D. R., & Hinshaw, S. P. (2002). Patterns of friendship among girls with and without attention-deficit/hyperactivity disorder. Journal of Abnormal Child Psychology, 30, 625−640. Bloomquist, M. L., & Schnell, S. V. (2002). Helping children with aggression and conduct problems. New York, NY: Guilford Press. Bonica, C., Arnold, D. H., Fisher, P. H., Zeljo, A., & Yershova, K. (2003). Relational aggression, relational victimization, and language development in preschoolers. Social Development, 12(4), 551−562. Bukowski, W. M., & Hoza, B. (1989). Popularity and friendship: Issues in theory, measurement, and outcome. In T. J. Berndt & G. W. Ladd (Eds.), Peer relationships in child development (pp. 15−45). New York: John Wiley and Sons. Burr, J. E., Ostrov, J. M., Jansen, E. A., Cullerton-Sen, C., & Crick, N. R. (2005). Relational aggression and friendship during early childhood: “I won't be your friend!”. Early Education and Development, 16, 161−183. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Ferguson, L. L., & Gariepy, J. L. (1989). Growth and aggression: Childhood to early adolescence. Developmental Psychology, 25, 320−330. Campbell, S. B., Shaw, D. S., & Gilliom, M. (2000). Early externalizing behavior problems: Toddlers and preschoolers at risk for later maladjustment. Development and Psychopathology, 12, 467−488. Cillessen, A. H. N., & Mayeux, L. (2004). From censure to reinforcement: Developmental changes in the association between aggression and social status. Child Development, 75, 147−163. Coie, J. D., & Dodge, K. A. (1998). The development of aggression and antisocial behavior. In W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol.3. Social, emotional, and personality development (5th ed., 776–862). New York: Wiley. Crick, N. R. (1996). The role of overt aggression, relational aggression, and prosocial behavior in the prediction of children's future social adjustment. Child Development, 67, 2317−2327. Crick, N. R. (1997). Engagement in gender normative versus non-normative forms of aggression: Links to social–psychological adjustment. Developmental Psychology, 33, 610−617. Crick, N. R., Casas, J. F., & Ku, H. C. (1999). Relational and physical forms of peer victimization in preschool. Developmental Psychology, 35(2), 376−385. Crick, N. R., Casas, J. F., & Mosher, M. (1997). Relational and overt aggression in preschool. Developmental Psychology, 33(4), 579−588. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social psychological adjustment. Child Development, 66, 710−722. Crick, N. R., Murray-Close, D., & Woods, K. (2006). Borderline personality features in childhood: A short-term longitudinal study. Development and Psychopathology, 17, 1051−1070. N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 267 Crick, N. R., Ostrov, J. M., Appleyard, K., Jansen, E., & Casas, J. F. (2004). Relational aggression in early childhood: You can't come to my birthday party unless…. In M. Putallaz & K. Bierman (Eds.), Duke series in child development and public policy: Aggressive antisocial behavior, and violence among girls: A developmental perspective, Vol. 1 (pp. 71-89). New York: Guilford Publications. Crick, N. R., Ostrov, J. M., & Kawabata, Y. (in press). Gender differences in aggression and violence. In I. Waldman, D. J. Flannery, & A. T. Vazsonyi (Eds.), The Cambridge handbook of violent behavior. Cambridge University Press. Crick, N. R., Ostrov, J. M., & Werner, N.E. (in press). A longitudinal study of relational aggression, physical aggression and children's social– psychological adjustment. Journal of Abnormal Child Psychology. Crick, N. R., Werner, N. E., Casas, J. F., O'Brien, K. M., Nelson, D. A., Gropeter, J. K., et al. (1999). Childhood aggression and gender: A new look at an old problem. In D. Bernstein (Ed.), The Nebraska symposium on motivation: Gender and motivation, Vol. 45 (pp. 75-141). Omaha, NE: University of Nebraska Press. Denham, S. A., Workman, E., Cole, P. M., Weissbrod, C., Kendziora, K., & Zahn-Waxler, C. (2000). Prediction of externalizing behavior problems from early to middle childhood: The role of parental socialization and emotion expression. Development and Psychopathology, 12, 23−45. Dubow, E. F., & Ippolito, M. F. (1994). Effects of poverty and quality of the home environment on changes in the academic and behavioral adjustment of elementary school-age children. Journal of Clinical Child Psychology, 23, 401−412. Fagot, B. I. (1984). The consequents of problem behavior in toddler children. Journal of Abnormal Child Psychology, 12, 385−395. Goldstein, N. E., Arnold, D. H., Rosenberg, J. L., Stowe, R. M., & Ortiz, C. (2001). Contagion of aggression in day care classrooms as a function of peer and teacher responses. Journal of Educational Psychology, 93, 708−719. Hart, C. H., Nelson, D. A., Robinson, C. C., Olsen, S. F., & McNeilly-Choque, M. K. (1998). Overt and relational aggression in Russian nurseryschool-age children: Parenting style and marital linkages. Developmental Psychology, 34, 687−697. Hart, C. H., Yang, C., Nelson, L. J., Robinson, C. C., Olsen, J. A., Nelson, D. A., et al. (2000). Peer acceptance in early childhood and subtypes of socially withdrawn behavior in China, Russia, and the United States. International Journal of Behavioral Development, 24, 73−81. Hawley, P. H. (2003). Strategies of control, aggression, and morality in preschoolers: An evolutionary perspective. Journal of Experimental Child Psychology, 85, 213−235. Hipwell, A. E., Loeber, R., Stouthamer-Loeber, M., Keenan, K., White, H. R., & Krone-Man, L. (2002). Characteristics of girls with early onset disruptive and antisocial behaviour. Criminal Behaviour & Mental Health, 12, 99−118. Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20, 1120−1134. Hymel, S. (1983). Preschool children's peer relations: Issues in sociometric assessment. Merrill Palmer Quarterly, 29, 237−260. Johnson, D. R., & Foster, S. L. (2005). The relationship between relational aggression in kindergarten children and friendship stability, mutuality, and peer liking. Early Education and Development, 16, 141−160. Keenan, K., & Shaw, D. (1997). Developmental and social influences on young girls' early problems. Psychological Bulletin, 121, 95−113. Levy-Shiff, R., & Hoffman, M. A. (1989). Social behavior as a predictor of adjustment among three-year-olds. Journal of Clinical Child Psychology, 18, 65−71. Linder, J. R., Crick, N. R., & Collins, W. A. (2002). Relational aggression and victimization in young adults' romantic relationships: Associations with perceptions of parent, peer, and romantic relationship quality. Social Development, 11, 69−86. Maccoby, E. (1998). The two sexes: Growing up apart, coming together. CA: Stanford. MacDonald, C., & O'Laughlin, E. (1997, April). Relational aggression and risk behaviors in middle school students. Poster presented at the biennial meeting of the Society for Research in Child Development, Washington, DC. McEvoy, M. A., Estrem, T. L., Rodriguez, M. C., & Olson, M. L. (2003). Assessing relational and physical aggression among preschool children: Inter-method agreement. Topics in Early Childhood Special Education, 23, 53−63. McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intra-class correlation coefficients. Psychological Methods, 1, 30−46. McNeilly-Choque, M. K., Hart, C. H., Robinson, C. C., & Olsen, S. F. (1996). Overt and relational aggression on the playground: Correspondence among different informants. Journal of Research in Childhood Education, 11(1), 47−67. Murray-Close, D., Ostrov, J. M., & Crick, N. R. (in press). A short-term longitudinal study of growth of relational aggression during middle childhood: Associations with gender, friendship intimacy, and internalizing problems. Development and Psychopathology. NICHD Early Child Care Research Network (2004). Trajectories of physical aggression from toddlerhood to middle childhood. Monographs of the Society for Research in Child Development, 69 (4, Serial No. 278). Olson, S. L., & Lifgren, K. (1988). Concurrent and longitudinal correlates of preschool peer sociometrics: Comparing rating scales and nomination measures. Journal of Applied Developmental Psychology, 9, 409−420. Olweus, D. (1979). Stability of aggressive reaction patterns in males: A review. Psychological Bulletin, 86, 852−875. Ostrov, J. M., & Keating, C. F. (2004). Gender differences in preschool aggression during free play and structured interactions. Social Development, 13, 255−277. Ostrov, J. M., Woods, K. E., Jansen, E. A., Casas, J. F., & Crick, N. R. (2004). 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, 19, 355−371. Pellegrini, A. D. (1996). Observing children in their natural worlds: A methodological primer. Mahwah, NJ: Erlbaum. Pellegrini, A. D. (2001). A longitudinal study of heterosexual relationships, aggression, and sexual harassment during the transition from primary through middle school. Journal of Applied Developmental Psychology, 22, 119−133. Pellegrini, A. D. (2001). Sampling instances of victimization in middle school: A methodological comparison. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 125−144). New York: The Guilford Press. Pellegrini, A. D., & Long, J. D. (2003). A sexual selection theory analysis of sexual segregation and integration in early adolescence. Journal of Experimental Child Psychology, 85, 257−278. 268 N.R. Crick et al. / Applied Developmental Psychology 27 (2006) 254–268 Prinstein, M. J., Boergers, J., & Vernberg, E. M. (2001). Overt and relational aggression in adolescents: Social–psychological adjustment of aggressors and victims. Journal of Clinical Child Psychology, 30, 479−491. Pulkkinen, L. (1992). The path to adulthood for aggressively inclined girls. In K. Bjorkqvist & P. Niemela (Eds.), Of mice and women: Aspects of female aggression (pp. 113−121). San Diego, CA: Academic Press, Inc. Rimm-Kaufman, S. E., Pianta, R. C., & Cox, M. J. (2000). Teachers' judgments of problems in the transition to kindergarten. Early Childhood Research Quarterly, 15, 147−166. Robins, L. (1986). The consequences of conduct disorder in girls. In D. Olweus, J. Block, & M. Radke-Yarrow (Eds.), Development of antisocial and prosocial behavior: Research, theories, and issues (pp. 385−414). New York: Academic Press. Rose, S. L., Rose, S. A., & Feldman, J. F. (1989). Stability of behavior problems in very young children. Development and Psychopathology, 1, 5−19. Russell, A., Hart, C. H., Robinson, C., & Olsen, S. F. (2003). Children's sociable and aggressive behavior with peers: A comparison of the U.S. and Australia, and contributions of temperament and parenting styles. International Journal of Behavioral Development, 27, 74−86. Rys, G. S., & Bear, G. G. (1997). Relational aggression and peer relations: Gender and developmental issues. Merrill-Palmer Quarterly, 43(1), 87−106. Sebanc, A. M. (2003). The friendship features of preschool children: Links with prosocial behavior and aggression. Social Development, 12, 249−268. Silverthorn, P., & Frick, P. J. (1999). Developmental pathways to antisocial behavior: The delayed-onset pathway in girls. Development and Psychopathology, 11, 101−126. Stauffacher, K., & DeHart, G. B. (2005). Preschoolers' relational aggression with siblings and with friends. Early Education and Development, 16, 185−206. Susser, S. A., & Keating, C. F. (1990). Adult sex role orientation and perceptions of aggressive interactions between girls and boys. Sex Roles, 23, 147−155. Tomada, G., & Schneider, B. H. (1997). Relational aggression, gender, and peer acceptance: Invariance across culture, stability over time, and concordance among informants. Developmental Psychology, 33, 601−609. Wasik, B. H. (1987). Sociometric measures and peer descriptions of kindergarten children: A study of reliability and validity. Journal of Clinical Child Psychology, 16, 215−224. Werner, N. E., & Crick, N. R. (1999). Relational aggression and social–psychological adjustment in a college sample. Journal of Abnormal Psychology, 108, 615−623. Werner, N. E., & Crick, N. R. (2004). Maladaptive peer relationships and the development of relational and physical aggression during middle childhood. Social Development, 13, 495−514. Wu, X., Hart, C. H., Draper, T. W., & Olsen, J. A. (2001). Peer and teacher sociometrics for preschool children: Cross-informant concordance, temporal stability, and reliability. Merrill-Palmer Quarterly, 47, 416−443. Yoshikawa, H. (1994). Prevention as cumulative protection: Effects of early family support and education on chronic delinquency and its risks. Psychological Bulletin, 115, 28−54. Zahn-Waxler, C. (1993). Warriors and worriers: Gender and psychopathology. Development and Psychopathology, 5, 79−89. Zahn-Waxler, C., Ianotti, R. J., Cummings, E. M., & Denham, S. (1990). Antecedents of problem behaviors in children of depressed mothers. Development and Psychopathology, 2, 271−291. Zalecki, C. A., & Hinshaw, S. P. (2004). Overt and relational aggression in girls with attention deficit hyperactivity disorder. Journal of Clinical Child and Adolescent Psychology, 33, 125−137. Zimmer-Gembeck, M., Geiger, T., & Crick, N. R. (2005). Relational aggression, physical aggression, prosocial behavior, and peer relations: Gender moderation and bidirectional associations. Journal of Early Adolescence, 25, 421−452. Journal of Applied Developmental Psychology 44 (2016) 52–62 Contents lists available at ScienceDirect Journal of Applied Developmental Psychology Associations of coolness and social goals with aggression and engagement during adolescence Sarah M. Kiefer ⁎, Joy Huanhuan Wang University of South Florida, United States a r t i c l e i n f o Article history: Received 30 October 2014 Received in revised form 23 February 2016 Accepted 24 February 2016 Available online 24 March 2016 Keywords: Coolness Social goals Aggression Academic engagement Middle school a b s t r a c t This study examined associations of coolness and social goals with aggression and academic engagement, and whether social goals and gender moderated associations across the fall and spring of sixth grade (first year of middle school). Students (N = 347; 49% females) self-reported social goals (popularity, dominance, intimacy) and engagement (involved, disruptive behavior) and peer-reported coolness and aggression (overt, relational). Results indicated relations of coolness and social goals with subsequent aggression and engagement, and goals and gender moderated associations. Cool youth who endorsed intimacy goals had higher overt aggression; cool boys with low popularity goals or high dominance goals had higher overt aggression. Cool youth endorsing dominance goals and cool girls endorsing popularity goals had higher relational aggression. Cool youth and boys endorsing dominance goals reported lower involved behavior. Youth endorsing popularity goals and cool youth endorsing dominance goals reported higher disruptive behavior. Implications for examining adolescent coolness and social motivation are discussed. © 2016 Elsevier Inc. All rights reserved. 1. Introduction Coolness is a salient feature of the peer context and has been found to have significant implications for social and academic functioning during early adolescence, especially during the first year of middle school (Bellmore, Villarreal, & Ho, 2011; Graham & Juvonen, 2002; Hoff, Reese-Weber, Schneider, & Stagg, 2009; Jamison, Wilson, & Ryan, 2015). Students are often concerned with fitting in with perceived peer group norms (Galván, Spatzier, & Juvonen, 2011) and being cool (Bellmore et al., 2011) as they navigate a larger, more complex peer system (Rubin, Bukowski, & Laursen, 2009). Due to changing social norms (e.g., becoming more favorable towards aggression and less favorable towards academic engagement), high peer status is increasingly related to aggressive and disengaged behavior in middle school (Bellmore et al., 2011; Bowker, Rubin, Buskirk-Cohen, Rose-Krasnor, & Booth-LaForce, 2010; Cillessen & Mayeux, 2007). Students also strive for a variety of social goals among peers, including popularity, dominance, and intimacy, which have significant implications for social and academic behaviors (Anderman, 1999; Kiefer & Ryan, 2008; Patrick, Anderman, & Ryan, 2002). Although related, coolness and social goals are conceptually distinct as coolness is an index of social status that reflects a consensus in the peer system (Jamison et al., 2015; Rodkin, Farmer, Pearl, & Van Acker, 2006), whereas social content goals reflect cognitive ⁎ Corresponding author at: University of South Florida, 4202 East Fowler Avenue, EDU 105, Tampa, FL 33620-5650, United States. E-mail address: kiefer@usf.edu (S.M. Kiefer). http://dx.doi.org/10.1016/j.appdev.2016.02.007 0193-3973/© 2016 Elsevier Inc. All rights reserved. representations of what individuals are trying to achieve in peer interaction (Wentzel, 2000). The present study makes a unique contribution to the field, studying the intersection of social and academic factors in school by examining the associations of coolness and social goals with subsequent aggression and engagement during the first year of middle school. Although cool peer reputation and social goals have important implications for students' functioning in school (Kiefer & Ryan, 2008; Rodkin, Ryan, Jamison, & Wilson, 2013), research examining relations among constructs across the school year is scarce (see Dawes & Xie, 2014; Ojanen & Findley-Van Nostrand, 2014 as exceptions). Studies have focused on aggression as an outcome, yet additional research is needed to distinguish between separate forms of aggression (overt and relational; Cillessen, Mayeux, Ha, de Bruyn, & LaFontana, 2014). Additionally, academic engagement has been understudied in relation to high social status and social goals. It is important to examine factors that shape engagement, as it is a precursor to achievement and amenable to change (Fredricks, Blumenfeld, & Paris, 2004). Examining the relations of status and goals with aggression and engagement may provide a comprehensive understanding of the behavior characteristic of high status youth motivated by certain goals. Further, dominance goals are relatively under-examined (Kiefer & Ryan, 2008) and most studies have not simultaneously investigated popularity, dominance, and intimacy goals in relation to adjustment. Although recent research has examined social goals (i.e., popularity goals) as moderating relations between high social status and aggression (Cillessen et al., 2014; Dawes & Xie, 2014; Ojanen & Findley-Van Nostrand, 2014), there is much we do not know. S.M. Kiefer, J.H. Wang / Journal of Applied Developmental Psychology 44 (2016) 52–62 The current study had two main aims: to examine the associations of coolness and social goals with later aggression (overt and relational) and academic engagement (involved and disruptive behavior), and to examine social goals and gender as moderators of these associations across the sixth grade (i.e., first year of middle school). Research indicates relations of high social status with aggression (Cillessen, Schwartz, & Mayeux, 2011; Dawes & Xie, 2014; Xie, Li, Boucher, Hutchins, & Cairns, 2006) and engagement (Schwartz, Gorman, Nakamoto, & McKay, 2006). However, there is not a consensus regarding the role of gender in moderating these relations. There is evidence for gender differences in relation to coolness, social motivation, aggression, and engagement (Cillessen et al., 2011; Xie et al., 2006), but relations have not been found in other studies (Dawes & Xie, 2014). Additional research is needed to clarify gender effects on cool status and adjustment (Rose, Glick, & Smith, 2011). We examined meanlevel gender differences and changes in variables across the first year of middle school due to puberty, social development, and school transition-related factors (Eccles, 2004; Pellegrini & Long, 2002). 2. Cool peer reputation: associations with aggression and academic engagement Cool peer reputation, or coolness, refers to “the embodiment of some combination of attributes that wins approval or earns the attention of others,” (Jamison et al., 2015, p. 384). Coolness encompasses a range of individual characteristics, including behaviors and attitudes, which not only win youth approval among peers but also express rebellion towards parents or other authorities, and eventually place them on the cutting edge of peer norms (Pountain & Robins, 2000). For example, coolness may include fashionable appearance and dress styles, trendy possessions, and sophisticated interpersonal skills (Adler & Adler, 1998; Adler, Kless, & Adler, 1992). Coolness is a social construction that is in constant flux (e.g., joining a jazz club in 1950's and punk explosion in 1970's; Pountain & Robins, 2000) and continually negotiated as youth determine a consensus of what is valued in the peer system (Adler et al., 1992). Cool youth are also characterized as being precocious (Pountain & Robins, 2000) and pseudomature (Allen, Schad, Oudekerk, & Chango, 2014). This aligns with Moffitt's (1993) “maturity gap”, which indicates early adolescence is a time youth strive to appear mature and adopt features associated with adult status that represent increased autonomy and less childlike behaviors. Cool peer reputation is used as an indicator of high social status. Research indicates the need to investigate coolness as an index of social status in addition to others, such as popularity (Bellmore et al., 2011; Rodkin et al., 2006). Youth can best define what coolness is for themselves (Pountain & Robins, 2000; Rodkin et al., 2006). Thus, peer nominated coolness is a developmentally appropriate index of high social status that allows adolescents to fully capture the authentic meaning of coolness. Although cool youth are a heterogeneous group with some subgroups identified as displaying aggressive and antiacademic behaviors and others displaying prosocial and positive academic behaviors (de Bruyn & Cillessen, 2006; Rodkin, Farmer, Pearl, & Van Acker, 2000), they are often viewed as the most mean and manipulative of all peers, and are often not well-liked (Eder, 1985; Merten, 1997; Rodkin et al., 2006). Conceptually, coolness is similar to perceived popularity as they both assess peer status. However, coolness is distinct from perceived popularity. Coolness focuses on individuality and gaining peer approval or attention (Jamison et al., 2015; Pountain & Robins, 2000), whereas popularity reflects social centrality and prominence within the peer group (Cillessen et al., 2011; Rodkin et al., 2006). Although youth nominated as cool may be popular, social centrality by itself is not sufficient for coolness. We anticipate that coolness may relate to study variables in similar and different ways from popularity and have unique implications for adjustment. For example, coolness and popularity may have similar relations with aggression and engagement as they indicate high social status, yet relations may differ 53 as cool youth focus on individual attributes rather than social network centrality. We expected that coolness will be relatively stable across time, given research examining cool peer nominations during early adolescence (Bellmore et al., 2011; Jamison et al., 2015). Coolness has been found to be associated with overt and relational aggression during early adolescence (Bellmore et al., 2011; Hoff et al., 2009; Rodkin et al., 2013; Rose, Swenson, & Waller, 2004). Social status is increasingly linked to peer-reported overt and relational aggression (Graham & Juvonen, 2002; Rose et al., 2004) and both forms of aggression are used to gain and maintain status during early adolescence (Cillessen & Mayeux, 2004; Hoff et al., 2009; Juvonen, Wang, & Espinoza, 2013). Research has identified a subset of overtly aggressive, high status youth (Rodkin et al., 2000; Rodkin et al., 2006) and ethnographic research indicates cool youth use relational aggression to maintain peer exclusivity and influence peers and group norms (Adler & Adler, 1998; Xie, Cairns, & Cairns, 2002). Although we expected coolness to be related to both overt and relational aggression, we hypothesized that the link with relational aggression would be stronger than overt forms as overt aggression is less approved by peers during this developmental period (Björkqvist, Österman, & Kaukiainen, 1992; Hoff et al., 2009). Compared to the abundant research examining links between social status and aggression, less is known regarding academic engagement. We examined relations of cool peer reputation with aggression and engagement as few studies have examined social status in relation to both social and academic functioning (see Cillessen & Mayeux, 2007 as an exception). Coolness is associated with a negative academic reputation across the first year of middle school, although this varies across subgroups and is contingent upon prior disruptive behavior (Jamison et al., 2015). Cool status is increasingly associated with academic disengagement (Schwartz & Gorman, 2011) and viewed as being incompatible with effort and achievement during adolescence (Anderman, 1999; Galván et al., 2011; Juvonen & Murdock, 1995). Aligning with changing norms is critical in ‘staying cool’ during the first year of middle school and has implications for aggression (Bellmore et al., 2011), yet it is unknown what implications coolness may have for academic engagement. Would adolescents consider “downplaying effort” or “being disruptive” as individual attributes that embody coolness? We anticipated that cool peer reputation will be positively related to subsequent disruptive behavior and negatively related to later involved, on-task classroom behavior. 3. Social goals: associations with aggression and academic engagement Two complementary approaches to social goals include an achievement orientation approach and a goal content approach. Achievement goal theory within the social domain focuses on the development or demonstration of social competence, as well as relations between students' goal orientations, beliefs, and behaviors (Dweck & Leggett, 1988; Ryan & Shim, 2006, 2008). The achievement orientation approach focuses on students' different orientations towards social competence and transcends specific goals (e.g., reasons why an individual is trying to achieve; Ryan & Shim, 2008). In the current study, we asked students what they like to strive for when they are with peers; thus the focus was on outcomes that would make students feel happy or successful. Specifically, we assessed popularity, dominance and intimacy goals. This is aligned with a goal content approach (e.g., cognitive representations of what an individual is trying to achieve) and often serves to direct their behavior towards outcomes (Ford, 1992; Wentzel, 2000). This conceptualization and operationalization is similar to prior social goals research (Jarvinen & Nicholls, 1996) and has been found to have significant implications for early adolescents' adjustment in school (Anderman, 1999; Patrick et al., 2002). Compared to popularity and intimacy goals, dominance goals have been relatively understudied (Kiefer & Ryan, 2008) and few studies 54 S.M. Kiefer, J.H. Wang / Journal of Applied Developmental Psychology 44 (2016) 52–62 have examined all three goals simultaneously. Further, social goal research has focused more on links to social functioning (e.g., aggression) than academic functioning (e.g., engagement), and these constructs are rarely studied together. Thus, our understanding of social goals during early adolescence is incomplete. Social goals play a key role in motivating behavior (Crick & Dodge, 1994; Erdley & Asher, 1999) and have implications for social and academic functioning in school (Kiefer & Ryan, 2008; Ojanen & Findley-Van Nostrand, 2014). Stored in the long-term memory, traitlike goals are activated by contextual cues and affect social information processing and behavior (Crick & Dodge, 1994). For instance, dominance goals may be activated when encountering an uncertain peer context and goal-concordant behavioral strategies like aggression are used to establish dominance in early adolescence (Pellegrini & Long, 2002). Social goals allow individuals to focus their attention on goalrelevant information, activating behavioral strategies associated with the goal, and influence adjustment (Dawes & Xie, 2014; Dweck & Leggett, 1988). Although moderately correlated, social goals (popularity, dominance, and intimacy goals) are distinct and differently related to adjustment (Kiefer & Ryan, 2008). Popularity goals refer to a focus on striving for high social status characterized by visibility and prestige within the larger peer group (Jarvinen & Nicholls, 1996) and are distinct from peer status (Cillessen et al., 2014). Popularity goals are positively related to teacher- and peer-reported overt and relational aggression during early adolescence (Dawes & Xie, 2014; Rodkin et al., 2013; Ryan & Shim, 2008). Although examined less frequently, popularity goals are negatively related to academic values, effort, and achievement (Anderman, 1999; Cillessen & Mayeux, 2007; Kiefer & Ryan, 2008). We expected that popularity goals will be positively related to subsequent overt and relational aggression as well as disruptive behavior, and negatively related to involved behavior. Popularity goals were also expected to increase or remain stable across time as the priority of popular status peaks during early adolescence (LaFontana & Cillessen, 2010). Dominance goals refer to a focus on having power over peers, characterized by getting peers to comply with one's wishes and instilling fear in others (Jarvinen & Nicholls, 1996). Dominance goals are related to peer- and teacher-reported overt and relational aggression (Ojanen, Grönroos, & Salmivalli, 2005; Ojanen, Findley, & Fuller, 2012). Although most research has examined implications of dominance goals for social functioning, less is known regarding academic functioning. Dominance goals are associated with off-task behavior and decreased achievement (Kiefer & Ryan, 2008), as these goals often conflict with teacher-focused goals. Examining implications of dominance goals for social and academic functioning can help us understand dominance goals more comprehensively. We expected dominance goals to have similar yet stronger relations as popularity goals: dominance goals will be positively related to subsequent aggression (overt and relational) and disruptive behavior, and negatively related to later involved behavior. Dominance goals were expected to increase across the first year of middle school when the peer hierarchy is in flux and students navigate a new social ecology (Pellegrini & Long, 2002). Intimacy goals refer to a focus on establishing peer relationships characterized by mutual support and disclosure of thoughts and feelings (Jarvinen & Nicholls, 1996) and reflect a salient concern in adolescent peer relationships (Cillessen & Mayeux, 2004). Closeness goals (similar to intimacy goals) are negatively related to overt aggression, but unrelated to relational aggression (Ojanen et al,. 2012). Although overt aggression (especially physical) undermines intimacy with peers, relational aggression (e.g., spreading rumors) requires some closeness with peers (Ojanen et al., 2012). Intimacy goals are related with positive academic adjustment, including prosocial behavior, engagement, and achievement (Kiefer & Ryan, 2008; Ojanen et al., 2005). We expected intimacy goals to be negatively related to overt aggression and to have little to no relation with relational aggression. Intimacy goals were also expected to be positively related to subsequent involved behavior and negatively related to later disruptive behavior. Intimacy goals were expected to decrease across time, as peer norms increasingly reflect negative social and academic behaviors during early adolescence (Cillessen & Mayeux, 2007; Galván et al., 2011; Kiefer & Ryan, 2011). 4. Social goals and gender as moderators Cool youth often have opportunities to enact social goals that are not available to individuals with lower status (Dawes & Xie, 2014; Rubin et al., 2009). Cool youth may also encounter social pressures to “fit in” with peer norms (Cillessen et al., 2011) and experience tension as they strive for individuality and peer attention or acceptance (Jamison et al., 2015; Pountain & Robins, 2000). Social goals may moderate relations of coolness with later aggression and engagement, with youth striving to engage in goal-congruent behavior. We anticipated that there will be enhanced negative relations of coolness with involved behavior as well as enhanced positive relations of coolness with aggression (overt and aggression) and disruptive behavior for youth who endorse popularity and dominance goals. These youth may develop socially dominant or manipulative behaviors (Cillessen et al., 2011) or negative attitudes towards school (Galván et al., 2011; Schwartz & Gorman, 2011), as popularity and dominance goals are congruent with strategies and outcomes associated with high status. Research indicates that popularity goals moderate relations among high social status and aggression during early adolescence. High status youth who strive for popularity report high levels of concurrent peer-reported aggression, although forms of aggression (overt and relational) were not examined (Cillessen et al., 2014). Popularity goals are related with concurrent relational aggression among high status youth (Dawes & Xie, 2014), whereas agentic and communal social goals moderate longitudinal links between popularity and aggression (Ojanen & Findley-Van Nostrand, 2014). The current study contributes to this research by examining popularity and dominance goals as moderating relations of coolness with aggression (overt and relational) and engagement (involved and disruptive behaviors) across the sixth grade. We expected that intimacy goals will diminish relations of coolness with aggression as these goals are incongruent with strategies and outcomes associated with cool status. Although coolness is related to aggression (overt and relational), we anticipated that there will be diminished relations for cool youth who endorse intimacy goals. While coolness is increasingly viewed as incompatible with engagement during adolescence (Galván et al., 2011; Jamison et al., 2015), we anticipated that there will be diminished relations of coolness with disruptive behavior and enhanced relations with involved behavior for youth who endorse intimacy goals. We also examined gender as moderating associations of coolness with aggression and engagement. Social cognitive theory suggests many attributes and roles promoted for boys and girls are influenced by a complex mix of experiences and guide gender-linked behavior (Rose & Rudolph, 2006; Ruble, Martin, & Berenbaum, 2006). Coolness differs among boys and girls, with physical aspects more salient among boys and social-relational aspects more salient among girls (Adler et al., 1992; Closson, 2009). Gender may moderate relations among constructs (Rose & Rudolph, 2006; Ruble et al., 2006), with stronger relations of coolness and goals with overt aggression for boys (Rodkin et al., 2000) and stronger relations of coolness and goals with relational aggression for girls (Cillessen & Mayeux, 2004; Eder, 1985; Merten, 1997). Despite recent advances, there are inconsistent findings regarding gender effects on high social status and aggression (Cillessen & Mayeux, 2004; Prinstein & Cillessen, 2003; Rose et al., 2004; Xie et al., 2006). Less research has examined gender differences in relations between cool status and engagement, and relevant longitudinal research is limited (Schwartz et al., 2006). Given a lack of consensus and need for additional research (Rose et al., 2011), no specific hypotheses were made and gender as a moderator was examined on an exploratory basis. S.M. Kiefer, J.H. Wang / Journal of Applied Developmental Psychology 44 (2016) 52–62 5. The present study This study addressed two main aims: (a) to examine the associations of coolness and social goals with aggression and academic engagement, and (b) to examine social goals and gender as moderating associations of coolness with aggression and engagement across the sixth grade. The present study advances prior research (Cillessen et al., 2014; Dawes & Xie, 2014; Ojanen & Findley-Van Nostrand, 2014) by examining three social goals (popularity, dominance, and intimacy) and gender as moderators of the associations of coolness with aggression and engagement. Examining these aims within the context of the first year of middle school may have implications for understanding early adolescent adjustment. An additional aim was to examine mean-level gender differences and changes in study variables across time. We expected boys to strive for dominance and popularity goals more and strive for intimacy goals less than girls (Jarvinen & Nicholls, 1996; Kiefer & Ryan, 2008). Consistent with peer nomination research, we expected boys to have higher levels of overt aggression and girls to have higher levels of relational aggression (Cillessen & Mayeux, 2004, 2007). We expected boys to report lower levels of involved behavior and higher levels of disruptive behavior compared to girls (Adler et al., 1992; Cillessen & Mayeux, 2007). Regarding change over time, we anticipated coolness to be relatively stable (Bellmore et al., 2011; Jamison et al., 2015). We expected popularity goals to increase or remain stable, dominance goals to increase, and intimacy goals to decrease across time (Kiefer & Ryan, 2011; LaFontana & Cillessen, 2010; Pellegrini & Long, 2002). We also expected involved behavior to decrease, and disruptive behavior and aggression to increase across time (Anderman, 1999; Cillessen & Mayeux, 2004, 2007; Galván et al., 2011; Pellegrini & Long, 2002). 6. Method 6.1. Pr...
Purchase answer to see full attachment