Vany 8hs波温 Influences of Gender Identity on Children's Maltreatment of Gender-Nonconforming Peers:A Person X Target Analysis of Aggression Rachel E. edwhether gen to t es abuse of gend eh is 7andanaltoconl adults who exhihit salien ol year nder-atypica individu simple disp nd even physical aggr ssion from (some of)their peers situ 9 ing gender. ical qualities (Bemdt Heller 1986:Kite n ieg likely to hand 1996: 1994: M Hymel.&R ubin. to be defended by pers)but rather b for cross-gender behay re at heiel arouse d ing dep sub ger,ar 2003: Lam Mishna.Ne tre 010 Walls Kane Wisneski 2010)Understanding the oots of wea con aggression toward gender-atypical persons is an important chal- the ognitive structures predict harsher treatment of gende armful attitudes and actions to with con t and I by has two cognitive pat ated wit tives underlyin negative treatment of in cts an der identity rooted in st ende nd att Tha trea d fema and who elves as hype culine(Herek.:Par ponden 1997S Pleck 1985:white 2001).Also.women who endorse the traditional female role some 43
Influences of Gender Identity on Children’s Maltreatment of Gender-Nonconforming Peers: A Person Target Analysis of Aggression Rachel E. Pauletti, Patrick J. Cooper, and David G. Perry Florida Atlantic University We investigated whether gender identity influences preadolescents’ tendency to single out genderatypical peers for abuse. Data were gathered from 195 boys and girls (M age 10.1 years) in the fall and spring of a school year. Children self-reported multiple dimensions of gender identity (intergroup bias, felt pressure for gender differentiation, felt gender typicality, gender contentedness); peers assessed each other’s social behavior (gender nonconformity, aggression toward each classmate). Using multilevel modeling, we examined how children’s attacks on gender-nonconforming peers (relative to their attacks on other peers) changed over the school year depending on their gender identity. There was modest support for the hypothesis that overconfident, arrogant gender identity promotes abuse of gender-atypical peers but considerable support for the hypothesis that insecure, self-questioning gender identity fosters this tendency. Implications for issues central to contemporary personality theory (e.g., Person Situation interaction) are discussed. New and somewhat surprising information about the cognitive and behavioral characteristics of gender-nonconforming preadolescents is provided. Keywords: aggression, gender identity, gender nonconformity, Person Situation interaction, victimization Children, adolescents, and adults who exhibit salient crossgender characteristics are at heightened risk for rejection, harassment, discrimination, and abuse. As early as the preschool years, children who behave in gender-atypical ways elicit dislike, teasing, and even physical aggression from (some of) their peers (Fagot, 1977; Langlois & Downs, 1980). Negative judgments and harsh treatment of gender-atypical peers continue across middle childhood and into adolescence; boys are more likely than girls both to treat their gender-atypical peers badly and to be harassed for having gender-atypical qualities (Berndt & Heller, 1986; Kite & Whitley, 1996; Lobel, 1994; Moller, Hymel, & Rubin, 1992; Poteat, Espelage, & Green, 2007). Moreover, youths who are victimized for cross-gender behavior are at heightened risk for additional negative outcomes, including depression, substance abuse, truancy, social isolation, hopelessness, and violence (Kimmel & Mahler, 2003; Lampinen et al., 2008; Mishna, Newman, Daley, & Solomon, 2009; Toomey, Ryan, Diaz, Card, & Russell, 2010; Walls, Kane, & Wisneski, 2010). Understanding the roots of aggression toward gender-atypical persons is an important challenge. Harmful attitudes and actions toward gender-atypical persons are perpetrated by a minority of individuals, raising questions about the unique psychological factors that motivate them. Our knowledge of the motives underlying negative treatment of individuals with cross-gender characteristics comes mainly from studies of adolescents and adults who dislike, harass, or assault persons whom they presume to be gay or lesbian. An implicit (but untested) assumption of this research is that negative treatment of gender-atypical individuals derives from more than simple dispositional or trait aggression (i.e., the tendency to be more aggressive than other people across diverse situations, including different target types) but rather is motivated, at least in part, by cognitive factors (e.g., values, beliefs, expectancies, goals, identities) specific to gender. In other words, it is believed that perpetrators do not treat gender-atypical persons harshly simply because such targets constitute a convenient victim group (e.g., likely to hand over resources, unlikely to be defended by peers) but rather because the perpetrators’ gendered cognitions are stimulated by the targets’ gender-atypical behaviors in ways that arouse disgust, threat, anger, and aggression specifically toward these targets. This literature has identified several cognitive predictors of harsh treatment of gender-atypical others, though many studies are characterized by methodological weaknesses (e.g., concurrent research designs, exclusive use of self-reports, failure to test whether the cognitive structures predict harsher treatment of genderatypical persons than of other target types, lack of control for trait aggression). This work, which is mainly with adolescent and adult males, has highlighted two cognitive patterns associated with aggression toward gender-nonconforming others. The first pattern reflects an overconfident gender identity rooted in strong gender stereotypes and attitudes. That is, much harsh treatment of genderatypical males is perpetrated by males who polarize the sexes (believe that males and females should act very differently), who espouse hypermasculinity (believe that males should inhibit tender emotions, take dangerous risks, dominate others, etc.), and who appraise themselves as hypermasculine (Herek, 1988, 2000; Parrott, 2009; Parrott, Adams, & Zeichner, 2002; Pryor & Whalen, 1997; Stark, 1991; Thompson, Grisanti, & Pleck, 1985; Whitley, 2001). Also, women who endorse the traditional female role someRachel E. Pauletti, Patrick J. Cooper, and David G. Perry, Department of Psychology, Florida Atlantic University. Correspondence concerning this article should be addressed to David G. Perry, Department of Psychology, Florida Atlantic University, Boca Raton, FL, 33431. E-mail: perrydg@fau.edu This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Journal of Personality and Social Psychology © 2014 American Psychological Association 2014, Vol. 106, No. 5, 843– 866 0022-3514/14/$12.00 DOI: 10.1037/a0036037 843
844 PAULETTL COOPER.AND PERRY diffe entiating beliefs and identities h nder-atv cal other night.individually and in combination with other yariables.affec ats to their se gender-atyp der favorably th the other) der-atypical ecting other people to recipr sec atment of nd gen with one's gender anxiety over one's adegu as a member of one be gender fit s of ende wel to the esire t th Tha he o th Felt d Eisler,Moore 2001;Kimme 1994:Parro 2001 is par ce adisturbin eater use of and b a gap betwec what one feels By guilt self-rejection ange 助d iated with elational difficultie 198( cr. (for oys:P.Ro 20:P. 1992).Males with a strongly gen 200 eithe lack self-effi d male c-typica two dimensions of gender identit gcnder to inhibit feminine behavi nder identity because they re children' ve own gen ption of salien and valued attribute 8.De and ng that the gender inadequacy reside nsions of gender ide ntity are usuall mentof both the o fiden ure cogni nent in gender-diff ng by et al Egan Perry,2001;Smith Leaper,2006 emulate the the result is an people hold the ail to one anoth (e.g..Egan persons to har ender-nonco nstruct (e.g.,s hr oh a pern der lens:Bem The Present Study he betw -gender and the within-gender forms of gender identity pt.desc ribed be ment of ger nder-atypical peers arch to be associated with maltre tment of aibltidim ollec e (As eers. There grounds for expecting th patters to apply to preadolescent
times dislike and are angered by gender-atypical females (Parrott & Gallagher, 2008). Presumably, people with strong genderdifferentiating beliefs and identities harass gender-atypical others because they perceive them as threats to their worldviews, as morally disgusting, as challenges to the positive distinctiveness of their gender collective, as disloyal and untrustworthy members of the collective, or as expecting other people to reciprocate their gender-atypical behavior. A second cognitive pattern associated with maltreatment of gender-atypical persons reflects insecure gender identity— uncertainty and anxiety over one’s adequacy as a member of one’s gender collective. Insecurity over one’s gender fit is especially conducive to attacks on gender-atypical others for individuals who also possess strong gender-differentiating stereotypes and attitudes (in this regard, they resemble persons with overconfident gender identity). That is, males who subscribe to rigid standards of hypermasculinity but perceive themselves as failing to live up to the standards tend to treat gender-atypical others harshly (Bernat, Calhoun, Adams, & Zeichner, 2001; Eisler & Skidmore, 1987; Franchina, Eisler, & Moore, 2001; Kimmel, 1994; Parrott, 2009; Vandello, Bosson, Cohen, Burnaford, &Weaver, 2008). Such males presumably experience a disturbing ought self-discrepancy (a gap between what one feels one should be and what one really is). Ought self-discrepancies are conducive to frustration, agitated distress, guilt, shame, self-rejection, anger, and loss of selfregulatory control (C. S. Carver, Lawrence, & Scheier, 1999; Higgins, 1987; Swann & Bosson, 2008; Tangney, Wagner, Fletcher, & Gramzow, 1992). Males with a strongly gendered ought self-discrepancy are said to experience masculine gender role stress (Levant, 2011). Although they strive to be male typical, they either lack self-efficacy for self-prescribed male-typical behaviors (e.g., dominance, risk taking) or fear that in certain situations they will be unable to inhibit feminine behavior. Presumably, they perceive in gender-atypical others what they fear in themselves, and this perception activates latent feelings of insecurity (fears of gender inadequacy, of being perceived as gay, of rejection by same-sex others) that motivate them to lash out defensively at gender-atypical persons, thereby removing the threat and demonstrating that the gender inadequacy resides in other people and not themselves. Note that a common element of both the overconfident and insecure cognitive patterns is investment in gender-differentiating stereotypes and attitudes. When people adopt such standards and successfully emulate them, the result is an overconfident gender identity. When people hold the standards but fail to fulfill them, the result is an insecure gender identity. Perhaps either pattern disposes persons to harass gender-nonconforming others. The Present Study To our knowledge, no study has investigated the factors that lead preadolescent children to single out gender-nonconforming peers for abuse. The principal purpose of the present study was to investigate the role of gender identity in preadolescents’ maltreatment of gender-atypical peers. Gender identity is a multidimensional construct encompassing a person’s felt compatibility with, quality of motivation to fit in with, and evaluation of a gender collective (Ashmore, Deaux, & McLaughlin-Volpe, 2004; Egan & Perry, 2001; Tobin et al., 2010). Using a model of gender identity proposed by Egan and Perry (2001), we examined how four dimensions of gender identity might, individually and in combination with other variables, affect children’s tendencies to abuse gender-atypical peers. These dimensions of gender identity were intergroup bias (perceiving one’s own gender more favorably than the other), felt pressure for gender differentiation (internalized pressure to avoid other-sex behaviors), felt gender typicality (perceived similarity to same-sex others), and gender contentedness (satisfaction with one’s gender). The first two dimensions—intergroup bias and felt pressure for gender differentiation—may be considered between-gender facets of gender identity in that they depend on perceiving, and likely exaggerating, differences between the sexes. Strong forms of these facets of gender identity may well contribute to the desire to interact with, fit in with, and emulate others of one’s gender, but they probably also encourage avoidance and disparagement of other-gender activities and persons. Felt pressure for gender differentiation is in fact assessed as the expectation of punishment (e.g., ridicule, alienation) for cross-gender behavior (Egan & Perry, 2001). Intergroup bias is part of a set of intergroup cognitions that also includes exaggeration of differences between the sexes, greater use of stereotypes, and homogenization of the other sex (Powlishta, 1995; Tajfel & Turner, 1979). By preadolescence, these gender-polarizing forms of identity are relatively immature and are associated with peer-relational difficulties, low selfesteem, internalizing problems (for girls), and externalizing problems (for boys; P. R. Carver, Egan, & Perry, 2004; P. R. Carver, Yunger, & Perry, 2003; Corby, Hodges, & Perry, 2007; Egan & Perry, 2001; Smith & Leaper, 2006; Yunger, Carver, & Perry, 2004). The other two dimensions of gender identity—felt gender typicality and gender contentedness—may be considered withingender facets of gender identity because they capture children’s feelings of compatibility with, and acceptance by, their own gender collective. They are largely a function of children’s selfperception of salient and valued same-gender attributes (e.g., same-sex friendships) and are not necessarily undermined by selfperception of a few other-gender attributes (Egan & Perry, 2001; Spence, 1993). These dimensions of gender identity are usually positively correlated with self-esteem, social skills (e.g., prosocial behavior), and peer acceptance (P. R. Carver et al., 2003, 2004; Corby et al., 2007; Egan & Perry, 2001; Smith & Leaper, 2006; Yunger et al., 2004). Because the four dimensions of gender identity typically show low correlations with one another (e.g., Egan & Perry, 2001; Yunger et al., 2004), they cannot be considered interchangeable reflections of a single, common underlying construct (e.g., gender schematicity, or a generalized tendency to view self and others through a pernicious gender lens; Bem, 1981). Moreover, as noted, the between-gender and the within-gender forms of gender identity usually relate to adjustment in opposite ways. Using measures of the four gender identity dimensions (and of other potentially relevant aspects of self-concept, described below), we examined whether the cognitive patterns found in previous research to be associated with maltreatment of gendernonconforming people by older persons (overconfident and insecure gender identity) also motivate preadolescents to attack gender-nonconforming peers. There are grounds for expecting both patterns to apply to preadolescents. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 844 PAULETTI, COOPER, AND PERRY
AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 845 These self-appraisals do not necessarily reflect felt inc patibilit ad ender-atypicalpeerso ective.but at t tend to juc veness).Ov identity is perhap nder others (Ha 2006 hias or felt er diffe Mo ith one Tgender (st e strong ocu ch f orm of p ing ridicule) and they f within-gen identity to spu betwen-endo of gender identiy and a ugender-nonspeci peer group but p Frick. ng sh0 engths,goals, den )it ssion.forboth adultsand hildrer r identity.low felt gender typicality.gende 996:Mo 200 self-effica lthof.2008).Thus.high m might en Howeve like onfident er identity.n ent)probab 's ability to satisfy e and att with s the exacerbated they Thus em or l self-effic y fo children with strone r-nolar tity to s their gender adequacy by assling gender gende and in well he alt identity may not equir preadolescents sment nonconforming withm-g ight also som s take on s ades of tity.We allowed for this possibility in our analytic s dentity is nother cognitiv riable (co teracts with self m or a measure of self-efficacy to promo hildren with low intergroup s).malt cal peer ren with hi mSuc由a ny of th pattem would constitute evidence for both pathways. their inner conflict atte ing to show that the deficienc elat identity (int bias or felt diffe Our ach to h ent of g stitute clear-cut gend elf-dis self-dis incapable of displaying dominance.popularity.or good looks
Children with overconfident gender identity might aggress toward gender-atypical peers for many of the same reasons that older persons do (e.g., moral outrage, distrust, threat to ingroup cohesiveness). Overconfident gender identity is perhaps most clearly represented when a strong form of between-gender identity (strong intergroup bias or felt pressure for gender differentiation) combines with strong felt compatibility with one’s gender (strong felt gender typicality or gender contentedness). Thus, we examined whether each form of between-gender identity interacts synergistically with each form of within-gender identity to spur children to attack gender-nonconforming peers. However, we also examined whether each between-gender form of gender identity combines with each of four gender-nonspecific self-appraisals to stimulate attacks on gender-nonconforming peers. These other selfappraisals were global self-esteem and perceptions of self-efficacy for dominance, popularity, and physical attractiveness. High selfesteem motivates people to act confidently on their self-perceived strengths, goals, identities, and values (Baumeister, 1998). When high self-esteem is paired with self-serving ideation (e.g., narcissism), it sometimes fosters aggression, for both adults and children (Baumeister, Smart, & Boden, 1996; Morf & Rhodewalt, 2001; Thomaes & Bushman, 2011; Thomaes, Bushman, Stegge, & Olthof, 2008). Thus, high self-esteem might encourage children with strong ingroup favoritism or felt pressure for gender differentiation to attack gender-nonconforming others. High selfefficacy for dominance, popularity, or attractiveness might as well. High self-efficacy in these domains reflects confidence in one’s social status and in one’s ability to satisfy demonstration goals, or desires to be admired by peers; children with such goals tend to display adjustment difficulties, including aggression (O’Moore & Kirkham, 2001; Rodkin, Ryan, Jamison, & Wilson, 2013; Rudolph, Abaied, Flynn, Sugimura, & Agoston, 2011; Salmivalli, 1998). It is conceivable that high self-efficacy in these domains empowers children with strong gender-polarizing identity to strive for admiration and status by discriminating against gendernonconforming peers. Overconfident gender identity may not always require strong between-gender identity, however. Strong within-gender identity— high felt gender typicality or gender contentedness—might also sometimes take on shades of overconfidence or narcissistic arrogance and promote attacks on gender-nonconforming peers. Perhaps this occurs when strong within-gender identity is accompanied by high selfesteem or self-efficacy for dominance, popularity, or attractiveness. Thus, we also examined whether each form of within-gender identity interacts with self-esteem or a measure of self-efficacy to promote harassment of gender-atypical peers. Children with insecure gender identity might also discriminate against gender-nonconforming others for many of the same reasons that older persons do (e.g., defensive lashing out at those who arouse their inner conflict, attempting to show that the deficiency lies in others and not themselves). Insecure gender identity is perhaps most clearly represented when a strong form of betweengender identity (intergroup bias or felt pressure for gender differentiation) combines with strong felt incompatibility with one’s gender (low felt gender typicality or low gender contentedness). Such pairings constitute clear-cut gender self-discrepancies. However, children with strong between-gender identity might also feel self-discrepant and anxious if they have low self-esteem or feel incapable of displaying dominance, popularity, or good looks. These self-appraisals do not necessarily reflect felt incompatibility with one’s gender collective, but at this age, children tend to judge their self-worth and competencies by comparing themselves to same-gender others (Harter, 2006), and self-appraisals on these dimensions are therefore likely to reflect how well children feel they are stacking up against same-sex others. Moreover, children with strong between-gender identity might focus on satisfying demonstration goals (e.g., feeling superior, being admired, avoiding ridicule), and if they view themselves as valueless or as incapable of fulfilling these goals, their frustration and insecurity may be acute. Although research on self-discrepancies in children is scant, children who ardently strive for status and admiration in the peer group but perceive themselves as falling short do tend to be aggressive (Barry, Frick, & Killian, 2003; Harter & McCarley, 2004; Pauletti, Menon, Menon, Tobin, & Perry, 2012); it is plausible that between-gender identity channels the aggression of these frustrated children toward gender-atypical targets. Thus, the insecure pathway to aggression toward gender-nonconforming peers would be supported if for children with a strong form of betweengender identity, lower felt gender typicality, gender contentedness, self-esteem, or self-efficacy for dominance, popularity, or attractiveness were to motivate attacks on gender-nonconforming peers. However, like overconfident gender identity, insecure gender identity may not always require strong between-gender identity. Children who feel incompatible with their gender (who feel gender atypical or discontent) probably experience some uncertainty over their adequacy as members of their gender collective even if they lack strong gender-differentiating stereotypes and attitudes; moreover, their anxiety may be exacerbated if they have low selfesteem or feel incapable of displaying attributes needed for status in their peer group. Thus, low self-esteem or low self-efficacy for dominance, popularity, or attractiveness might potentiate a tendency on the part of children who feel gender incompatible to demonstrate their gender adequacy by hassling gendernonconforming peers. The overconfident and insecure pathways may well be alternate routes to preadolescents’ harassment of gender-nonconforming peers: Some children might harass gender-atypical peers because of overconfident gender identity, others because of insecure gender identity. We allowed for this possibility in our analytic strategy by probing for a quadratic effect of one cognitive variable (e.g., self-esteem) at different levels of another cognitive variable (conceptualized as the moderator; e.g., intergroup bias). For example, it might be that for children with high intergroup bias (but not for children with low intergroup bias), maltreatment of genderatypical peers is greater for children with either high or low self-esteem than for children with intermediate self-esteem. Such a pattern would constitute evidence for both pathways. The Present Study in Relation to Contemporary Personality Theory Our approach to conceptualizing and evaluating cognitive pathways to harassment of gender-atypical others is grounded in contemporary cognitive metatheories of personality (Cervone, 2004; Mischel & Shoda, 1995, 2008; Read et al., 2010; Zakriski, Wright, & Underwood, 2005). These theories offer conceptual strategies for generating hypotheses linking people’s stable cognitive processing structures—their values, expectancies, goals, beliefs, idenThis document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 845
846 PAULETTL COOPER.AND PERRY tities,and the like- use the th ries depart in critical wavs ch o examine cognitive pathw to children's har ssment o the that play con and of thei vior (e.g..agg ab his or her situations they ensitive to and to react re psych evant to them theories do n meas sive thar irections of influe aple it is cono able that children' who d partic it is clea hich the nd stable personal profiles indicatin nt gende-aypica pee ers,that is,to di ect more aggre children'sggre sion in the ions.To dete severity of children' dren aggression in sc on toward their victim e nay be child" (ehave simpositivethn-child bet).but one child hild'smv (ac ations is the child's aggress sign a children diffe )re istinc address the impact of social identity only on the former. fter the ative influences of situatio and trai erences acr Characteristics of Gender-Nonconforming Children on the cog here some effort identify er-atyp with bility in ag bou t the children whose victimiz the focus of stu ion partn (e.g..pee vs. or is pred othe tud than hy 200 Zakriski et al 2005).These data support Four ing ch nsk th o m)if the a problematic form of tity (high al with who one is inter ty)th for s dfcrcentiationorfeldisconmcicdiomthcrow a central ten of cogniti ender may be particularly distresse Method ing this issue by examinine whether children's coonitive struc Participants and Procedure to aggress more toward All chilren n the oys (M =10.2 ears)and 94 girls (Ma e=10.0ye gender atypical as a situationa
tities, and the like—to their social behavior. Because our research questions and paradigm were stimulated by these theories and because the theories depart in critical ways from other approaches to investigating human aggression, we describe certain features of the theories and tell how they guided the present study. Cognitive theories emphasize that many persons display considerable situational specificity in a social behavior (e.g., aggression) owing to interactions between their relatively enduring processing structures and features of the situations they encounter. That is, people’s unique cognitive structures cause them to be sensitive to, and to react in particular ways to, aspects of situations that are psychologically relevant to them. These theories do not deny that aggression also possesses the qualities of a personality trait (i.e., that some people are generally more aggressive than others), but they suggest that much can be learned about aggressive motivation by studying how individuals organize their aggression around particular eliciting circumstances. Although data are still accumulating, it is clear that many children do possess unique aggression signatures (Mischel & Shoda, 1995)— distinctive and stable personal profiles indicating how their aggression varies over different situations (relative to other children’s aggression in the same situations). To determine children’s signatures, children are observed for aggression in several situations. For each situation, each child’s aggression is converted to a z score indicating the child’s aggression relative to other children, thus removing the nomothetic influence of the situation. A child’s idiographic profile of z scores across the situations is the child’s aggression signature. The child’s average z score is a measure of trait aggression (Fournier, Moskowitz, & Zuroff, 2008; Mischel & Shoda, 1995). Many children differ markedly across situations in their aggression, and this variability cannot always be accounted for by trait aggression. In other words, after the normative influences of situations and trait aggression are removed, considerable intraindividual differences across situations (signatures) often remain to be explained. There has been some effort to identify the salient features of situations that organize within-child variability in children’s aggression. Chief among these situational features are the characteristics and behaviors of a child’s interaction partners (e.g., peer vs. adult, male peer vs. female peer, liked peer vs. disliked peer; Coie et al., 1999; Hodges, Peets, & Salmivalli, 2009; Matthys, Maassen, Cuperus, & Van Engeland, 2001; Mischel & Shoda, 1995; Peets, Hodges, & Salmivalli, 2011; Rodkin & Berger, 2008; Veenstra et al., 2007; Zakriski et al., 2005). These data support Fournier et al.’s (2008, p. 533) observation that “the most salient psychological features of the situation are found in the behavior of the individual with whom one is interacting.” Although there has been progress in identifying some features of situations (e.g., target characteristics) that underpin within-child situational variability in aggression, a central tenet of cognitive theories—the idea that individual differences in within-child situational variability in aggression are predictable from enduring cognitive processing structures— has, to our knowledge, yet to be investigated. The present study illustrates a paradigm for researching this issue, by examining whether children’s cognitive structures predict their tendency to aggress more toward genderatypical peers than toward other peers. Thus, we conceptualize the degree to which a child’s interaction partner is characteristically gender atypical as a situational variable that interacts with children’s processing structures to affect aggression. Our paradigm therefore is one of Person Target interaction. To examine cognitive pathways to children’s harassment of gender-atypical peers, we used multilevel modeling. In both the fall and spring of a school year, we collected measures of children’s cognitions, of their aggression toward each classmate, and of their gender-atypical behavior. For each child, a beta coefficient (slope) was calculated indicating how the child’s aggression toward his or her classmates changed over the school year as a function of the classmates’ gender-atypical behavior in the fall (within-subjects analyses). To test the hypotheses, these within-child slopes were predicted from measures of participants’ cognitions (between-subjects analyses). The longitudinal design was used to help rule out certain alternative directions of influence. For example, it is conceivable that children’s cognitions are reactive to their aggression toward gender-atypical others (e.g., boys who attack boys for playing with girls may afterward feel more gender typical or more dominant). It is important to note that our hypotheses address only the degree to which the cognitive variables affect children’s tendencies to discriminate against gender-atypical peers, that is, to direct more aggression toward gender-atypical peers than toward gender-typical peers. Our hypotheses do not address the absolute severity of children’s aggression toward their victims. Two children may be similarly inclined to victimize gender-atypical peers more than gender-typical ones (i.e., have similar positive within-child betas), but one child may aggress generally (across all peer targets) at a higher level than the other and therefore show a more generally elevated pattern (i.e., a higher intercept). Sensitivity to a target feature (captured by slope) and average aggression across targets (i.e., trait or general aggression, captured by intercept) are distinct constructs, and our hypotheses address the impact of social identity only on the former. Characteristics of Gender-Nonconforming Children Although the focus of this investigation was on the cognitive determinants of children’s aggression toward gender-atypical peers, the measures we collected afforded an opportunity to learn more about the children whose victimization was the focus of study— gender-nonconforming preadolescents. We describe the self-concepts and social-behavioral qualities of these children. We examine whether gender nonconformity predicts, or is predicted by, each other study variable over time. Although this aspect of our study was largely exploratory rather than hypothesis driven, we examined whether gender-nonconforming children are at greater risk for adjustment difficulties (e.g., peer victimization, internalizing problems, low selfesteem) if they possess a problematic form of gender identity (high between-gender identity, low within-gender identity) than if they do not. For example, gender-nonconforming children who feel strong pressure for gender differentiation or feel disconnected from their own gender may be particularly distressed. Method Participants and Procedure All children in the fourth through seventh grades of a university laboratory school were invited to participate. Participants were 101 boys (M age 10.2 years) and 94 girls (M age 10.0 years) who received written informed parental consent; they represented 75% This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 846 PAULETTI, COOPER, AND PERRY
AGGRESSION TOWARD GENDER-NONCONEORMING PEERS 847 oftheir grades.The numbers (and meana 43 ( cale ind The lts with this of the it al child preted cautiously aisals es of the school ensure th attractivenes wa asured by ha ving childre har e of the entire popul t each o hildren were tes individually ina library cubice in both the ce self-efficacy was asse hilnd spring the sch ght it feel like I am i stating that they were free to iny questio n the self-effica was a with si items ."Being popular hing suit is me. Being good looking is sment of Cognitive Variables 84.and87. identity uch they liked ates(Ho Assessment of Social Behavior nAchil's a e from sepa tion instrument per sex to indey ate were included on these instr gender nonconformity (a ified versions of Fo and Per es:th ixth and I cor not have 0n4 tha for me and e.g My fifth graders'classm nates and to sixth and seventh graders'grad chil nat girl )and the as administered.Item tems from p 5203 )bu the bous b For each item.children read descripti s of tw with bov "He lik of chac to do things that girls us lly do ones depic H alway inations):children cro list so th with higher so on of classmates (of either sex)who nominated the chilo the BUT Oth the ad very high loa s for all four gende for the The nents promin ce they were BUT all their life BUTOther girlsare glad they'beagr all their life and internalizing problems(e.gsadness.anxiety).A similar anal
of the children in their grades. The numbers (and mean ages in the fall) of children in Grades 4 through 7, respectively, were 52 (8.7 years), 56 (9.7 years), 43 (10.7 years), and 44 (11.7 years). Four additional children were tested in the fall but did not participate in the spring; participation in the spring was limited to children who had participated in the fall, and thus spring participants represented 98% of fall participants. The sample was approximately 51% White, 21 % Black, 20% Hispanic, and 8% other. The admissions procedures of the school ensure that the student body is representative of the entire population of Florida in terms of ethnicity/race and socioeconomic level. Children were tested individually in a library cubicle in both the fall and spring of the school year. At the start of each session, children signed an assent form describing the questionnaires and stating that they were free to leave blank any question they preferred not to answer. At the conclusion, children were asked not to discuss their responses or the study with other children. Assessment of Cognitive Variables Gender identity. Four dimensions of gender identity were assessed. Intergroup bias was calculated from children’s ratings of how much they liked each of their participating classmates (“How much do you like each kid?”). Ratings were on a 4-point scale with 1 indicating Not at all and 4 indicating A lot. A child’s rating of a peer was converted to a z score indicating the child’s liking of the peer relative to all children’s liking of that peer. The average other-sex z was subtracted from the average same-sex z to index intergroup bias, or ingroup favoritism. The other three facets of gender identity were assessed with modified versions of Egan and Perry’s (2001) scales; the scales were shortened to retain items with high item-total correlations in previous studies. Felt pressure for gender differentiation was assessed with seven items. Children rated statements on 4-point scales with 1 indicating Not at all true for me and 4, Very true for me. Items described possible reactions for gender-atypical behavior from parents (two items; e.g., “My parents would be upset if they saw me acting like a girl”), same-sex peers (two items; e.g., “The boys I know would be upset if I wanted to learn an activity that girls usually do”), and the self (three items; e.g., “I wouldn’t like myself if I heard myself talking or laughing like a girl”). Scale totals were averages of the seven items. Cronbach’s alpha was .81. Items on the felt gender typicality and gender contentedness scales were formatted alike. Each had five items, with about half reverse-scored. For each item, children read descriptions of two kinds of children— ones depicted as high on the gender identity dimension and ones depicted as low—and chose which kind of children they resembled more; they then indicated whether this choice was Very true or Sort of true for them. This yielded a 4-point scale, with higher scores indicating greater gender identity. Scale scores were item averages. Items assessing gender typicality included (from the girls’ form) “Some girls feel they are different from other girls BUT Other girls feel they are similar to other girls” and “Some girls have the same feelings that other girls have BUT Other girls don’t have the same feelings that other girls have.” Gender contentedness items included “Some girls are happy that they were born a girl BUT Other girls are not happy that they were born a girl” and “Some girls wish they didn’t have to be a girl all their life BUT Other girls are glad they’ll be a girl all their life.” Cronbach’s alpha was .68 for gender typicality and .59 for gender contentedness. The marginal reliability of the gender contentedness scale indicates that results with this measure should be interpreted cautiously. Gender-nonspecific self-appraisals. Self-esteem was assessed with Harter’s (1985) six-item global self-worth scale; each item is scored from 1 to 4 (averaged for the scale total); Cronbach’s alpha was .73. Self-efficacy for dominance, popularity, and attractiveness was measured by having children rate how hard versus easy it was for them to enact each of 18 behaviors. Children rated each behavior on a 4-point scale, with 1 indicating HARD! and 4 indicating EASY! Dominance self-efficacy was assessed with eight items (e.g., “Bossing others around is ____ for me,” “Making others feel like I am in charge is ____ for me”). Popularity self-efficacy was assessed with six items (e.g., “Being popular is ____ for me,” “Being cool is ____ for me”). Attractiveness selfefficacy was assessed with four items (e.g., “Looking good in a bathing suit is ____ for me,” “Being good looking is ____ for me”). Cronbach’s alphas for these scales, respectively, were .79, .84, and .87. Assessment of Social Behavior The study required assessments of children’s gender-atypical behavior as well as of their aggression toward each classmate. These measures came from separate peer nomination instruments, described in turn. Only the names of children with parental permission to participate were included on these instruments. Assessment of gender nonconformity (and other social behaviors) in the peer group. Fourth and fifth graders provided peer nomination data for all of their classmates. Because sixth and seventh graders did not have a homeroom class but knew all the children in their grade very well, each child provided nominations for 14 –15 grade-mates (a randomization procedure ensured that each child could nominate, and be nominated by, a similar number of peers). We use the term classmates to refer both to fourth and fifth graders’ classmates and to sixth and seventh graders’ grademates. The mean number of peers serving as nominators for a child was 16.1 (minimum 14). A 23-item peer nomination inventory was administered. Items covered a broad array of social behaviors in the peer group. Most items came from previous studies (e.g., Pauletti et al., 2012), but four that assessed gender-atypical behavior were written for the present study (from the boys’ form: “He’d rather play with girls than with boys,” “He likes to do things that girls usually do,” “He acts like a girl,” and “He always plays with girls”). Children checked off the names of classmates who fit each item (unlimited nominations); children crossed their own name off the list so they did not nominate themselves. A child’s score on each item was the proportion of classmates (of either sex) who nominated the child for the item. A principal components analysis (with varimax rotation) on the data of all children in the fall yielded five components, one of which had very high loadings for all four gender-atypicality items (.83 to .92) and very low loadings for the remaining 19 items (.16 to .23); it was labeled gender nonconformity. The other components were labeled social prominence (e.g., popularity, sports competence), prosocial qualities (e.g., helpful behavior, physical attractiveness), coercion (e.g., dominance, manipulation), and internalizing problems (e.g., sadness, anxiety). A similar analThis document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 847
848 PAULETTL COOPER.AND PERRY SPSS Program which At each time,each child's raw ent of target-specific aggression.We needed a mea was a for the child).we assessed global aggression ony (i.ebeing mean ation by female peers was similarly computed. Results Descriptive Statistics ificant ways.The following instructions ere read to the h Table I gives the mean and standard deviation of each measure h kid subsequent analyses,me aggression and victim c m an to other kids.For ple,a kid could hit or pu nch so e measure re ted in Table2 ble3 fo nc nted a hoklet Ea had the erm gender of children'v bcha conformiry to refer to oral gender noncon lace for stion Four pag lly,each model involved two steps.First 14-15 bed bo which the child'sg on loward pee need over the ad pr al 10n target gender r 4 sion to ard his or her classmate gender nonconformity in t child's ntage nomina his or her aggression to vard gend step of th were ion tos ard bovs early all children (17 f195. n 0r91.8%) toward girls.)Tw rally victimized by peers (to hetic target influ mcan numbe sin the fall (spring)respectively. mate ion).A Level I equation was used to compute ildren' red victims,each child's cHLM model,the within-child slope d as the depend in analyses i al linear modeling (HLM) analyses eg Nezlek 2011)Thu thes were between-child 02).The analyses cont hild 63(3fo encra (36)for boys aggr on tov ard girls,53 (.22)for girls a lso ex ined).A vel this pur ose.A sion toward boys,and 46(.15)for girls'aggression towar d girls effect of the cognitive term in indicate
ysis on the spring data yielded a nearly identical structure. Component scores were computed for both the fall and the spring by the SPSS program, which yields uncorrelated factors. Children’s scores for gender nonconformity were of particular interest in this study. Assessment of target-specific aggression. We needed a measure of each child’s aggression toward each classmate. Because obtaining such data is time consuming and burdensome (especially for the child), we assessed global aggression only (i.e., being mean to someone); we did not repeat the assessment process for different subtypes of aggression (e.g., physical, relational). This procedure allowed children to consider a wide range of acts as mean, presumably increasing the likelihood that children would nominate all peers whom they perceived to be mean to someone in one or more significant ways. The following instructions were read to the child: We want to find out which kids sometimes do mean things to other kids, and who they are doing it to. There are many different ways to be mean to other kids. For example, a kid could hit or punch someone, tease someone or say something mean about them, or do something mean over the cell phone or the Internet. For each kid, we want you to tell us whether he or she is sometimes mean to each kid on the list. Children were then presented a booklet. Each page had the name of a different peer at the top, followed by a list of all of that peer’s classmates. Next to each of the classmates’ names there was a place for the child to check either “Yes” or “No” in response to the question, “Is (name of the peer at the top of the page) mean to (name of the classmate)?” Thus, the peer whose name appeared at the top of the page was a potential aggressor, and the children whose names followed were potential victims. Fourth and fifth graders’ booklets had a page for each classmate, and sixth and seventh graders had a page for each of the 14 –15 peers for whom they had provided the nominations described above. However, all children’s lists of potential victims included all of the potential aggressor’s classmates (both male and female). The mean number of peers listed as potential victims was 28.5 (minimum 14, maximum 43, minimum number of peers of a given sex 7). Participants’ own names were never in the booklet, so they did not nominate themselves as aggressors or as victims. A child’s raw aggression toward a peer was the percentage of nominators who named the child as mean to that peer. These scores were used in the first step of the multilevel analyses described later. A child’s raw victimization by a peer was the percentage of nominators who named the peer as mean to the child. Nearly all children (179 of 195, or 91.8%) were nominated by at least one classmate (in either fall or spring) as mean to a peer. Of the aggressor–victim pairs so identified, the minimum, maximum, and mean numbers of nominators in the fall (spring), respectively, were one (one), 13 (16), and 2.01 (2.19). To estimate stability over the year in children’s profiles of preferred victims, each child’s profile of raw aggression toward classmates in the spring was predicted from the child’s profile of raw aggression toward the same classmates in the fall, using the hierarchical linear modeling (HLM) program (Raudenbush & Bryk, 2002). The mean (and standard deviation) of these withinchild betas were .63 (.31) for boys’ aggression toward boys, .54 (.36) for boys’ aggression toward girls, .53 (.22) for girls’ aggression toward boys, and .46 (.15) for girls’ aggression toward girls. These moderate coefficients indicate there was some stability but also room for change in children’s targets over the year. At each time, each child’s raw aggression toward male peers was averaged to estimate the child’s general aggression toward male peers; a measure of general aggression toward female peers was similarly computed. Also at each time, each child’s raw victimization by male peers was averaged to estimate the child’s general victimization by male peers; a measure of general victimization by female peers was similarly computed. Results Descriptive Statistics Table 1 gives the mean and standard deviation of each measure in both the fall and the spring of the school year, as well as the stability of each measure from fall to spring. (To be consistent with subsequent analyses, measures of general aggression and victimization are given separately in relation to same-sex and other-sex peers.) Correlations among the measures are presented in Table 2 for the fall and in Table 3 for the spring. Henceforth, for clarity, we use the term felt gender typicality to refer to children’s self-rated gender typicality and the term gender nonconformity to refer to peers’ perceptions of children’s overt, behavioral gender nonconformity. Does Gender Identity Predict Children’s Aggression Toward Gender-Nonconforming Peers? Analysis strategy. To answer this question, we ran a series of HLM models. Conceptually, each model involved two steps. First, a measure was computed for each child capturing the degree to which the child’s aggression toward peer targets changed over the school year as a function of target gender nonconformity. That is, a within-child beta coefficient (slope) was calculated for each child predicting the child’s raw aggression toward his or her classmates in the spring from the classmates’ gender nonconformity in the fall. The beta controlled for the child’s raw aggression toward each classmate in the fall, so in effect it assessed change over the school year in the child’s propensity to attack peers based on their gender nonconformity. The larger the slope, the more the child increased his or her aggression toward gender-nonconforming peers (relative to other targets) over the year. (These within-child associations were computed separately for children’s aggression toward boys and children’s aggression toward girls.) Two other variables were controlled in these betas—the degree to which a classmate was generally victimized by peers (to remove nomothetic target influences) and classmates’ aggression toward the participant (to reduce the possibility that participants’ aggression was reactive to classmates’ aggression). A Level 1 equation was used to compute these within-child betas. In the second step of each HLM model, the within-child slopes just described served as the dependent variable in analyses in which the cognitive variables were predictors (slopes-as-outcomes analyses; e.g., Nezlek, 2011). Thus, these were between-child analyses. These analyses controlled for child general aggression (trait aggression), age, and sex (though interactions with sex were also examined). A Level 2 equation served this purpose. A significant effect of the focal cognitive term in this equation indicated This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 848 PAULETTI, COOPER, AND PERRY
AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 849 得点程用制 858话短85658自85吉台 行片g目 /s 8始粉学组838吉首台 8同斜然片88的第自第目 公云营8多居首三酱兰爸昌昌 3丙号有网g日昌 船8：8号含5吉首昌 莲店总君号时活 早实号t年8R8实司安F是的时 弦出
Table 1 Stability, Mean, and Standard Deviation of Each Measure Measure Stability Fall wave Spring wave Wave difference (t)a Boys Girls Sex difference in the fall (t) Boys Girls Sex difference in the spring (t) M SD M SD M SD M SD Cognitive variables Intergroup bias .46 0.48 0.53 0.55 0.52 0.74 0.56 0.56 0.60 0.50 0.17 1.61 Felt pressure for gender differentiation .83 2.54 0.79 1.69 0.50 8.80 2.56 0.82 1.60 0.49 9.74 1.14 Felt gender typicality .56 3.05 0.65 2.81 0.68 2.34 3.30 0.63 2.79 0.77 4.86 2.50 Gender contentedness .47 3.79 0.32 3.76 0.36 0.58 3.82 0.35 3.73 0.39 1.41 0.03 Self-esteem .49 3.56 0.47 3.60 0.42 0.70 3.62 0.45 3.60 0.54 0.38 0.98 Dominance self-efficacy .69 2.66 0.59 2.56 0.59 1.08 2.77 0.58 2.58 0.61 2.02 1.95 Popularity self-efficacy .79 3.32 0.61 3.29 0.56 0.43 3.41 0.62 3.34 0.56 0.91 2.49 Attractiveness self-efficacy .67 3.04 0.76 3.21 0.68 1.53 3.16 0.73 3.26 0.71 0.82 2.11 Behavioral variables Gender nonconformity .82 0.12 0.97 0.13 1.04 1.67 0.09 1.02 0.10 0.97 1.24 0.07 Social prominence .91 0.31 1.04 0.32 0.86 4.37 0.29 1.00 0.31 0.91 4.21 0.24 Prosocial qualities .85 0.36 0.87 0.42 0.96 5.77 0.37 0.92 0.40 0.93 5.66 0.42 Coercion .77 0.21 0.88 0.21 1.08 3.07 0.13 0.94 0.13 1.05 1.82 0.13 Internalizing problems .71 0.06 1.04 0.06 0.95 0.57 0.16 1.06 0.17 0.90 2.37 0.03 General aggression toward same-sex peers .84 0.06 0.08 0.04 0.06 2.07 0.08 0.10 0.05 0.07 3.06 5.30 General aggression toward other-sex peers .71 0.04 0.07 0.06 0.07 1.30 0.06 0.08 0.09 0.12 1.96 5.21 General victimization by same-sex peers .78 0.06 0.05 0.04 0.05 2.71 0.08 0.07 0.05 0.04 4.98 6.63 General victimization by other-sex peers .54 0.06 0.04 0.05 0.07 0.56 0.09 0.08 0.07 0.05 2.83 5.52 a Entries in this column are t values from repeated measures tests. p .05. p .01. p .001. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 849
PAULETTL COOPER.AND PERRY 949房8黄】 高特转1 导月并形是话 9日明 8兰6的：为话1头 月司9药肉：是1月 9智兰二2马新房兰房 9科青话为名两两 : 甲号9e991gg胡 月消酱的9的三18品 等月月为9片1后5日 45名站155活为二将月 9鼎1片 8活1行：8698等 0 居华内1号活超药二名9三国 含名名号1的遇酱将43行 4a6点两店南目 监8二时锅没！的8=两两只酱房月 书者前兰乌科舍5片普病8 319的六鸡=8的兰兰昌5 名18949号8用内 鸣为1常为三89989共95 车兰1内将生8将三村趋巴的9 吉1亏的莉多当后的司名上号=首丹 198符有9的方晋名5 19西月99片8片商338
Table 2 Correlations Among Variables in the Fall Measure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1. Intergroup bias — .04 .13 .00 .04 .08 .15 .00 .25 .12 .04 .05 .16 .11 .11 .08 .17 2. Felt pressure for gender differentiation .17 — .03 .02 .05 .18 .15 .12 .04 .01 .10 .12 .08 .13 .11 .01 .09 3. Felt gender typicality .03 .01 — .17 .43 .05 .20 .14 .05 .03 .08 .07 .08 .02 .03 .02 .05 4. Gender contentedness .24 .13 .17 — .40 .03 .07 .10 .01 .06 .03 .04 .02 .07 .05 .03 .12 5. Self-esteem .09 .23 .17 .38 — .05 .20 .25 .00 .01 .17 .15 .13 .18 .19 .18 .05 6. Dominance self-efficacy .10 .09 .01 .01 .15 — .25 .43 .16 .28 .31 .23 .36 .14 .18 .08 .03 7. Popularity self-efficacy .22 .12 .09 .12 .23 .47 — .61 .03 .40 .14 .19 .42 .19 .09 .31 .26 8. Attractiveness self-efficacy .17 .05 .10 .23 .21 .20 .51 — .01 .33 .05 .15 .45 .01 .07 .09 .09 9. Gender nonconformity .17 .01 .02 .03 .05 .20 .04 .06 — .31 .12 .03 .18 .19 .33 .29 .30 10. Social prominence .11 .13 .08 .05 .11 .33 .25 .04 .50 — .27 .20 .09 .16 .11 .40 .45 11. Prosocial qualities .06 .01 .08 .12 .06 .05 .10 .06 .24 .07 — .13 .12 .47 .47 .44 .44 12. Coercion .25 .09 .14 .12 .21 .16 .30 .28 .03 .07 .00 — .08 .58 .42 .19 .13 13. Internalizing problems .17 .17 .21 .15 .18 .21 .38 .13 .16 .04 .05 .08 — .31 .25 .55 .29 14. General aggression toward same-sex peers .21 .10 .25 .04 .09 .11 .13 .09 .02 .25 .50 .36 .27 — .84 .60 .45 15. General aggression toward other-sex peers .08 .11 .23 .04 .15 .04 .15 .15 .10 .17 .50 .53 .14 .75 — .51 .48 16. General victimization by same-sex peers .09 .08 .14 .23 .04 .00 .25 .14 .05 .26 .35 .12 .38 .15 .10 — .63 17. General victimization by other-sex peers .01 .12 .05 .14 .07 .08 .02 .09 .11 .18 .21 .12 .23 .20 .25 .57 — Note. Correlations for boys are above the diagonal; correlations for girls are below the diagonal. Entries are partial correlations with age controlled. p .05. p .01. p .001. Table 3 Correlations Among Variables in the Spring Measure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1. Intergroup bias — .11 .18 .13 .26 .24 .18 .03 .21 .11 .06 .12 .08 .12 .07 .09 .17 2. Felt pressure for gender differentiation .16 — .08 .12 .02 .08 .09 .10 .02 .00 .12 .07 .08 .22 .18 .04 .04 3. Felt gender typicality .01 .10 — .37 .38 .21 .17 .09 .01 .15 .04 .01 .09 .00 .01 .05 .19 4. Gender contentedness .01 .01 .23 — .47 .18 .12 .06 .09 .09 .03 .08 .08 .01 .03 .03 .02 5. Self-esteem .10 .34 .28 .36 — .19 .39 .24 .06 .13 .20 .01 .33 .24 .27 .26 .24 6. Dominance self-efficacy .03 .25 .14 .29 .10 — .45 .54 .12 .33 .27 .21 .51 .01 .10 .32 .16 7. Popularity self-efficacy .13 .19 .20 .14 .49 .33 — .69 .07 .47 .25 .16 .44 .27 .15 .49 .33 8. Attractiveness self-efficacy .15 .03 .28 .06 .27 .22 .53 — .04 .51 .13 .19 .40 .09 .04 .39 .29 9. Gender nonconformity .09 .02 .14 .05 .06 .12 .06 .23 — .34 .08 .09 .00 .21 .26 .22 .10 10. Social prominence .14 .20 .22 .16 .05 .35 .11 .04 .51 — .27 .23 .17 .20 .13 .42 .31 11. Prosocial qualities .02 .04 .05 .06 .13 .09 .10 .08 .21 .01 — .11 .10 .55 .51 .35 .31 12. Coercion .17 .02 .19 .16 .17 .17 .23 .26 .11 .17 .01 — .08 .41 .42 .15 .04 13. Internalizing problems .03 .04 .12 .13 .33 .12 .37 .12 .03 .11 .05 .08 — .28 .10 .63 .34 14. General aggression toward same-sex peers .24 .02 .21 .10 .10 .15 .10 .13 .07 .16 .47 .42 .02 — .84 .66 .38 15. General aggression toward other-sex peers .04 .13 .15 .07 .06 .08 .08 .09 .19 .03 .27 .48 .02 .72 — .48 .41 16. General victimization by same-sex peers .04 .09 .02 .07 .20 .24 .30 .14 .09 .16 .31 .00 .54 .28 .18 — .51 17. General victimization by other-sex peers .00 .21 .07 .08 .27 .00 .21 .11 .08 .11 .13 .24 .51 .39 .51 .58 — Note. Correlations for boys are above the diagonal; correlations for girls are below the diagonal. Entries are partial correlations with age controlled. p .05. p .01. p .001. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 850 PAULETTI, COOPER, AND PERRY
AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 851 that the cognitive variable moderated the influence of targets' +yaa*(ParAgel）+ya*(ParCogl,）+ta ParAggToTar2y=bo+by*(ParAggToTarl +b2*(NomVicOfTarlu) nths.ParCogl is the participant's fall cognitive meas +by*(TarAggToParl) The main-effect impact of ach of our eight individua sion tow (fouon chang target gender-atypical behavior e r-nonconform as examined in this way ation for ed by combinations of variables than b ba)in ated how ch n the year single 20 wo-way nr f the cognitive var pendent d create a able (the vithin ngare list following basic Level 2 equation wasused: equation (above)was replaced by three new terms the main e ua Selective Targeting of Gender-Nonconforming Peers f Gender Ider e Over Time in Children'? Fixed effect Random effects Interaction term SE SD 67 cac 3893 21236 8888 nt.Iw
that the cognitive variable moderated the influence of targets’ gender nonconformity on change in participants’ aggression over the school year. Thus, the effect indicated a cross-level interaction of a subject variable with a target variable. The Level 1 equation used to calculate the within-child betas was ParAggToTar2ij b0j b1j * (ParAggToTar1ij) b2j * (NomVicOfTar1ij) b3j * (TarAggToPar1ij) b4j * (TarGenderNoncon1ij) rij. In this equation, ParAggToTar2 and ParAggToTar1 are participant’s aggression toward target in spring and fall, respectively; b0j is the intercept; NomVicOfTar1 is nomothetic victimization of target in fall; TarAggToPar1 is target’s aggression toward participant in fall; TarGenderNoncon1 is target gender-atypical behavior in fall; and rij is error. All variables were group-mean centered. The beta yielded by the equation for the TarGenderNonconp1 term (b4j) indicated how change in the child’s aggression over the year depended on target’s gender-atypical behavior in the fall. To evaluate effects of the cognitive variables on the dependent variable (the within-child betas), we ran a series of HLM models. To evaluate the main effect of a single cognitive variable, the following basic Level 2 equation was used: B4j 40 41 * (ParGenAgg1j ) 42 * (ParSexj ) 43 * (ParAge1j ) 44 * (ParCog1j ) u4j . In this equation, B4j is the within-child beta yielded by the Level 1 equation for the TarGenderNoncon1 term, 40 is the intercept, ParGenAgg1 is participant’s general aggression in fall, ParSex is the participant’s sex, ParAge1 is the participant’s fall age in months, ParCog1 is the participant’s fall cognitive measure, and u4j is error. All variables were grand-mean centered. The significance of the coefficient of the cognitive variable indicated whether that variable predicted change in children’s aggression as a function of peers’ gender nonconformity. To evaluate a Cognition Sex interaction, the model was rerun to include the interaction term. The main-effect impact of each of our eight individual cognitive variables (four gender identity dimensions, four gendernonspecific self-appraisals) on change in children’s targeting of gender-nonconforming peers was examined in this way. However, overconfident and insecure forms of gender identity are better captured by combinations of cognitive variables than by any single variable. Thus, we also evaluated 20 two-way interactions among the cognitive variables, representing the pairings of variables that we suggested create an overconfident or insecure gender identity. The pairings are listed in Table 4. To evaluate a two-way interaction, the single cognitive term in the basic Level 2 equation (above) was replaced by three new terms—the main Table 4 Results of Hierarchical Linear Modeling Analyses Evaluating Influences of Gender Identity on Change Over Time in Children’s Selective Targeting of Gender-Nonconforming Peers Fixed effects Random effects Interaction terma SE t SD 2 df Intergroup bias Felt gender typicality 0.33 .39 0.87 .35 219.06† 188 Gender contentedness 0.80 .67 1.20 .33 218.09† 188 Self-esteemb 2.64 .78 3.40 .35 113.02 95 Dominance self-efficacy 0.20 .35 0.58 .36 219.29† 188 Popularity self-efficacyc 1.28 .45 2.86 .37 207.03 186 Attractiveness self-efficacyd 0.93 .35 2.64 .33 207.57 186 Felt pressure for gender differentiation Felt gender typicalitye 0.92 .36 2.55 .36 210.46 186 Gender contentedness 1.16 .44 2.61 .30 212.31 188 Self-esteem 0.95 .33 2.85 .27 209.47 188 Dominance self-efficacy 0.46 .24 1.91 .34 215.47† 188 Popularity self-efficacy 0.30 .25 1.19 .35 217.91† 188 Attractiveness self-efficacy 0.26 .22 1.22 .35 216.55† 188 Felt gender typicality Self-esteem 0.71 .44 1.62 .35 215.89† 188 Dominance self-efficacy 0.46 .36 1.25 .35 218.05† 188 Popularity self-efficacy 0.86 .35 2.44 .32 213.40 188 Attractiveness self-efficacy 0.86 .29 2.94 .34 211.45 188 Gender contentedness Self-esteem 0.90 .88 1.03 .33 217.53† 188 Dominance self-efficacy 1.11 .54 2.06 .38 215.75† 188 Popularity self-efficacy 0.88 .46 1.91 .39 216.45† 188 Attractiveness self-efficacy 1.21 .46 2.65 .38 212.36 188 a When an interaction term was significant, it was followed up to see if the gender identity variable preceding moderated the relation of the other variable to selective targeting of gender-nonconforming peers. b Data in this row are for boys only. c Data in this row are for the interaction of intergroup bias and the quadratic effect of popularity self-efficacy. d Data in this row are for the interaction of intergroup bias and the quadratic effect of attractiveness self-efficacy. e Data in this row are for the interaction of felt pressure and the quadratic effect of felt gender typicality. † p .10. p .05. p .01. p .001. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. AGGRESSION TOWARD GENDER-NONCONFORMING PEERS 851
PAULETTL COOPER.AND PERRY riable of nd uadratic term.as described next).we exa ined its pattern If athway was ted if high scores on nder ure o way interaction and the tw ctions of sex with to omote hacas nent of gender- ning p variable in turn (in order of Table 4 9 variahle this was done by runnine s ent models tha nt of gender. ning peers,but there was the tw main effects e evidence r the overconfide ntial mod If this der identity variable with thre 10n c self-app elf ef was he i inte ith self-e uded at Level one predicting the Bias Self-Esteen included such an ea cept m Leve ollo towith fal (19 d at 1 and SD of t n.Ou influ nces of cognition on trait with s ng ingroup fay tism (p 001 simulta d thei rd h s and for to be moder ately effects.Th attr rate sim wi h other-s 003 ut not signif ing child) arget Because the b towa nent of s ing peers That childr high in bot ion toward othe sex p ports the overconfide t pathway:that childr ted few r set of analyses We first summ ergrolf-di s but low win self-efficacy also do so supports the ggre sion to ard gende onconforming ame. ex peers pressure for er ntiation Felt pres ssure fo edict in ever,the inte on of sex and self-esteem wa cks on gender-nor onforming peers.These interac e-estem fore 081D0291 in a level 2 equation predictine slopes ure.but it is more sistent with th To evalu cure ant pathwa than ray interactions list that high s res on nigh felt was more conducive than lov felt e gender identity variable combined with high scores on anothe ictimization of gender-nonconforming peers (b 0.75.p-
effects of the two cognitive variables and their product. If the interaction was significant (and not qualified by child sex or by a quadratic term, as described next), we examined its pattern. If, however, the three-way interaction of child sex and the cognitive variables was significant (in a subsequent model that included the three-way interaction and the two-way interactions of sex with each cognitive variable), we evaluated the interaction of the cognitive variables separately for each sex. Also for each two-way interaction, we examined whether the first-named member of each variable pair listed in Table 4 moderated a quadratic effect of its partner variable; this was done by running subsequent models that included not only the two main effects and their interaction but also the quadratic effect of the partner variable and the interaction of this quadratic variable with its potential moderator. If this interaction was significant, we examined its pattern. (We also ran models to see if any such interactions were qualified by child sex; none was.) When a Level 2 equation is used to predict the slopes from a Level 1 equation, the model may be misspecified unless a second equation is included at Level 2— one predicting the intercepts from Level 1 (from the same terms as those predicting the slopes). We therefore always included such an equation at Level 2. An intercept from Level 1 estimates a participant’s general (trait) aggression in the spring. Because this extra equation predicts the intercepts from cognition with fall general aggression controlled, the coefficients for the cognitive terms indicate whether cognition predicts change over the year in trait aggression. Our hypotheses do not address influences of cognition on trait aggression, but we later comment on these effects. Although we have described the analysis as proceeding in two steps (from Level 1 to Level 2), in reality the HLM program estimates all equations simultaneously. Each HLM analysis included the data for boys’ aggression toward boys and for girls’ aggression toward girls. Participant sex is therefore confounded (deliberately) with target sex, but as noted, sex was examined both as a main effect and as a moderator of cognitive effects. This analytic strategy is appropriate when testing hypotheses expected to operate similarly in both boys’ and girls’ interactions with same-sex peers but not necessarily in their interactions with other-sex peers (e.g., Rodkin & Berger, 2008; Sainio, Veenstra, Huitsing, & Salmivalli, 2012). We also ran a parallel series of analyses examining children’s aggression toward othersex targets. Because the base rate of children’s aggression toward other-sex peers is very low and because there is less reason to expect gender identity to affect aggression toward other-sex peers, we expected few results for this set of analyses. We first summarize results for same-sex targets. Aggression toward gender-nonconforming same-sex peers. No cognitive variable predicted change in children’s maltreatment of gender-nonconforming peers over the school year as a main effect. However, the interaction of sex and self-esteem was significant (1.40, p .009, t 2.64). For boys only, low self-esteem forecast increased targeting of gender-nonconforming peers (0.81, p .029, t 2.22). (The gamma symbol denotes an unstandardized beta coefficient from a cognitive term in a Level 2 equation predicting slopes.) To evaluate the overconfident and insecure pathways to aggression toward gender-nonconforming peers, we tested the 20 twoway interactions listed in Table 4. Evidence that high scores on a gender identity variable combined with high scores on another variable to encourage maltreatment of gender-nonconforming peers was taken as support for the overconfident pathway. The insecure pathway was seen as supported if high scores on a measure of between-gender identity or low scores on a measure of within-gender identity combined with low scores on another variable to promote harassment of gender-nonconforming peers. Of the 20 interactions tested, 10 were significant, only one of which was moderated by child sex. We describe the significant interactions for each gender identity variable in turn (in order of Table 4). As we shall see, most results supported the insecure pathway to maltreatment of gender-nonconforming peers, but there was also some evidence for the overconfident route. Intergroup bias. Intergroup bias did not interact with any other gender identity variable but did interact with three gendernonspecific self-appraisals—self-esteem, popularity self-efficacy, and attractiveness self-efficacy—to predict increased victimization of gender-nonconforming peers. Because the interaction of intergroup bias with self-esteem was qualified by child sex (three-way interaction 2.63, p .018, t 2.40), the two-way interaction of Intergroup Bias Self-Esteem was examined for each sex. This interaction was significant for boys (p .001) but not girls (details of significant interactions are in Table 4). Significant interactions were followed up using procedures recommended by Aiken and West (1991), that is, the effect of one variable was estimated at 1 and 1 SD of the other (conceptualized as the moderator). As seen in Figure 1A, low self-esteem predicted victimization of gender-nonconforming peers only for boys with strong ingroup favoritism (p .001). This supports the insecure, self-discrepant pathway to harassment of gendernonconforming peers: Boys who saw their own gender as superior but nonetheless questioned their own worth grew increasingly mean to gender-nonconforming peers with time. (This was the only interaction to be moderated by child sex; all remaining interactions in this section capture patterns for the total sample.) Intergroup bias also interacted with the quadratic form of both popularity self-efficacy and attractiveness self-efficacy (see Figures 1B and 1C). In each case, the quadratic effect of the selfefficacy variable was highly significant for children with strong intergroup bias (both ps .003) but not significant for children with little intergroup bias: For children with high intergroup bias, either very low or very high self-efficacy led to selective harassment of gender-nonconforming peers. That children high in both intergroup bias and self-efficacy for a demonstration goal harass these peers supports the overconfident pathway; that children high in intergroup bias but low in self-efficacy also do so supports the insecure, self-discrepant pathway. Felt pressure for gender differentiation. Felt pressure for gender differentiation interacted with three variables—felt gender typicality, gender contentedness, and self-esteem—to predict increased attacks on gender-nonconforming peers. These interactions are depicted in Figures 2A, 2B, and 2C. Felt pressure interacted with the quadratic form of felt typicality (p .012; see Figure 2A). The pattern does not provide strong support for either pathway because the quadratic effect of felt typicality is not significant at high felt pressure, but it is more consistent with the insecure, self-discrepant pathway than with the overconfident pathway. This is because for children who felt gender atypical, high felt pressure was more conducive than low felt pressure to victimization of gender-nonconforming peers (b 0.75, p This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 852 PAULETTI, COOPER, AND PERRY
