C 2004) Journal of Abnormal Child Psychology, Vol. 32, No. 5, October 2004, pp. 551–563 ( Effect of Children’s Perceived Rejection on Physical Aggression Vito S. Guerra,1,3 Steven R. Asher,1 and Melissa E. DeRosier2 Received July 29, 2003; revision received March 4, 2004; accepted March 10, 2004 This study investigated whether the perception of self as socially rejected might contribute to increased physical aggression among elementary-school children. It was hypothesized that physically aggressive children would become more physically aggressive over time if they perceived that they were rejected and tended to blame peers for social failure experiences. Third-grade boys and girls (n = 941) were assessed in the Fall and Spring of the school year. Peer-report data on physical aggression and social preference were collected, along with self-report data on perceived rejection and attributions for social failure experiences. Results for boys were consistent with hypotheses, whereas the results for girls revealed a different pattern of relations. These results constitute prospective evidence that children’s self-perceptions of social rejection can uniquely influence externalizing behavior. Results are discussed in terms of mechanisms that might mediate the relation between perceived rejection and physical aggression. KEY WORDS: perceived rejection; physical aggression; social preference; social attributions. INTRODUCTION their own social acceptance. They also described, anecdotally, marked individual differences in the degree to which such self-perceptions were concordant with children’s “actual” social status, an observation strongly supported in later empirical work (Boivin & Begin, 1989; Cillessen & Bellmore, 1999; Hughes, Cavell, & Grossman, 1997; Hymel, Bowker, & Woody, 1993; Patterson, Kupersmidt, & Griesler, 1990; Zakriski & Coie, 1996). Hymel and Franke hypothesized that these individual differences might mediate associations between rejection by peers and later differential outcomes, and urged researchers in the field to move beyond questions of perceptual accuracy, asserting that self-perceptions of social acceptance would influence children’s affective responses and social behavior whether or not those perceptions were veridical. Although Hymel and Franke (1985) did not make specific predictions regarding the affective and behavioral sequelae likely associated with individual differences in children’s perceptions of social status, Hymel and her colleagues, among others, have since focused on the mediational role played by such perceptions in proposed “pathways” linking sociometrically assessed rejection to later internalizing problems, such as depressive Children’s social status, a construct indexing the extent to which a child is broadly liked or accepted by peers, has received a great deal of empirical attention over the past two decades, owing largely to evidence suggestive of a link between children’s rejection by peers and later internalizing, externalizing, and academic problems (e.g., Bierman & Wargo, 1995; Boivin, Hymel, & Bukowski, 1995; Coie, Terry, Lenox, Lochman, & Hyman, 1995; Deater-Deckard, Dodge, Bates, & Pettit, 1998; Dodge et al., 2003; Kupersmidt & Patterson, 1991; Ollendick, Weist, Borden, & Greene, 1992; Renshaw & Brown, 1993). Several years into the current era of peer relations research, however, Hymel and Franke (1985) noted the dearth of information available regarding children’s perceptions of 1 Department of Psychology: Social and Health Sciences, Duke University, Durham, North Carolina. 2 Frank Porter Graham Child Development Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina. 3 Address all correspondence to Vito S. Guerra, Department of Psychology: Social and Health Sciences, Duke University, Box 90085, Durham, North Carolina 27708; e-mail: [email protected]. 551 C 2004 Springer Science+Business Media, In. 0091-0627/04/1000-0551/0 552 symptomatology, loneliness, and withdrawal (e.g., Boivin et al., 1995; Panak & Garber, 1992; Rubin, Hymel, Mills, & Rose-Krasnor, 1991; Sandstrom & Zakriski, 2004). Panak and Garber (1992) tested such a model in a shortterm longitudinal investigation and found that low perceived social acceptance fully mediated the association between sociometric rejection and depressive symptomatology among the elementary-school children in their sample. At least two other studies employing prospective designs have likewise found low perceived social acceptance to be predictive of internalizing problems (Hymel, Rubin, Rowden, & LeMare, 1990; Kistner, Balthazor, Risi, & Burton, 1999). In Kistner et al.’s (1999) research, negative social self-perceptions were associated with heightened depressive symptomatology even in the absence of “actual” peer rejection, supporting Hymel and Franke’s (1985) hypothesis that the perception of social rejection can have significant affective–motivational consequences irrespective of the accuracy of that perception. Internalizing sequelae notwithstanding, Coie and his colleagues (e.g., Coie, 1990; Coie et al., 1995; Sandstrom & Coie, 1999; Zakriski & Coie, 1996) have suggested that the perception of rejection might ultimately advantage children who are, in fact, rejected by their peers, especially in those cases where rejected children acknowledge responsibility for their own social circumstances. The perception of rejection functions as a double-edged sword in this view, possibly precipitating the internalizing problems discussed above, but also engendering socially adaptive behavior change. That is, one can paint a picture of a child who is saddened and troubled by the rejection the child accurately perceives, but who is thereby motivated to make the changes, behavioral and otherwise, necessary to improve his or her social fortunes. For example, although data suggest that many (physically) aggressive–rejected children maintain overly optimistic views of their own social standing (e.g., Hughes et al., 1997; Hymel et al., 1993; Zakriski & Coie, 1996), those aggressive–rejected children who are aware of their negative status may make efforts to decrease their physically aggressive behavior in order to secure a better standing among peers. From this perspective, it is children who perceive themselves to be accepted in the face of actual sociometric rejection who may be at most risk for continued social and behavioral problems (for similar views, see Boivin, Poulin, & Vitaro, 1994; Hughes, Cavell, & Prasad-Gaur, 2001; Patterson et al., 1990). At least two studies have, in fact, reported a concurrent association between inflated perceptions of acceptance and heightened levels of physical aggression (David & Kistner, 2000; Hughes et al., 1997), although to our knowledge such a relation has not been Guerra, Asher, and DeRosier demonstrated prospectively (see Hughes et al., 2001). In summary, then, while the perception of peer rejection has been hypothesized to be a risk factor for internalizing outcomes, perceived rejection has been hypothesized to have, if anything, a salubrious influence on externalizing behavior. It is suggested here, however, that the perception of rejection can, under certain conditions, predict an increase in at least one salient type of externalizing behavior, that of physical aggression. As such, the primary goal of this study was to examine prospectively the influence of perceived rejection on children’s physical aggression. For reasons to be articulated below, the direction and strength of this influence was expected to depend jointly upon children’s attributions for social failure experiences, and upon children’s initial level of physical aggression. Attributions for Social Failure Experiences A number of investigators (e.g., Asher, Parkhurst, Hymel, & Williams, 1990; Sobol & Earn, 1985) have suggested that children’s causal attributions for social failure experiences might influence their affective and behavioral responses to peer rejection. (Note that these investigators make the implicit assumption that the child is aware of his or her social rejection.) Although several dimensions of causal attributions have been identified by attribution theorists (see Weiner, 1990), the degree to which children internalize versus externalize responsibility for social failure (e.g., self- vs. peer-blaming) has received the most consistent attention within the peers literature (e.g., Crick & Ladd, 1993; Ollendick et al., 1992; Sandstrom & Coie, 1999). As has been true with respect to the existing literature on perceived rejection, children’s attributions regarding social failure have been discussed primarily in the context of internalizing outcomes. Specifically, researchers have hypothesized that sociometrically rejected children who tend to make internal attributions for social failures (e.g., an attribution of “personal incompetence”) are those at particular risk for experiencing internalizing symptomatology. Although no direct test of this moderation model has been published, internal (and stable) attributions for social failure experiences have predicted depressive symptomatology (Panak & Garber, 1992), loneliness (Renshaw & Brown, 1993), and withdrawal (Goetz & Dweck, 1980). It is more common, however, for children to make external attributions for social failure experiences (Sobol & Perceived Rejection and Physical Aggression Earn, 1985), a self-serving bias observed throughout the broader attribution literature (see Weiner, 1990). Blaming peers for one’s social difficulties is one clear example of an external attribution. Recent research suggests that sociometrically rejected children, especially those who maintain their negative status over time, make this type of external attribution at a higher rate than do other children (Crick & Ladd, 1993; Sandstrom & Coie, 1999). The broader attribution literature suggests that blaming others for personally relevant negative outcomes often elicits anger (see Weiner & Graham, 1984). Moreover, it seems reasonable to predict that children who tend to blame peers for negative social experiences will likely subscribe to the belief that peers are generally hostile and “rejecting” and therefore maintain fairly negative expectations regarding future interactions with such peers. These types of beliefs and expectations have been linked to both physically aggressive behavior and aggressionencouraging cognitive biases in a number of studies (Burks, Laird, Dodge, Pettit, & Bates, 1999; Downey, Lebolt, Rincon, & Freitas, 1998; Lochman & Dodge, 1998; MacKinnon-Lewis, Rabiner, & Starnes, 1999). Downey et al. (1998, study 3), for example, presented fifth- through seventh-grade boys and girls with hypothetical vignettes depicting ambiguous social rejection scenarios. They reported that children’s “angry expectations of rejection” in response to these vignettes predicted increased self- and teacher-reported physical aggression 1 year later, after controlling for earlier levels of aggression. In another study, Burks et al. (1999) found that eighth-grade children who held “hostile” generalized beliefs about peers (e.g., “mean” rather than “friendly”; “jerks” rather than “cool”) were more likely than others to display externalizing symptomatology, as well as biases in social information-processing (a latent variable indexing hostile attribution biases, aggressive response generation, and aggressive response selection) that have frequently been linked to physically aggressive responding (see Crick & Dodge, 1994; Dodge, Pettit, McClaskey, & Brown, 1986). The influence of children’s expectations/beliefs about peer interactions on their social information-processing was also examined by Lochman and Dodge (1998), who found that boys’ expectations regarding the level of aggression likely to be displayed by both self and partner during a dyadic interaction task predicted their postinteraction perceptions of enacted aggression, even after controlling for actual observed aggression. This would suggest that children who expect aggression are biased to perceive it, a bias linked to increased aggressive behavior (Lochman, 1987). 553 Baseline Physical Aggression Children who both perceive that they are rejected and who tend to blame their peers for social failure experiences may, then, be primed to be physically aggressive towards their peers. Of course, not all children thus biased towards aggressive responding will necessarily behave aggressively (Erdley & Asher, 1996; Quiggle, Garber, Panak, & Dodge, 1992). Countervailing factors can function to inhibit actual aggressive responding. For example, Erdley and Asher (1996) studied children who tended to attribute hostile intent to peers (i.e., children vulnerable to “reactive” aggression; see Crick & Dodge, 1994), and found that there existed a sizable subgroup of these children who tended to endorse nonaggressive behavioral responses; children in this subgroup were both less confident in their ability to enact aggressive behavior (low self-efficacy for aggression; see Bandura, 1977) and less likely to report antisocial goals (e.g., revenge-seeking) than were children who tended to endorse aggressive behavioral responses. Although these and other potential countervailing constraints, such as negative outcome expectancies for aggressive behavior (see Crick & Werner, 1998; Perry, Perry, & Rasmussen, 1986) and negative beliefs about the morality or appropriateness of aggression (see Erdley & Asher, 1998; Huesmann & Guerra, 1997), were not assessed in the current study, it is presumed that children already engaging in significant aggressive behavior are those least likely to be inhibited by countervailing factors when “primed” by the anger, negative expectations, and perceptual/attributional biases that are expected to be manifest when children both perceive they are rejected and blame others for their social failures. On the basis of this reasoning, a three-way interaction between perceived rejection, social failure attribution (i.e., self- vs. peer-blame for social failure experiences), and baseline physical aggression was hypothesized. That is, among children high in baseline aggression, the perception of rejection coupled with the tendency to blame peers for social failure experiences was expected to lead to increased physically aggressive behavior. Children low in baseline aggression, on the other hand, were not expected to engage in increased aggression in response to the combined effects of perceived rejection and peer-blaming for social failure experiences. Further, and consistent with a hypothesis forwarded by Sandstrom and Coie (1999), the perception of rejection coupled with the tendency to blame oneself for social failure experiences was expected to lead to decreased aggressive behavior among children high in baseline aggression. 554 No a priori hypotheses regarding possible gender differences were generated. However, it has been common for investigators interested in children’s physical aggression to use all-male samples or, where gender is mixed, to employ analytical approaches (e.g., extreme-group analyses with relatively small, unbalanced cell sizes) that make it difficult to conduct tests of gender interactions with satisfactory statistical power (though see Crick, 1997; Egan, Monson, & Perry, 1998; Erdley & Asher, 1998; Feldman & Dodge, 1987). The hypotheses forwarded here are, then, based to some degree on prior research that may or may not be generalizable to girls. As such, a final goal of the present study was to consider whether the hypothesized three-way interaction described above would hold for girls as well as boys. METHOD Participants Eligible participants included all third-grade children (in mainstream classes) attending any of 11 selected schools. Each of these schools operated within a large County-wide public school system serving both urban and rural communities in the southeastern United States. The third-grade population within each of these 11 schools ranged from 100 to 169 children. The ethnic makeup of the total sample reflected that of the larger County— approximately 69% Caucasian, 27% African American, and 4% other ethnic minority groups (e.g., Hispanic; Asian). Individual-level data on family income and socioeconomic status (SES) were not obtained. However, the full range of SES was represented in the County from which the sample was drawn; approximately 30% of the children in these schools qualified for a free or reduced price lunch. Consent letters were sent to the parents of 1,374 children in September of the children’s third-grade year, and permission to participate was obtained for 1,255 children (91.34% consent rate). Because of either school absence or noncompletion of specific measures, full data on the instruments pertinent to this study were available for 1,071 children at Time 1 (October of the third-grade year). Of the 1,071 children with full data at Time 1, 941 also had full data available at Time 2 (April of the third-grade year). The attrition of approximately 12% between Time 1 and Time 2 was attributable either to school transfer or to school absence at the Time 2 assessment. All analyses to be reported below included the 941 children (470 boys and 471 girls) for whom complete Time 1 and Time 2 data were available. A series of t tests were conducted using Time 1 assessments to examine whether children who completed Guerra, Asher, and DeRosier both waves of data (n = 941) differed significantly from children lost to attrition (n = 130). The two groups did not significantly differ with respect to either self-report measure (perceived rejection; social failure attribution). However, children lost to attrition had higher peer-assessed aggression scores, t(1069) = 4.09, p < .01, and lower peer-assessed social preference scores, t(1069) = 4.16, p < .01, than did study completers. Procedure The data presented here were collected as part of a larger longitudinal investigation conducted by the third author. All instruments were group-administered within children’s classrooms. Time 1 (Fall) assessment occurred during two 30 to 45-min sessions in October of the children’s third-grade year. Time 2 (Spring) assessment occurred during two 30 to 45-min sessions in April of that same school year. The school year in this County had begun in mid-August, allowing children approximately 2 months of interaction with grademates prior to the initial assessment. Instruments Social Preference Participants were provided with rosters listing all of the children in their grade. They were asked to circle the names “of all the kids that you like the very most” using the first roster, and to circle the names “of all the kids that you like the very least” using the second roster. The use of an unlimited nomination procedure was guided by research suggesting that this methodology is psychometrically preferable to a limited nomination procedure (see Terry, 2000). The number of “like most” (LM) and “like least” (LL) nominations each participant received was calculated and then standardized within each of the 11 schools. Following the often utilized Coie, Dodge, and Coppotelli (1982) method, social preference (SP) scores were computed by subtracting children’s standardized LL score from their standardized LM score, and then standardizing that difference. In the current study, SP scores ranged from −4.53 to 3.07 at Time 1, and from −5.56 to 2.78 at Time 2. Physical Aggression During the session in which the sociometric assessment was completed, participants were provided with Perceived Rejection and Physical Aggression another roster listing all children in their grade and were asked to circle the names “of all the kids in your grade who fight a lot.” Single-item peer assessments of children’s physical aggression, such as the one employed here, are commonly used (e.g., Hecht, Inderbitzen, & Bukowski, 1998; Hughes et al., 1997; Hymel et al., 1993; Kupersmidt & Coie, 1990; MacKinnon-Lewis et al., 1999; Parkhurst & Asher, 1992; Patterson et al., 1990; Sandstrom & Coie, 1999) and have been found to be stable over time (e.g., Coie & Dodge, 1983; Deater-Deckard et al., 1998). A proportion score was calculated for each child by dividing the number of nominations that child received by the number of grade mates responding to the item (for a similar approach to calculating aggression scores, see Egan et al., 1998; Erdley & Asher, 1996; Parkhurst & Asher, 1992). So, for example, if child A was nominated by half of those grade mates completing the assessment, child A would have an aggression score equal to .50. This method was preferred to an often-used alternative means of determining a child’s aggression score, wherein nominations are summed and then standardized by classroom or school. Use of the proportion score, which was not standardized by school, more flexibly allows for the possibility that the schools may vary in the prevalence of physical aggression. While there are subtle conceptual implications of using one versus the other of these computations, it is worth noting that the correlation between the two derived variables is generally quite high. In this study, this correlation was .92 at Time 1 and .94 at Time 2. At Time 1, proportion scores ranged from .000 (i.e., a child nominated by 0 grade mates) to .382 (i.e., a child nominated by 38.2% of grade mates). At Time 2, scores ranged from .000 to .458. Perceived Rejection Children’s perceived rejection was assessed using a single-item measure from the Social Interactions Survey (SIS; DeRosier, 2001). SIS was designed to measure children’s perceptions regarding the quality of their relationships with peers, parents, teachers, and other adults. Each SIS item consists of a brief vignette describing a child in a particular social scenario. Participants are instructed to indicate the degree to which they fit the description of the child depicted within the vignette by drawing a mark along a line that is anchored on one end by the descriptor Not at all, and on the other end by the descriptor Exactly. The item specifically pertaining to the perception of rejection reads as follows: Some kids are very disliked or rejected by lots of other kids in their grade. 555 Often, other kids don’t like to play with them or even be on the same team with them. Other kids may try to avoid them or leave them out of their group. How much are you the kind of child that most other kids really don’t like? Exactly Not at all Participants’ responses were transformed using a 100point analog scale (a score of 50 was assigned to a mark at the midpoint of the line), such that scores ranged from 0 to 100, with higher scores indicative of greater perceived rejection. In a pilot study with third- through fifth-grade students, the 1-week test–retest reliability of the perceived rejection measure was .67 (DeRosier & Parkhurst, 2004). Drawing from the same sample used in the current study, DeRosier and Parkhurst also found the item to correlate in the expected direction (r = −.42, p < .001) with the “Social Acceptance” subscale of the Self-Perception Profile for Children (SPPC; Harter, 1982, 1985), the measure most frequently used in past research to measure children’s perceived acceptance/rejection (e.g., Boivin & Begin, 1989; Boivin & Hymel, 1997; Hughes et al., 1997, 2001; Kistner et al., 1999; Kupersmidt & Patterson, 1991; Panak & Garber, 1992; Patterson et al., 1990). The “Social Acceptance” subscale of SPPC, however, actually consists of items assessing perceptions about friendships and general social competence in addition to perceptions of acceptance, and thereby confounds these three constructs (see Zakriski & Coie, 1996, for a similar analysis). Moreover, there is no specific item on SPPC that taps whether children perceive themselves to be actively disliked (i.e., rejected). For these reasons, SPPC was not used in this study. Self- Vs. Peer-Blame for Social Failure Experience Children completed a modified version of the Peer Attribution Questionnaire (PAQ; see Panak & Garber, 1992), an instrument that assesses children’s attributions for both social success and social failure situations along three dimensions: internal–external (self vs. peer), global– specific, and stable–unstable. Each of the instrument’s 18 items presents two possible attributions that might be made in response to the situation described. Participants must decide which of the two attributions they are more likely to make, and then indicate the degree to which they endorse that attribution. For this study, we used the two items that gauged children’s tendency to place blame on either the self or peers in social failure situations. On the questionnaire, these items were separated by a number of other items. The two items were 556 Guerra, Asher, and DeRosier as follows: The regression model was also reestimated after excluding only the 17 children (2% of sample) who responded most inconsistently to these two items at Time 1 (i.e., assigned scale scores of 1 and 4 on the items). Again, results were very similar. RESULTS Mean-Level Differences as a Function of Gender Table I presents full sample and gender-specific means and standard deviations on each of the four continuous variables employed in this study. Gender differences on the two self-report variables were not statistically significant. Consistent with previous research, however, boys received significantly higher physical aggression scores than did girls at both Time 1 ( p < .001) and Time 2 ( p < .001), and received significantly lower social preference scores than did girls at both Time 1 ( p < .001) and Time 2 ( p < .01). Distribution of Perceived Rejection Scores Item responses were assigned a scale score of 1– 4, with higher scores indicative of a greater tendency to blame peers for one’s social failures. A summary score was derived by averaging children’s scores on these two items. Across the 941 children in the sample, item scores were moderately correlated (Time 1: r = .40, p < .001; Time 2: r = .41, p < .001). Detailed inspection of children’s responses suggested that children were responding to the two items in an internally consistent manner. At Time 1, 93% (n = 874) of the children gave responses to the two items that were either identical or within one scale score. The comparable statistic at Time 2 was 92% (n = 870). As a check, the full regression model described in the Results section was reestimated without the 67 children (7% of sample) who responded in a generally inconsistent manner to the two items at Time 1. Findings were very similar. Table II summarizes the distribution of perceived rejection scores at Time 1. The distribution of perceived rejection scores at Time 2 was similar and is not displayed. For the purpose of presenting this distribution of scores, children were classified as either aggressive or nonaggressive, and also as either rejected or nonrejected. Children were categorized as aggressive if they received physical aggression scores that were at least 1 standard deviation above the mean score; all other children were considered nonaggressive. At Time 1, 113 children (approximately 12% of the sample) met “aggressive” criteria. Children were categorized as rejected if they received a social preference score less than −1, a standardized “like most” score less than 0, and a standardized “like least” score greater than 0 (for an overview of sociometric Table I. Time 1 and Time 2 Summary Statistics on Continuous Variables Entire sample (N = 941) Boys (n = 470) Girls (n = 471) Measure Time 1 Time 2 Time 1 Time 2 Time 1 Time 2 Aggression Social Preference Perceived Rejection Social Failure Attrib. 0.04 (0.05) 0.07 (0.98) 29.43 (30.49) 3.18 (0.62) 0.04 (0.05) 0.07 (0.99) 28.26 (30.09) 3.08 (0.66) 0.06 (0.06) −0.07 (0.96) 30.21 (31.20) 3.18 (0.67) 0.06 (0.06) −0.02 (1.02) 27.87 (30.29) 3.03 (0.67) 0.03 (0.03) 0.21 (0.97) 28.64 (29.77) 3.19 (0.58) 0.03 (0.04) 0.16 (0.95) 28.64 (29.91) 3.13 (0.65) Note. Aggression is expressed as a proportion score (e.g., on average, boys were nominated by approximately 6% of their peers as being a child who “fights a lot”); Social Preference, a Z-Score, has a non-zero sample mean because SP was standardized by school; Perceived Rejection ranged from 0-100, with higher scores reflecting greater perceived rejection; Social Failure Attrib[ution] ranged from 1–4, with higher scores reflecting a greater tendency to blame peers for social failure. Values represent mean(standard deviation). Perceived Rejection and Physical Aggression 557 Table II. Distribution of Time 1 Perceived Rejection Scores Nonrejected (n = 831) Rejected (n = 110) Perceived rejection Entire sample (N = 941) Nonaggressive (n = 760) Aggressive (n = 71) Nonaggressive (n = 68) Aggressive (n = 42) 0–25 26–50 51–75 76–100 554 (58.87%) 168 (17.85%) 107 (11.37%) 112 (11.90%) 466 (61.32%) 141 (18.55%) 85 (11.18%) 68 (8.95%) 34 (47.89%) 11 (15.49%) 11 (15.49%) 15 (21.13%) 31 (45.59%) 11 (16.18%) 9 (13.24%) 17 (25.00%) 23 (54.76%) 5 (11.90%) 2 (4.76%) 12 (28.57%) Note. Median score for entire sample was 18. classification methodology, see Coie et al., 1982; Terry, 2000); all other children were considered nonrejected. At Time 1, 110 children (approximately 12% of the sample) met “rejected” criteria. As can be seen in the table, 23% of the full sample endorsed the perceived rejection item at a level greater than 50, the midpoint on the scale, and 12% of the full sample endorsed the item at a level greater than 75, a score that is approximately 1.5 standard deviations above the mean. Note also that although previous studies have found aggressive children to exhibit a tendency to inflate their self-perceptions of social acceptance (e.g., Hughes et al., 1997; Zakriski & Coie, 1996), the current data suggest that there are, nevertheless, a considerable percentage of aggressive children (some of them sociometrically rejected) who believe themselves to be quite strongly disliked. Zero-Order Correlations The within-time and across-time zero-order correlations of the four continuous measures are summarized in Tables III and IV, respectively. Gender-specific correlations are not displayed because correlations were of the same general strength and direction for boys and girls. With respect to the within-time correlations reported in Table III, several points are worth highlighting. First, the zero-order associations (or lack thereof) between variables Table III. Zero-Order Correlations Among Measures at Each Time of Testing Measure Aggression (AGG) Social Preference (SP) Perceived Rejection (PR) Social Failure Attribution (ATT) AGG SP — −.44∗ .06 −.11∗ −.44∗ — −.16∗ −.05 PR ATT .09∗ −.05 .02 .01 — −.10∗ — .01 Note. Fall correlations are above the diagonal; spring correlations are below the diagonal. ∗ p < .01. at the Fall assessment were quite similar to those obtained at the Spring assessment. Second, consistent with previous research (see Coie et al., 1990), social preference and aggression scores were negatively (and moderately) correlated. Third, at neither assessment was there a significant association between perceived rejection and social failure attributions. The independence of these two self-report measures is helpful with regard to testing the major hypotheses of the study. Finally, the correlation between perceived rejection and peer-assessed social preference was only −.10 in the Fall and −.16 in the Spring. The small magnitude of these two correlations is entirely consistent with data from previous studies (e.g., Hughes et al., 2001; Kistner et al., 1999; Panak & Garber, 1992). The across-time correlations displayed in Table IV reveal the high degree of stability of physical aggression and social preference scores. The stability coefficients for perceived rejection and social failure attributions were more modest. Note that there was not a significant zeroorder association between Fall perceived rejection and Spring aggression (r = .04, ns); nor did there exist a significant zero-order association between Fall social failure attribution and Spring aggression (r = −.08, ns). Testing the Moderation Model The hypotheses forwarded in this study were tested within a multiple regression framework, in which Time 1 Table IV. Zero-Order Correlations of Fall Measures With Spring Measures Spring measure Fall measure AGG SP PR ATT Aggression (AGG) Social Preference (SP) Perceived Rejection (PR) Social Failure Attribution (ATT) .83∗ −.43∗ .04 −.08 −.40∗ .79∗ −.08 .03 .06 −.16∗ .23∗ .02 −.09∗ −.03 −.06 .22∗ ∗p < .01. 558 (Fall) data was used to predict Time 2 (Spring) physical aggression. All variables except gender, which was dummy coded (girl = 0; boy = 1), were continuous. Following the suggestion of McClelland and Judd (1993), an alpha level of .10 was established for statistical tests of interactions. For all other tests, an alpha level of .05 was used. To examine whether the positively skewed distribution on the criterion variable (skewness = 2.85) would unduly bias tests of inference, the model to be described below was estimated twice, once using the raw proportion scores for aggression and once using an arcsine square root transformation of that proportion score. Results using the transformed and nontransformed variables were very similar. Since the nonnormal distribution of the criterion does not bias ordinary least squares (OLS) estimates of population parameters, only those results using the nontransformed proportion score are reported. The estimated model included all 941 children in the sample and formally tested for the presence of a four-way interaction between perceived rejection, social failure attribution, baseline (i.e., Fall) aggression, and gender. The procedures recommended by Aiken and West (1992) for the testing and probing of interaction effects were followed. When estimating the model, all lower order terms comprising the four-way interaction of interest were included. This constituted the four variables involved in the interaction, 6 two-way interactions, and 4 three-way interactions. Moreover, to facilitate the interpretation of regression coefficients and to minimize multicollinearity between predictor variables and their interaction terms, children’s scores on continuous measures other than social preference were standardized prior to estimating the regression model. Social preference scores had already been standardized by school, so these scores were centered, but not restandardized, prior to estimating the regression model. With the variables thus standardized, the “raw” regression coefficients output by statistical software packages are appropriately interpreted as standardized regression coefficients (betas [β]). These are the coefficients that are reported below. In the context of this study, betas index the standard deviation change in Spring aggression (SD = 0.054 [5.4%]) associated with a 1 standard deviation change in the Fall predictor. Variables were entered into the model in two steps. In the first step, Spring aggression was regressed on a block of control variables that included gender, Fall aggression, and Fall social preference. Approximately 69% of the variance in Spring aggression was accounted for by the linear combination of these three variables, F(3, 937) = 702.80, p < .0001. Fall aggression was strongly predictive of Spring aggression (β = .77, p < .0001). Gender and social preference also made significant unique contri- Guerra, Asher, and DeRosier Table V. Regression Analysis With All Predictor Terms Included Outcome (Time 2) Predictors (Time 1) .702∗∗∗ Aggression Gender Aggression (AGG) Social Preference (SP) Perceived Rejection (PR) Social Failure Attribution (ATT) PR × ATT PR × AGG PR × Gender AGG × ATT AGG × Gender ATT × Gender PR × ATT × AGG PR × ATT × Gender PR × AGG × Gender AGG × ATT × Gender PR × ATT × AGG × Gender †p R2 β .15∗∗∗ .65∗∗∗ −.10∗∗∗ −.05† −.02 −.03 −.11∗∗ .01 −.01 .15∗∗∗ −.01 −.05 .06 .12∗∗ −.02 .08∗ < .10. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. butions to the prediction of Spring aggression, t(937) = 3.52, p = .0004, and t(937) = −4.18, p < .0001, respectively. Boys were more likely than girls to behave aggressively in the Spring (β = .13), and higher Fall social preference scores were associated with lower levels of Spring aggression (β = −.09). The remaining predictor terms—perceived rejection, social failure attribution, 6 two-way interactions, 4 threeway interactions, and the four-way interaction between perceived rejection, social failure attribution, aggression, and gender—were entered in the second step. Perceived rejection and social failure attribution were not entered in an intermediate step because all hypotheses pertaining to these variables involved interactios.1 The full model accounted for just over 70% of the variance in Spring aggression, and the increment in the proportion of variance explained (.0145) was significant, F(13, 924) = 3.52, p < .01. This 1.5% increment in explained variance represents 5% of the variance in Spring aggression unaccounted for by the control variables. Table V displays the betas 1 We recognize that some readers will have an interest in viewing data that describe the relation between these two variables and Spring aggression without consideration of any of the hypothesized interactions. As such, in a separate analysis, Fall perceived rejection and Fall social failure attribution were entered as a second block of variables. There was a marginally significant inverse association between Fall perceived rejection and Spring aggression (β = −.03, p = .0797), and a significant inverse relation between Fall external attributions and Spring aggression (β = −.04, p = .0496). Caution need be taken in trying to interpret these betas as “main” effects, however, giving the significant interaction to be described. Perceived Rejection and Physical Aggression 559 Table VI. Relation (β) of Fall Perceived Rejection to Spring Aggression as a Function of Fall Aggression, Fall Social Failure Attribution, and Gender Fall aggression Boys Girls Fall Attribution 10th .8% 50th 4.0% 85th 10.9% 95th 17.6% 10th 0.0% 50th 1.9% 85th 5.2% 95th 8.3% Self-blaming Peer-blaming −.07 −.03 −.12 .00 −.23∗∗∗ .08† −.33∗∗∗ .15∗ .00 .07 .01 .00 .03 −.12∗∗ .05 −.24∗∗∗ Note. Column headings index level of Fall Aggression by both percentile and gender-specific raw proportion score. † p < .10. ∗ p < .05. ∗∗ p < .01. ∗∗∗ p < .001. associated with each predictor term included in the full model. Fall social preference, the only measured variable not involved in higher order predictor terms, continued to make a unique contribution to the prediction of Spring aggression (β = −.10, p < .0001). Gender, Fall aggression, 2 two-way interactions, 1 three-way interaction, and the four-way interaction also made significant unique contributions to the prediction of Spring aggression. Given the significant four-way interaction (β = .08, p = .0356), the standardized slopes (betas) associated with the five lower order predictor terms are best interpreted as conditional effects, characterizing the relation between the predictor and Spring aggression when all other predictors involved in the four-way interaction are at their mean. These lower-order effects will not be discussed further. To understand the nature of the significant four-way interaction, a series of betas was calculated, each describing the association between Fall perceived rejection and Spring aggression, given a particular constellation of “levels” on the three moderating variables (i.e., gender, Fall aggression, and Fall social failure attribution). The choice of probe levels for gender was self-evident. The choice of probe levels for Fall aggression and Fall social failure attribution was informed by both theoretical and empirical considerations. Selection of probe levels for physical aggression took into account the much wider range of proportion scores among children in the top, as opposed to the bottom, quartile. As such, probing was done at the 10th, 50th, 85th, and 95th percentiles of aggression for each gender (in absolute terms, this reflects eight probe levels for aggression). Two levels of social failure attributions were selected, one to index relatively high peer-blame for social failure (scale score of 4.0) and one to index relatively strong self-blame for social failure (scale score of 1.5). These somewhat asymmetrical probes reflect the negatively skewed distribution of attribution scores generally (only 1% of children received a scale score of 1.0 on the measure, whereas 22% of children received a scale score of 4.0). Probing “self-blame” at 1.5 rather than 1.0 greatly reduces the standard error associated with parameter estimation. Once levels had been chosen, probing of the four-way interaction was conducted within the full regression model, such that associations between Fall perceived rejection and Spring aggression at different levels of the moderator variables were considered only after controlling for all other predictors entered into the model. Table VI summarizes the simple slopes (βs) that characterize the conditional relations between Fall perceived rejection and Spring physical aggression. There was no significant relation between these variables among children who exhibited either low (10th percentile) or moderate (50th percentile) levels of Fall aggression. The relation of Fall perceived rejection to Spring aggression among children who displayed high levels of Fall aggression (85th and 95th percentiles) varied as a function of both gender and children’s attribution for social failure. Results for boys were consistent with hypotheses. Among boys who were aggressive and self-blaming in the Fall, higher levels of Fall perceived rejection were associated with lower levels of Spring aggression. By contrast, among boys who were aggressive and peer-blaming in the Fall, higher levels of Fall perceived rejection were associated with higher levels of Spring aggression. A different pattern of results emerged for aggressive girls. Among girls who were both aggressive and self-blaming in the Fall, there was no significant association between Fall perceived rejection and Spring aggression. Moreover, higher levels of Fall perceived rejection were associated with lower levels of Spring aggression among girls who were aggressive and peer-blaming in the Fall. To summarize, perceived rejection was significantly related to subsequent physical aggression for three “groups” of children, groups that can be described by their standing on measures of Fall aggression and Fall social failure attribution as follows: (a) aggressive boys who 560 tend to blame peers for social failure experiences, (b) aggressive boys who tend to blame themselves for social failure experiences, and (c) aggressive girls who tend to blame peers for social failure experiences. Perceived rejection was positively associated with later aggression among children meeting the criteria defining the first group, whereas perceived rejection was negatively associated with Spring aggression among children meeting the criteria defining either of the latter two groups. Although the findings pertaining to the two groups of boys are consistent with hypotheses, taken alone they are not unambiguously supportive, as regression coefficients do not necessarily index absolute change. Take, for example, only those boys who were aggressive (85th or 95th percentile) and peer-blaming in the Fall. The positive association (β = .08 and β = .15, respectively) between perceived rejection and later aggression could reflect either an increase in aggression among boys who were high on perceived rejection in the Fall (the hypothesis), or it could instead reflect a maintenance of aggression that is “positive” only in the sense that it contrasts with a reduction in aggression exhibited by boys who were low on perceived rejection in the Fall. To better gauge the meaning of the conditional betas obtained during probing of the interaction, average raw change scores were calculated for each of the three “groups” just defined.2 These further analyses suggested that the obtained betas did reflect absolute changes in aggression over time within this sample. One such analysis, for example, considered the group of boys who were aggressive (i.e., at or above the 85th percentile) and peer-blaming (i.e., at or above an attribution score of 3.0) at the Fall assessment. Of these boys, those who were high on perceived rejection at Time 1 (i.e., at or above 75 [1.5 SD above the mean]) received aggression scores that increased from Time 1 (M = 0.166) to Time 2 (M = 0.186). Moreover, this increase was evident among both the sociometrically rejected and nonrejected boys in this group. DISCUSSION Results of this study support the position that children’s perceptions of their own social acceptance can influence the degree to which they engage in physical ag2 For the entire sample, approximately 44% of the children (197 girls; 219 boys) received a greater aggression score at Time 2 (Spring) than at Time 1 (Fall). Approximately 47% (216 girls; 226 boys) received a lower aggression score at Time 2 than at Time 1. Aggression scores did not change for approximately 9% of the children (58 girls; 25 boys). Recall that participants were not limited in the number of nominations they could make, so the changes in aggression scores do not merely reflect a redistribution of a fixed number of nominations. Guerra, Asher, and DeRosier gression, even after controlling for children’s “actual” social acceptance by their peers. The nature of this influence, however, seems to depend upon several other participant characteristics. For boys, findings were consistent with the hypothesized three-way interaction. Perceived rejection was prospectively related to changes in physical aggression only for boys already exhibiting high levels of aggression, and the direction of those changes varied as a function of boys’ attributions for social failure experiences. The perception of rejection predicted decreased aggression among aggressive boys who tended to blame themselves for social failure experiences, whereas the perception of rejection predicted increased aggression among aggressive boys who tended to blame peers for social failure. The decrease in aggression among self-blaming aggressive boys is consistent with Sandstrom and Coie’s (1999) hypothesis that the perception of rejection, coupled with some acceptance of personal responsibility for social failure experiences, might motivate a child to make efforts to decrease the problematic behavior(s) thought to be “causing” the rejection. Although this does seem a rather plausible account of the reduction in physical aggression among these boys, at least one alternative hypothesis also merits consideration. As discussed above, prior theory and research suggests that children who both perceive themselves to be rejected and blame themselves for social failures may be at particular risk for developing internalizing problems (Asher et al., 1990; Goetz & Dweck, 1980; Panak & Garber, 1992). It seems possible that the decrease in aggression evidenced by this group of aggressive boys may reflect not an adaptive, motivated choice, but rather a serendipitous “side effect” of internalizing symptoms such as social withdrawal or a diminished energy level. Further research is needed to determine which, if either, of these mediation hypotheses is viable. Identification of the mechanism(s) mediating the association between perceived rejection and increases in aggression among aggressive boys also awaits future empirical investigation. In this study, it was hypothesized that perceived rejection, coupled with the externalization of blame for social failure, would precipitate increases in children’s anger, negative expectations, and attributional and perceptual biases during peer interactions, increases that would, in turn, be associated with a heightened propensity to behave aggressively. Of course, it may be that other processes either partially or fully account for this link between perceived rejection and increased aggression. Consider, for example, how children’s selection of friends may play a mediating role. Most children who believe themselves to be generally disliked will, nevertheless, continue to seek a sense of “belongingness” within Perceived Rejection and Physical Aggression the peer culture (see Baumeister & Leary, 1995). Because physical aggression is not uncommon among boys in elementary school (Coie & Dodge, 1998), an aggressive boy who perceives that he is rejected by the mainstream peer culture stands a good chance of finding other physically aggressive boys among his grade mates who share a similar predicament, boys who may manifest a more accepting attitude towards him. In this way, then, the perception of mainstream rejection may motivate a physically aggressive boy to spend more time with other aggressive children. Indeed, this is consistent with evidence that the friends of physically aggressive boys tend, themselves, to be rather aggressive (Bagwell, Coie, Terry, & Lochman, 2000; Cairns, Cairns, Neckerman, Gest, & Gariepy, 1988). Once involved in such friendships, a child’s aggressive behavior is likely to further consolidate, as he regularly interacts with peers who both model and reinforce aggression (Bagwell, Andreassi, & Keeley, 2001; Dishion, Patterson, & Griesler, 1994). Results obtained for girls in this study contrasted sharply with the results obtained for boys. Perceived rejection bore no significant relation to later physical aggression among aggressive girls who tended to blame themselves for social failure experiences, while perceived rejection predicted a decrease in physical aggression among aggressive girls who tended to blame their peers for social failure experiences. Perhaps the latter result may be at least partially understood by again considering friendship selection processes. During elementary school, physical aggression is much less common among girls than it is among boys (Coie & Dodge, 1998). It is therefore unlikely that a physically aggressive girl who perceives that she is rejected will find many other physically aggressive girls with whom to associate. Cross-gender friendships are relatively rare among children this age (see Maccoby, 1990, 1998), so it is also improbable that a physically aggressive girl will establish many ties to her male counterparts. The opportunity for peer modeling and reinforcement of physical aggression should, then, be much reduced. This hypothetical girl, believing that peers are generally rejecting and hostile (i.e., peer-blaming for social failure experiences), may wind up withdrawing from most, if not all, of her peers. In such a case, the reduction in peer-assessed physical aggression would simply index a more general reduction in social interaction. Research is needed to test the validity of this interpretation. Future research might also examine whether and how the perception of rejection can influence forms of aggressive behavior that are not manifest physically. Consider, for example, relational or social aggression, behavior enacted with intent to inflict harm upon another person’s relationships or social reputation (e.g., spreading false rumors 561 about a person; see Crick, 1997; Underwood, 2003). Relational aggression occurs far more frequently than does physical aggression among girls (Crick, 1997; Crick & Grotpeter, 1995). If the perception of rejection influences friendship selection in the manner described above, the relation of perceived rejection to subsequent relational aggression may be quite different from the relation between perceived rejection and subsequent physical aggression. Girls who engage in high levels of relational aggression should be able to find other relationally aggressive girls to befriend if and when they perceive that those in the peer “mainstream” are rejecting of them. As such, the perception of rejection might lead to increased relational aggression among already aggressive girls who tend to blame others for their social misfortunes. Finally, we would offer a cautionary note regarding the self-report instruments used in this study. Although we have argued that the measure of perceived rejection employed herein is an improvement over existing multiitem scales that confound perceived acceptance/rejection with other social self-perceptions (e.g., general social competence; self-concept; friendship), we are aware that single-item self-report measures have psychometric drawbacks. It may be prudent to develop alternative measurement strategies that capitalize on the psychometric benefits of multi-item scales while still retaining a narrow focus on the construct of perceived rejection. Likewise, psychometric considerations would argue for expanding upon the two-item attribution measure used in this study. These suggestions for future empirical investigation, whether pertaining to mediation pathways, gender differences, relational aggression, or construct measurement, highlight limitations of the research presented here. Even without further explication, however, these data speak effectively to the import of considering ways in which children’s self-perceptions of social status might influence behavior independent of their actual level of acceptance among peers. Whereas previous research has demonstrated a unique effect of social self-perceptions upon internalizing emotional experience (Kistner et al., 1999; Panak & Garber, 1992), the present results constitute the first prospective evidence that self-perceptions of social rejection can uniquely influence externalizing behavior. ACKNOWLEDGMENTS This research was supported, in part, by a grant to the third author from the National Institute of Mental Health (MH54227-01A1). The authors thank the staff and students of the Wake County Public School System for their 562 cooperation. They also thank David Gardner for his role in data management. REFERENCES Aiken, L. S., & West, S. G. (1992). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Asher, S. R., Parkhurst, J. T., Hymel, S., & Williams, G. A. (1990). Peer rejection and loneliness in childhood. In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 253–273). New York: Cambridge University Press. Bagwell, C. L., Andreassi, C. L., & Keeley, S. J. (2001, April). The friendships of aggressive boys: Relationship quality and conflict management. Poster presented at the biennial meeting of the Society for Research in Child Development, Minneapolis, MN. Bagwell, C. L., Coie, J. D., Terry, R. A., & Lochman, J. E. (2000). Peer clique participation and social status in preadolescence. MerrillPalmer Quarterly, 46, 280–305. Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. Bierman, K. L., & Wargo, J. B. (1995). Predicting the longitudinal course associated with aggressive–rejected, aggressive (nonrejected), and rejected (nonaggressive) status. Development and Psychopathology, 7, 669–682. Boivin, M., & Begin, G. (1989). Peer status and self-perception among early elementary school children: The case of the rejected children. Child Development, 60, 591–596. Boivin, M., & Hymel, S. (1997). Peer experiences and social selfperceptions: A sequential model. Developmental Psychology, 33, 135–145. Boivin, M., Hymel, S., & Bukowski, W. M. (1995). The roles of social withdrawal, peer rejection, and victimization by peers in predicting loneliness and depressed mood in childhood. Development and Psychopathology, 7, 765–785. Boivin, M., Poulin, F., & Vitaro, F. (1994). Depressed mood and peer rejection in childhood. Development and Psychopathology, 6, 483– 498. Burks, V. S., Laird, R. D., Dodge, K. A., Pettit, G. S., & Bates, J. E. (1999). Knowledge structures, social information processing, and children’s aggressive behavior. Social Development, 8, 220– 236. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Gest, S. D., & Gariepy, J. (1988). Social networks and aggressive behavior: Peer support or peer rejection? Developmental Psychology, 24, 815–823. Cillessen, A. H. N., & Bellmore, A. D. (1999). Accuracy of social selfperceptions and peer competence in middle childhood. MerrillPalmer Quarterly, 45, 650–676. Coie, J. D. (1990). Toward a theory of peer rejection. In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 365–401). New York: Cambridge University Press. Coie, J. D., & Dodge, K. A. (1983). Continuities and changes in children’s social status: A five-year longitudinal study. Merrill-Palmer Quarterly, 29, 261–282. Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.), Handbook of child psychology (pp. 779–862). New York: Wiley. Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982). Dimensions and types of social status: A cross-age perspective. Developmental Psychology, 18, 557–570. Coie, J. D., Dodge, K. A., & Kupersmidt, J. B. (1990). Peer group behavior and social status. In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 17–59). New York: Cambridge University Press. Guerra, Asher, and DeRosier Coie, J., Terry, R., Lenox, K., Lochman, J., & Hyman, C. (1995). Childhood peer rejection and aggression as predictors of stable patterns of adolescent disorder. Development and Psychopathology, 7, 697– 713. Crick, N. R. (1997). Engagement in gender normative versus nonnormative forms of aggression: Links to social–psychological adjustment. Developmental Psychology, 33, 610–617. Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children’s social adjustment. Psychological Bulletin, 115, 74–101. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social–psychological adjustment. Child Development, 66, 710– 722. Crick, N. R., & Ladd, G. W. (1993). Children’s perceptions of their peer experiences: Attributions, loneliness, social anxiety, and social avoidance. Developmental Psychology, 29, 244–254. Crick, N. R., & Werner, N. E. (1998). Response decision processes in relational and overt aggression. Child Development, 69, 1630–1639. David, C. F., & Kistner, J. A. (2000). Do positive self-perceptions have a “dark side”? Examination of the link between perceptual bias and aggression. Journal of Abnormal Child Psychology, 28, 327– 337. Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1998). Multiple risk factors in the development of externalizing behavior problems: Group and individual differences. Development and Psychopathology, 10, 469–493. DeRosier, M. E. (2001). Social interactions survey. Unpublished manuscript, University of North Carolina at Chapel Hill, Frank Porter Graham Child Development Center. DeRosier, M. E., & Parkhurst, J. T. (2004). Linking social selfperceptions with behavioral and emotional adjustment at school. Manuscript in preparation. Dishion, T. J., Patterson, G. R., & Griesler, P. C. (1994). Peer adaptations in the development of antisocial behavior: A confluence model. In L. R. Huesmann (Ed.), Aggressive behavior: Current perspectives (pp. 61–95). New York: Plenum Press. Dodge, K. A., Lansford, J. E., Burks, V. S., Bates, J. E., Pettit, G. S., Fontaine, R., et al. (2003). Peer rejection and social informationprocessing factors in the development of aggressive behavior problems in children. Child Development, 74, 374–393. Dodge, K. A., Pettit, G. S., McClaskey, C. L., & Brown, M. M. (1986). Social competence in children. Monographs of the Society for Research in Child Development, 51(2, Serial No. 213). Downey, G., Lebolt, A., Rincon, C., & Freitas, A. L. (1998). Rejection sensitivity and children’s interpersonal difficulties. Child Development, 69, 1074–1091. Egan, S. K., Monson, T. C., & Perry, D. G. (1998). Social–cognitive influences on change in aggression over time. Developmental Psychology, 34, 996–1006. Erdley, C. A., & Asher, S. R. (1996). Children’s social goals and selfefficacy perceptions as influences on their responses to ambiguous provocation. Child Development, 67, 1329–1344. Erdley, C. A., & Asher, S. R. (1998). Linkages between children’s beliefs about the legitimacy of aggression and their behavior. Social Development, 7, 321–339. Feldman, E., & Dodge, K. A. (1987). Social information processing and sociometric status: Sex, age, and situational effects. Journal of Abnormal Child Psychology, 15, 211–227. Goetz, T. Z., & Dweck, C. S. (1980). Learned helplessness in social situations. Journal of Personality and Social Psychology, 39, 246– 255. Harter, S. (1982). The perceived competence scale for children. Child Development, 53, 87–97. Harter, S. (1985). Manual for the Self-Perception Profile for Children. Denver, CO: University of Denver. Hecht, D. B., Inderbitzen, H. M., & Bukowski, A. L. (1998). The relationship between peer status and depressive symptoms in children and adolescents. Journal of Abnormal Child Psychology, 26, 153– 160. Perceived Rejection and Physical Aggression Huesmann, L. R., & Guerra, N. G. (1997). Children’s normative beliefs about aggression and aggressive behavior. Journal of Personality and Social Psychology, 72, 408–419. Hughes, J. N., Cavell, T. A., & Grossman, P. B. (1997). A positive view of self: Risk or protection for aggressive children? Development and Psychopathology, 9, 75–94. Hughes, J. N., Cavell, T. A., & Prasad-Gaur, A. (2001). A positive view of peer acceptance in aggressive youth risk for future peer acceptance. Journal of School Psychology, 39, 239–252. Hymel, S., Bowker, A., & Woody, E. (1993). Aggressive versus withdrawn unpopular children: Variations in peer and self-perceptions in multiple domains. Child Development, 64, 879–896. Hymel, S., & Franke, S. (1985). Children’s peer relations: Assessing selfperceptions. In B. H. Schneider, K. H. Rubin, & J. E. Ledingham (Eds.), Children’s peer relations: Issues in assessment and intervention (pp. 75–91). New York: Springer-Verlag. Hymel, S., Rubin, K. H., Rowden, L., & LeMare, L. (1990). Children’s peer relationships: Longitudinal prediction of internalizing and externalizing problems from middle to late childhood. Child Development, 61, 2004–2021. Kistner, J., Balthazor, M., Risi, S., & Burton, C. (1999). Predicting dysphoria in adolescence from actual and perceived peer acceptance in childhood. Journal of Clinical Child Psychology, 28, 94–104. Kupersmidt, J. B., & Coie, J. D. (1990). Preadolescent peer status, aggression, and school adjustment as predictors of externalizing problems in adolescence. Child Development, 61, 1350–1362. Kupersmidt, J. B., & Patterson, C. J. (1991). Childhood peer rejection, aggression, withdrawal, and perceived competence as predictors of self-reported behavior problems in preadolescence. Journal of Abnormal Child Psychology, 19, 427–449. Lochman, J. E. (1987). Self and peer perceptions and attributional biases of aggressive and nonaggressive boys in dyadic interactions. Journal of Consulting and Clinical Psychology, 55, 404–410. Lochman, J. E., & Dodge, K. A. (1998). Distorted perceptions in dyadic interactions of aggressive and nonaggressive boys: Effects of prior expectations, context, and boys’ age. Development and Psychopathology, 10, 495–512. Maccoby, E. E. (1990). Gender and relationships: A developmental account. American Psychologist, 45, 513–520. Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Belknap Press/Harvard University Press. MacKinnon-Lewis, C., Rabiner, D., & Starnes, R. (1999). Predicting boys’ social acceptance and aggression: The role of mother–child interactions and boys’ beliefs about peers. Developmental Psychology, 35, 632–639. McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interactions and moderator effects. Psychological Bulletin, 114, 376–389. Ollendick, T. H., Weist, M. D., Borden, M. C., & Greene, R. W. (1992). Sociometric status and academic, behavioral, and psychological adjustment: A five-year longitudinal study. Journal of Consulting and Clinical Psychology, 60, 80–87. 563 Panak, W. F., & Garber, J. (1992). Role of aggression, rejection, and attributions in the prediction of depression in children. Development and Psychopathology, 4, 145–165. Parkhurst, J. T., & Asher, S. R. (1992). Peer rejection in middle school: Subgroup differences in behavior, loneliness, and interpersonal concerns. Developmental Psychology, 28, 231–241. Patterson, C. J., Kupersmidt, J. B., & Griesler, P. C. (1990). Children’s perceptions of self and of relationships with others as a function of sociometric status. Child Development, 61, 1335–1349. Perry, D. G., Perry, L. C., & Rasmussen, P. (1986). Cognitive social learning mediators of aggression. Child Development, 57, 700–711. Quiggle, N. L., Garber, J., Panak, W. F., & Dodge, K. A. (1992). Social information processing in aggressive and depressed children. Child Development, 63, 1305–1320. Renshaw, P. D., & Brown, P. J. (1993). Loneliness in middle childhood: Concurrent and longitudinal predictors. Child Development, 64, 1271–1284. Rubin, K. H., Hymel, S., Mills, R. S. L., & Rose-Krasnor, L. (1991). Conceptualizing different developmental pathways to and from social isolation in childhood. In D. Cicchetti & S. L. Toth (Eds.), Internalizing and externalizing expressions of dysfunction (pp. 91–122). Hillsdale, NJ: Erlbaum. Sandstrom, M. J., & Coie, J. D. (1999). A developmental perspective on peer rejection: Mechanisms of stability and change. Child Development, 70, 955–966. Sandstrom, M. J., & Zakriski, A. L. (2004). Understanding the experience of peer rejection. In J. B. Kupersmidt & K. A. Dodge (Eds.), Children’s peer relations: From development to intervention (pp. 101– 118). Washington, DC: American Psychological Association. Sobol, M. P., & Earn, B. M. (1985). Assessment of children’s attributions for social experiences: Implications for social skills training. In B. H. Schneider, K. H. Rubin, & J. E. Ledingham (Eds.), Children’s peer relations: Issues in assessment and intervention (pp. 93–110). New York: Springer-Verlag. Terry, R. (2000). Recent advances in measurement theory and the use of sociometric techniques. In A. H. N. Cillessen & W. M. Bukowski (Eds.), Recent advances in the measurement of acceptance and rejection in the peer system (pp. 27–53). San Francisco: JosseyBass. Underwood, M. K. (2003). Social aggression among girls. New York: Guilford. Weiner, B. (1990). Attribution in personality psychology. In L. A. Pervin (Ed.), Handbook of personality: Theory and research (pp. 465– 485). New York: Guilford Press. Weiner, B., & Graham, S. (1984). An attributional approach to emotional development. In C. E. Izard, J. Kagan, & R. B. Zajonc (Eds.), Emotions, cognition, and behavior (pp. 167–191). New York: Cambridge University Press. Zakriski, A. L., & Coie, J. D. (1996). A comparison of aggressive– rejected and nonaggressive–rejected children’s interpretations of self-directed and other-directed rejection. Child Development, 67, 1048–1070.
© Copyright 2026 Paperzz