Perceived Social Support among Bullies, Victims, and Bully-Victims ...

J Youth Adolescence (2007) 36:984–994 DOI 10.1007/s10964-006-9153-3

ORIGINAL PAPER

Perceived Social Support among Bullies, Victims, and Bully-Victims Melissa K. Holt · Dorothy L. Espelage

Received: 4 August 2006 / Accepted: 1 November 2006 / Published online: 13 December 2006 C Springer Science+Business Media, LLC 2006

Abstract Research indicates that social support plays a protective role among adolescents, but little research has explicitly evaluated its function among youth involved in bullying. Accordingly, this study examined relations among social support, bully/victim status, and psychological distress in a sample of 784 ethnically diverse youth. We assessed differences in perceived social support across bully/victim subtypes, and evaluated peer and maternal social support as protective factors among victims, bullies, and bully-victims. Youth were classified as uninvolved (61.6%), as bullies (14.3%), as victims (12.5%), and as bully-victims (11.6%). Uninvolved youth reported the most peer and maternal social support and the least anxiety/depression. Multivariate analyMelissa K. Holt is a Research Assistant Professor at the University of New Hampshire Crimes Against Children Research Center and Family Research Laboratory. She earned her Ph.D. from the University of Illinois at Urbana-Champaign in Counseling Psychology. Her major research interests include bullying, multiple victimization among adolescents, and contextual influences on victimization. Dorothy L. Espelage is an Associate Professor of Counseling Psychology in the Department of Educational Psychology at the University of Illinois, Urbana-Champaign. She earned her Ph.D. from Indiana University in Counseling Psychology. Her major research interests include youth aggression, victimization, childhood sexual abuse, and eating disorders. M. K. Holt () Research Assistant Professor, Family Research Laboratory and Crimes Against Children Research Center, University of New Hampshire, 10 West Edge Dr., Suite 106, Durham, NH 03824 e-mail: [email protected] D. L. Espelage Associate Professor, Department of Educational Psychology, University of Illinois at Urbana-Champaign, 220A Education, Champaign, IL 61820 e-mail: [email protected] Springer

ses revealed that there was a significant interaction between bully/victim groups and peer social support. Specifically, bullies, victims, and bully-victims who reported moderate peer social support also indicated the least anxiety/depression. Results highlight the importance of encouraging youth to develop and effectively use peer support networks as part of bullying intervention programs. Keywords Bullying . Victimization . Social support . Adolescence Approximately 30% of American students are involved in bullying as bullies, victims, or bully-victims (Nansel et al., 2001) and often suffer adverse consequences as a result of this involvement. Specifically, bullying victimization is associated with depression (Denny et al., 2004; Kaltiala-Heino et al., 2001; Neary and Joseph, 1994), loneliness (Hawker and Boulton, 2000), and low self-esteem (Olweus, 1995). Bullies also experience psychological distress, including depression (Seals and Young, 2003) and suicidal ideation (Rigby and Slee, 1999). Finally, considerable research has documented that bully-victims are a particularly high-risk group. For instance, bully-victims demonstrate more externalizing problem behaviors (Nansel et al., 2003; Nansel et al., 2001), more physical symptoms of anxiety (Swearer et al., 2001), and have a greater probability of being referred for psychiatric consultation than their peers (Kumpulainen et al., 1998). In recent years researchers have started to move beyond simply relating bullying involvement to psychological functioning and have attempted to address heterogeneity in adjustment (Kochenderfer-Ladd and Skinner, 2002; Schwartz et al., 2000). This movement toward identifying moderating variables follows the broader literature on

J Youth Adolescence (2007) 36:984–994

adolescence that has sought to identify factors that buffer youth against deleterious effects of stress. Through identifying areas that promote positive youth psychosocial functioning in the face of adversity, more effective intervention and prevention programs can be designed. To date researchers have evaluated key factors including perceptions of the environment (Jessor et al., 1995), gender (Bond et al., 2001), coping styles (Grant et al., 2000), and religiosity (Smith, 2003). Further, one of the most widely acknowledged protective factors among multiple age groups, including among adolescents, is perceived social support (Cohen and Wills, 1985). However, little research has evaluated social support in the context of bullying involvement, or examined the extent to which social support serves as a buffer between bullying involvement and psychological distress. As such, this study builds on existing research that broadly considers the function of social support among adolescents and applies it to our understanding of bullying perpetration and victimization dynamics. Results will inform anti-bullying programs. If social support indeed minimizes the distress associated with experiencing bullying victimization, information on accessing and effectively using social support networks should become an integral component of bullying prevention and intervention programs. Perceived social support Research indicates that youth can derive social support from a number of sources for beneficial effects to result. The primary source of social support for youth often varies as a function of age. In childhood youth tend to seek support primarily from parents but as youth transition into adolescence peer support becomes more salient (Furman and Buhrmester, 1992). Also, older adolescents typically report less support from all sources than younger adolescents (Malecki and Elliott, 1999). Given that youth report that they receive different social provisions from parents than from peers (Furman and Buhrmester, 1985), however, both sources of support are important to consider. Finally, some research has addressed gender differences in social support. Broadly, girls tend to report more perceived social support than boys (Furman, 1996; Malecki and Elliott, 1999), and girls’ friendships tend to be more intimate (Parker and Asher, 1993). Social support as a buffer Investigators have often studied social support in the context of how it affects youth who have experienced adverse events, and have examined both familial and peer social support in this capacity. With respect to familial support, adolescents exposed to stress who perceived greater maternal social support reported fewer internalizing and externalizing behaviors (Grant et al., 2000). In addition, among African American

985

males maternal support buffered youth victimized in dating relationships against psychological distress (Holt and Espelage, 2005). Finally, among sexually abused adolescents, those individuals who described high levels of support from one or both parents had less negative psychological outcomes than those individuals without such high levels of support (Luster and Small, 1997). There have been mixed findings about whether the buffering effects of familial support vary by gender. For instance, an investigation showed that girls, but not boys, exposed to family violence were better able to maintain their selfworth when they reported high levels of support from their guardians (Kolbo, 1996). Conversely, in a study of African American youth parental support had a positive influence on educational outcomes above and beyond the impact of sex and adverse conditions (Connell et al., 1994). Peer support has also been shown to promote psychological well-being despite adversity (McCreary et al., 1996). For example, youth who felt supported by their close friends experienced fewer negative effects of marital discord than youth with less support (Wasserstein and La Greca, 1996). Peer social support also appears to play a salient role with respect to self-esteem. Specifically, one study showed that greater peer and teacher support positively influenced adolescents’ self-esteem over time, whereas maternal and paternal social support did not significantly affect youth self-esteem (Colarossi and Eccles, 2003). Perceived social support and its relation to bullying Despite strong evidence linking perceived social support to adjustment among youth, the role of perceived social support among youth involved with bullying in particular is not yet clear. Thus far, researchers have evaluated social support in three primary capacities in studies on bullying — perceptions of social support among bully/victim subtypes; the extent to which victims seek social support as a coping mechanism to being victimized; and social support as a protective factor — yet our knowledge in each of these arenas remains sparse. Perceptions of social support and bully/victim subtypes Few studies have examined perceptions of social support by bully/victim subtype. One study of 5th through 12th graders found that victims reported less teacher and peer support than their peers (Furlong et al., 1995). Similarly, in a recent study of predominately Hispanic middle school students, those classified as victims and bully-victims indicated receiving substantially less social support from classmates than those students in the comparison group (Demaray and Malecki, 2003). Additionally, in this same study researchers investigated differences in perceptions of the importance of Springer

986

social support across bully/victim groups. Results indicated that victims and bully-victims rated total social support as more important relative to the bully and comparison groups. As such, although victims and bully-victims reported less perceived social support than their peers, they gave the most weight to its importance. Given the clearly established association between levels of social support and positive psychological outcomes, the implications of these findings are troublesome; youth at greatest risk for negative outcomes given their status as victims and bully-victims value social support more than their peers yet report less of it. Social support as a coping mechanism Social support represents one potential coping mechanism students can use to deal with bullying. In recognition of this, some schools have developed formal peer social support systems for victims to access (Naylor et al., 2001). However, it appears that the victims who likely need social support the most have the greatest difficulty obtaining it. Specifically, research has shown that chronic victims have more difficulty utilizing social support as a coping mechanism than victims whose experiences with bullying are more transitory in nature. For example, Hunter and Boyle (Hunter and Boyle, 2004) found that targets of bullying who had experienced bullying for more than four weeks indicated using less social support than targets who had experienced bullying for less than four weeks. To address such findings, some researchers (Smith and Talamelli, 2001) have suggested training children in friendship skills or teaching them how to engage in broader help-seeking behaviors. This could assist victims in both accessing support in their environment and utilizing existing support. Protective capacity of social support Only a handful of studies have evaluated the protective capacity of social support as it relates to bullying. In a study by Hodges and colleagues (Hodges et al., 1999), peer victimization was linked to internalizing behavior problems only for children without sufficient social support, specifically that which could be derived from a mutual best friend. The degree of perceived protection from the friend was relevant as well, whereby only children who perceived a high degree of protection from their friend experienced the positive buffering effect of social support. In another study, among ethnically diverse adolescents relational victimization (e.g., threatening to withdraw friendship if adolescent does not go along with aggressor) was related to externalizing behaviors only for adolescents with low perceived peer social support (Prinstein et al., 2001). In contrast to the two studies described above, although low social support and peer victimization independently predicted poor mental health in a Springer


sample of 845 school-aged children, neither peer nor familial social support moderated the association between victimization and well-being (Rigby, 2000). As such, more research is needed in this area to clarify the extent to which social support plays a buffering role among youth involved in bullying. Distinguishing among peer group status, social support, and friendship It is important to note that although the research above documents differences across bully/victim groups on perceptions and use of social support, and in particular highlights associations between victimization and social support, there are important distinctions among peer group status, social support, and friendship. More specifically, Parker and Asher (1993) note that it is important to consider not only friendship quality (which includes social support), but also the extent to which youth have mutual friendships. With respect to victims in particular, it is not necessarily the case that youth who are victimized do not have friends, but they might encounter more difficulties negotiating their friendships than non-victims (Parker and Asher, 1993). For instance, victims often have limited social problem solving skills, and are less able to negotiate conflict that arises in relationships (Champion et al., 2003). Or, their mutual friendships might be characterized by high levels of intimacy or selfdisclosure, which in turn have been shown to be associated with relational aggression within the friendship and jealousy (Crick and Grotpeter, 1996; Crick and Nelson, 2002; Parker et al., 2005). Accordingly, although victims might have friends, the quality of their relationships might be lower than the quality of relationships experienced by non-victims. Further, because victims tend to be friends with other victims, they often do not have friendships with youth who might be able to protect them from bullies (Champion et al., 2003; Pellegrini et al., 1999). This is supported by research showing that although having friends in the victim group did not protect youth against victimization, having friends in the bullying and uninvolved groups did (Pellegrini et al., 1999). Hypotheses Extant research has documented the link between social support and well-being among adolescents and has highlighted the protective nature of social support against stressors. However, little research has explicitly evaluated the role of social support among bullies, victims, and bully-victims. Thus, the current investigation extends previous work through its comprehensive evaluation of relations among perceived social support, bullying involvement, and psychological functioning. Specifically, we hypothesized that uninvolved students would report the highest amounts of perceived social support


and the least anxiety/depression. Further, we expected that perceived social support would moderate the association between bullying involvement and psychological functioning. In particular, we hypothesized that among victims, bullyvictims, and bullies, those who also perceived high levels of social support would experience less anxiety/depression than their peers who reported low levels of social support.

Method Participants Participants were 784 youth, including 424 (54%) middle school and 360 (46%) high school students. The middle school was located in a suburb of a large Midwestern city and the high school was located in a small Midwestern city. All students from these schools were invited to participate, and 100% of students whose parents agreed to their participation (see below) elected to complete the survey. There were 371 (47%) males and 413 (53%) females. With respect to race, respondents identified themselves as White, Non-Hispanic (n = 415; 52.9%), African American (n = 269; 34.3%), Hispanic (n = 45; 5.7%), Asian (n = 10; 1.3%), Native American (n = 9; 1.1%), and “Other” (n = 36; 4.6%). Of these students, 222 were seventh graders (28.3%), 202 were eighth graders (25.8%), 125 were ninth graders (15.9%), 25 were tenth graders (3.2%), 120 were eleventh graders (15.3%), and 90 were twelfth graders (11.5%). Few tenth grade students participated in the study because tenth graders in the study site did not typically enroll in physical education classes, the setting in which surveys were administered. The mean age for the sample was 14.51 (SD = 1.97).1 With respect to socio-economic status, although we did not collect this information directly from students, school-district records indicate that approximately 42% of middle school students and 43% of high school students were classified as lowincome.

Measures Each participant first completed a demographic questionnaire that included questions about his/her sex, age, grade, and race/ethnicity.

987

Self-reported bullying behavior The nine-item University of Illinois Bully Scale (UIBS; Espelage and Holt, 2001) was used to assess bullying behavior including teasing, social exclusion, name-calling, and rumor spreading. Researchers developed this scale based on interviews with middle school students, a review of the research literature on existing bullying measures, and extensive factor analytic investigations (Espelage et al., 2000; Espelage et al., 2003). Students are asked to indicate how often in the past 30 days they have engaged in each behavior (e.g., “I teased other students” and “I upset other students for the fun of it”). Response options include “Never,” “1 or 2 times,” “3 or 4 times,” “5 or 6 times,” and “7 or more times.” These response options allow for the assessment of the persistence of the bullying. Higher scores indicate more self-reported bullying behaviors. The construct validity of this scale has been supported via exploratory and confirmatory factor analysis (Espelage and Holt, 2001). Further, the Bullying Scale was found to be moderately correlated (r = .65) with the Youth Self-Report Aggression Scale (Achenbach, 1991), suggesting convergent validity. Also, this scale was not significantly correlated with the Illinois Victimization Scale (r = .12), providing evidence of discriminant validity (Espelage et al., 2003). Finally, the UIBS was found to converge with peer nomination data (Espelage et al., 2003). A Cronbach alpha coefficient of .90 was found for the current sample. Self-reported victimization Victimization from peers was assessed using the University of Illinois Victimization Scale (UIVS; Espelage and Holt, 2001). Students are asked how often the following things have happened to them in the past 30 days (i.e., “Other students called me names,” “Other students made fun of me,” “Other students picked on me,” and “I got hit and pushed by other students”). Response options include “Never,” “1 or 2 times,” “3 or 4 times,” “5 or 6 times,” and “7 or more times.” Higher scores indicate more self-reported victimization. As noted above, this scale was not significantly correlated with the Illinois Bullying scale (r = .12), providing evidence of discriminant validity (Espelage et al., 2003). This scale was also found to converge with peer nomination data (Espelage et al., 2003). A Cronbach alpha coefficient of .86 was found for current study. Psychological functioning

For 7th graders ages ranged from 12–14 (M = 12.49, SD = .52); for 8th graders ages ranged from 12–15 (M = 13.45, SD = .54); for 9th graders ages ranged from 14–16 (M = 14.89, SD = .51); for 10th graders ages ranged from 15–18 (M = 16.28, SD = .61); for 11th graders ages ranged from 16–19 (M = 16.80, SD = .53); and for 12th graders ages ranged from 17–19 (M = 17.81, SD = .48). 1

The anxiety/depression scale from the Youth Self-Report (YSR; Achenbach, 1991) was used to assess psychological functioning. This scale consists of 16 self-report items and students are asked to indicate the degree to which particular Springer

988

statements apply to them (e.g., “I feel lonely,” “I am nervous or tense”). Response options are “Not true,” “Somewhat or sometimes true,” and “Very true or Often true.” Numerous studies have supported this instrument’s reliability and validity (Achenbach, 1991; Bennett and Bates, 1995; Visser et al., 1999). The coefficient alpha coefficient was .90 for this study.


Survey administration

The Support/Cohesion Microsystem Scale (Seidman et al., 1995) is a 21-item inventory designed to measure the degree to which adolescents perceive individuals in their lives support them. For the purpose of the current investigation participants were asked to rate support in relationships with their mothers and close friends. Three components of support — emotional, instrumental, and satisfaction — are evaluated and summed to create a total score for each relationship. Specifically, individuals rate how helpful the person is “when I have a personal problem” (emotional support) and “when I need money and things” (instrumental support) and also indicate how much “I have fun with this person” (satisfaction). Response options range from 1 (not at all) through 3 (a great deal). Coefficient alphas were: .81 (middle school) and .86 (high school) for the mother subscale, and .66 (middle school) and .75 (high school) for the peer subscale.

Six trained research assistants, the primary researcher, and a faculty member collected data. At least two of these individuals administered surveys to each physical education class, which ranged in size from 25 to 50. Students completed the surveys on bleachers. Students were first informed about the general nature of the investigation. Surveys were anonymous. Next, researchers made certain that students were sitting far enough from one another to ensure confidentiality. Students were then given survey packets and asked to answer all questions honestly. Researchers were available to answer questions that emerged once students began responding to survey items. When students had completed the surveys they were given the opportunity to have their data removed from analyses if they had not carefully considered each question. Each participant was also provided with a list of phone numbers to call (e.g., community counseling agencies) should they experience an emotional reaction to the questionnaires, and were told about school-based resources. Further, at every administration a doctoral-level psychologist was in the room that was appropriately trained to provide an immediate response to any distressed participants and direct him/her to appropriate resources. Last, a raffle was held in each group in which one student won a $10 gift certificate to a local mall. On average it took students approximately 45 minutes to complete the survey.

Procedure

Results

Parental consent

Preliminary analyses

We obtained permission from the University of Illinois at Urbana-Champaign Institutional Review Board to use passive parental consent in this investigation. Passive parental consent was used to maximize participation. Parents of all students were sent letters informing them about the purpose of the study. They were asked to sign and return the form only if they were unwilling to have their child participate in the investigation. Three middle school parents and zero high school parents returned this form. In addition to passive parental consent, students were asked to assent to participate in the study through a form included in the questionnaire packet. As noted above, 100% of students whose parents tacitly permitted their participation using the passive consent procedure elected to complete the survey. Accordingly, the overall participation rate was over 99%. At the beginning of each survey administration teachers removed students from the room if they were not allowed to participate, and researchers also reminded students that they should not complete the survey if their parents had returned a form.

We first classified all students in the sample as bullies, victims, bully-victims, or uninvolved so that subsequently we could examine group differences on social support and anxiety/depression. In line with previous research (Demaray and Malecki, 2003; Haynie et al., 2001), students were considered “bullies” if their scores were in the top 25th percentile on the bullying measure. This resulted in 112 (14.3%) youth categorized as bullies. The 98 (12.5%) students classified as “victims” had scores in the top 25th percentile on the peer victimization measure. Youth whose scores were in the top 25th percentile on both the bullying and peer victimization measures were considered “bully-victims”; ninety-one (11.6%) participants received this classification. Finally, students in none of the above categories were considered “uninvolved” (N = 483; 61.6%). Table 1 displays mean bullying and peer victimization scores for each bully/victim subtype. Next, we examined whether the composition of bully/victim subtypes differed across sex, grade, and race. Chi-square analyses indicated that there were significant sex differences across groups (χ 2 = 4.43; p < .05,

Support/cohesion microsystem scale

Springer

J Youth Adolescence (2007) 36:984–994 Table 1

989

Mean bullying and peer victimization scores for bullies, victims, bully-victims, and uninvolved youth

Bullying Peer victimization

Uninvolved youth (n = 483) M SD

Bullies (n = 112) M SD

Victims (n = 98) M SD

Bully-victims (n = 91) M SD

ANOVA F (3, 783)

η2

16.94a 5.04c

34.65b 5.46c

18.71a 12.36d

34.13b 13.30d

460.69∗ 656.68∗

0.64 0.72

3.26 1.41

10.33 1.53

3.36 3.22

8.91 3.58

Bullying mean is significantly different from all other groups, p < .01.

a b

Bullying mean is significantly different from uninvolved youth and victims, p < .01.

c

Victimization mean is significantly different from bully-victims and victims, p < .01.

c

Victimization mean is significantly different from all other groups, p < .01.

∗

p < .01.

N = 784). Specifically, males were more represented in the bully group (males = 16.2%; females = 12.6%) and the bully-victim group (males = 16.2%; females = 7.5%), whereas females were more represented in the uninvolved group (males = 55.8%; females = 66.8%). The percentages of males and females in the victim group did not differ significantly (males = 11.9%; females = 13.1%). Chisquare analyses also indicated that bully/victim subtypes differed significantly by race/ethnicity (χ 2 = 24.98; p < .001; N = 784). For these analyses, African-American, Hispanic, Asian, and Native American students were collapsed into one category to achieve an adequate sample size and compared to White students. For the bully and bullyvictim groups, non-White students were more represented than White students (White = 7.5%, 8.7% respectively; Non-White = 22.0%, 14.9% respectively). For the victim group, White children were represented more than nonWhite youth (White = 14.7%; Non-White = 10.0%). Finally, Chi-square analyses yielded significant differences for grade levels across bully/victim groups (χ 2 = 9.59; p < .01; N = 784). With respect to the bully group, approximately 6% of 9th and 12th graders were classified in this category, in contrast to a range from 16 to 20% among students in the other grades. Eleventh and twelfth graders had lower per-

Table 2

centages in the victim group than other students (10% versus a range of 12% to 16%). Eighth graders had the highest percentage in the bully-victim group (19.3%), followed by the 7th graders (12.2%), with lower percentages for 9th through 12th graders (4.8, 4.0, 9.2, and 7.8% respectively). Given that the bully/victim groups differed across sex, race, and grade, these demographic variables were considered as additional independent variables in subsequent analyses. Bullying group comparisons on social support We compared bullies, victims, bully-victims, and uninvolved youth on perceived social support using ANOVAs (Table 2). The dichotomous variables of sex and race (White, nonWhite) were entered as additional independent variables. Given that the major grade differences were found between 7th and 8th graders and the rest of the grades, the grade levels were collapsed into a dichotomous variable (7th & 8th versus 9th–12th). In the first ANOVA we examined differences in peer social support, and there was a significant, but modest, main effect for group. Tukey follow-up tests indicated that uninvolved youth perceived significantly more peer social support than bully-victims or victims. However, bullies and uninvolved youth did not report significantly different

Mean perceived social support and anxiety/depression scores for bullies, victims, bully-victims, and uninvolved youth

Perceived social support Peer Maternal Anxiety/Depression

Uninvolved youth (n = 483) M SD

Bullies (n = 112) M SD

Victims (n = 98) M SD

Bully-victims (n = 91) M SD

Anova F (3, 752)

η2

7.87a 7.47c 4.66e

7.77 7.32 7.41f

7.39b 6.83d 9.35f

7.36b 7.01 9.48f

2.80∗ 5.73∗∗ 29.00∗∗∗

.01 .02 .10

1.40 1.64 5.20

1.57 1.83 7.29

1.71 1.83 7.73

1.47 2.56 6.53

Peer social support mean is significantly different from bully-victims and victims, p < .01.

a b

Peer social support mean is significantly different from uninvolved youth, p < .01.

c

Maternal social support mean is significantly different from Victims, p < .01.

d

Maternal social support mean is significantly different from uninvolved youth, p < .01.

e

Anxiety/depression mean is significantly different from all other groups.

f

Anxiety/depression mean is significantly different from uninvolved youth.

∗

p < .05, ∗∗ p < .01, ∗∗∗ p < .001. Springer

990


amounts of perceived peer social support. A significant main effect was also found for sex, whereby females reported more peer social support (M = 8.20, SD = 1.26) than males (M = 7.22, SD = 1.54). Peer social support also differed by race/ethnicity; White students reported slightly more peer social support (M = 8.30, SD = 1.35) than non-White students (M = 7.40, SD = 1.56). No interactions were significant. The second ANOVA, testing group differences on perceived maternal social support, yielded a significant main effect for group. In this case, uninvolved students reported more maternal social support than victims. Bullies and bully-victims did not, however, report significantly different amounts of maternal social support than uninvolved youth or victims. Main effects for sex and race were not significant. However, a significant interaction between bully/victim groups and race was found. In each group, White students reported less maternal support than nonWhite students, with the exception of bullies. Among bullies, White students reported more maternal support (M = 7.54, SD = 1.52) than non-White students (M = 7.23, SD = 1.94). Of note, these differences were minimal. Finally, a significant main effect was found for grade level; 7th and 8th graders as a group reported slightly more maternal social support (M = 7.57, SD = 1.75) than 9th through 12th graders (M = 7.00, SD = 1.88). Two three-way interactions were significant, but were not interpreted because they involved cell sizes less than ten. No other interactions were significant. Bullying group comparisons on anxiety/depression We next tested whether anxiety/depression scores differed for bullies, victims, bully-victims, and uninvolved youth (Table 2). Results from the ANOVA were statistically significant for bully/victim groups. Post-hoc Tukey comparisons showed that uninvolved youth reported significantly less anxiety/depression than youth in other groups and bullies reported significantly less anxiety/depression than victims or bully-victims. Significant differences were also found Table 3

Social support as a moderator of victimization & depression/anxiety Next, we computed two additional ANOVAs to examine the buffering effects of peer and maternal social support on the relation between bully/victim subtypes and anxiety/depression. Peer and maternal social support variables were converted from continuous to three-level variables (low, moderate, high) using a tertial split. This allowed us to use these variables as independent variables along with the bullyvictim grouping variable. The dependent variable in both models was anxiety/depression. Given the modest effect size associated with sex differences in anxiety/depression levels and literature on the higher rates of depressive symptoms among girls, sex was entered as an additional independent variable. Race and grade were not included in analyses, however, because although they were significantly related to social support, they did not yield large enough effect sizes to warrant including them in the models. Further, including race and grade would have introduced a number of cells with fewer than ten participants. In the first ANOVA we examined peer social support and found a significant bully/victim group by peer social support interaction (Table 3). This suggested that associations between bully/victim groups and anxiety/depression depended on perceptions of peer social support. Specifically, as depicted in Figure 1 students in the uninvolved group had the lowest levels of anxiety/depression, and their levels of anxiety/depression were similar across the three levels of

Mean depression/anxiety scores for bullies, victims, bully-victims, and uninvolved youth across three levels of support Uninvolved youth (n = 483) M SD

Peer social support Low Moderate High Maternal social support Low Moderate High ∗

for sex, grade and race. Females (M = 6.88, SD = 6.33) reported more anxiety/depression than males (M = 5.43, SD = 6.34). Students in grades 9 through 12 reported more anxiety/depression (M = 6.50, SD = 6.54) than youth in grades 7 and 8 (M = 5.94, SD = 6.22). With respect to race, White students reported more anxiety/depression (M = 6.53, SD = 6.56) than non-White students (M = 5.83, SD = 6.14). No two-way or three-way interactions were significant.

p < .05. Springer

Bullies (n = 112)

Victims (n = 98)

M

M

SD

SD

Bully-victims (n = 91) ANOVA Group × Support M SD F (6, 760) 2.17∗ , η2 = .02

4.57 4.50 4.80

5.91 4.75 5.18

10.33 6.24 6.84

8.73 5.80 7.21

10.03 7.26 10.23

7.70 6.79 8.24

9.43 8.57 11.00

6.76 5.25 8.03

6.14 4.69 3.53

5.71 5.06 4.61

9.97 6.00 6.68

7.89 5.69 7.77

11.51 8.54 6.74

8.82 6.34 6.32

10.92 8.91 8.04

7.89 6.44 3.34

.90


991

12

10

Anxiety/Depression

8

6

4

2

0 Uninvolved

Bully-victims

Victims

Bullies

Bully/Victim Groups Low

Moderate

High

Fig. 1 Anxiety/Depression scores by bully/victim groups

peer social support. Conversely, for bully-victims, victims, and bullies, those youth reported moderate levels of social support also reported the least anxiety/depression. For bullyvictims and victims, the highest levels of anxiety/depression were associated with high levels of perceived support from peers. For bullies, however, youth reporting the most anxiety/depression perceived their peer social support as low. In the second ANOVA we examined maternal social support. Results indicated that the interaction between bully/victim groups and maternal social support was not significant, suggesting that, regardless of bully/victim status, as maternal social support increased anxiety/depression decreased. Across groups individuals with low maternal social support reported the most anxiety/depression. Of note, in both ANOVAs main effects of sex and bully/victim groups remained significant.

Discussion The current investigation examined differences in perceived peer and maternal social support across bully/victim subtypes. First, we considered the influence of sex, age and race/ethnicity in perceptions of social support. Second, we analyzed differences in perceived social support among bul-

lies, victims, bully-victims, and uninvolved youth. Finally, we evaluated perceived peer and maternal social support as buffers against psychological distress across each of the bully/victim subtypes. Preliminary analyses supported extant research on social support among adolescents broadly, in particular research that has delineated differences in social support by gender and age. Specifically, females reported more peer social support than males, as has been found previously (Furman, 1996; Malecki and Elliott, 1999), although they did not report more maternal social support. Also, middle school youth from this investigation perceived more maternal social support than high school students, adding support to other studies that have found lower rates of social support among older students (Malecki and Elliott, 1999). Finally, results indicated that White students tended to report somewhat more peer social support, and somewhat less maternal social support, than non-White students. Consistent with the study hypotheses and previous research (Demaray and Malecki, 2003), uninvolved students reported greater perceived peer social support from their mothers and friends than victims and bully-victims, and greater maternal social support than victims. It is also consistent with previous research (Demaray and Malecki, Springer

992

2003) that bullies did not differ from uninvolved students in their perception of social support from peers. Given that bullies often affiliate with one another, and accordingly peer group members support bullying perpetration, this result is not surprising. Social identity theory (Tajfel and Turner, 1979) posits that individuals derive their social identity from the group to which they belong, consequently favoring behaviors that are consistent with group norms and emphasizing differences between their group and other groups that seem to be similar, to preserve the group’s identity. Perceiving positive social support from their immediate peer group could help preserve the group identity and thus maintain bullying behavior. It would be important for future research on social support and bullying involvement to consider more explicitly how peers are supportive of one another. It is probable that they are promoting their group identity through bullying perpetration. The current study’s finding that perceived social support is greater for students who are uninvolved in bullying perpetration and victimization adds to the extant literature, but says little about how and when students utilize their peers and families for social support. Do they turn to their friends or family when they experience victimization? In one study of young children, having a friend help or telling a teacher when victimized were more successful strategies in adjustment outcomes than fighting back or walking away (Kochenderfer and Ladd, 1997). This suggests that when support is available and utilized then adjustment outcomes are more positive. In contrast, another study found that students rated nonchalance (i.e., acting as if one did not care) as a more constructive response than either fighting back or helplessness (crying and avoiding) (Salmivalli et al., 1996). So, in some ways seeking social support might be seen as a less viable option than keeping one’s reaction to victimization to oneself. Future studies to move beyond the assessment of “perceived” social support and assess more comprehensively when certain coping strategies are employed, under which conditions, and at what point. Findings from the current investigation also point to the powerful influence of peer social support for youth involved in bullying in any capacity. Among bullies, those with low levels of social support reported the most anxiety/depression. It is possible that this subgroup of bullies did not use aggression proactively and therefore were not popular within their peer group (Pellegrini et al., 1999), and accordingly did not feel supported. For bully-victims and victims, however, high levels of peer social support were associated with more anxiety/depression. It might be that although these youth perceived high levels of social support they were not able to use it effectively, and accordingly the social support did not buffer against negative mental health symptoms. Or, it might be that those individuals who perceived high social support were not accurate in this perception. Another possibility is Springer


that a high level of social support is coupled with other negative friendship aspects (e.g., jealousy) among victims and bully-victims, which results in social support not buffering against psychological distress. This study additionally found that bullies, victims, and bully-victims who reported moderate social support also indicated the least anxiety/depression. Future research is needed to support and clarify this finding. Additional research also is needed on the effects of maternal social support among youth involved in bullying. In the current investigation although peer social support played a moderating role, maternal social support did not. This might be because youth are less willing to talk with their mothers about bullying involvement, and accordingly do not access support from their mothers that is directly related to these experiences. It is important to note, however, that across all bully/victim subtypes those youth with the least maternal social support reported the most anxiety/depression. This highlights that maternal social support is an important factor in promoting healthy psychological functioning. Although findings from the current study add valuably to our understanding of the function of perceived social support among youth across bully/victim subtypes, the study had a number of limitations. First, a methodological limitation of the present study is the reliance on students’ self-reports. More confidence in such findings could be gained by using a multi-informant approach that would prevent shared method variance. The present study also was limited in that it did not assess relational forms of aggression and victimization, which might be more common for adolescent girls. Also, the study did not consider contextual factors such as school climate. Numerous studies have demonstrated that school climate factors such as sense of belonging, school policies regarding bullying, and teacher and staff attitudes all impact students’ experiences of peer victimization. Future research might address this issue by considering how school climate factors relate to the perception and use of social support among peers. Finally, the sample consisted of middle and high school students from two Midwestern communities. Associations among perceived social support, victimization and bullying, and anxiety/depression might be different in other samples. Implications for public policy Findings from this study have important implications for public policy. Polices against bullying are becoming increasingly prevalent; as of 2005 17 states had enacted anti-bullying legislation (Dounay, 2005). Some of these state statutes mention the benefit of implementing bullying prevention programs (Limber and Snell, 2003), although given the increasing number of bullying prevention programs available it is often daunting for schools to select an effective one. Findings


from the current investigation suggest that one component crucial to effective programs is addressing student social support networks, from helping youth to access networks to assisting youth in developing the skills to effectively use these networks. This is supported by other research showing that victims who seek support in response to bullying are more likely to “escape” their victimization status than those youth who do not seek such support (Smith et al., 2004). Accordingly, state legislation might want to identify specific anti-bullying policies in their statutes that not only have empirical evidence supporting their efficacy, but also address social support as an integral component.

Conclusion This is one of the first studies to examine perceived social support across the bully/victim continuum. Bullies, victims, and bully-victims reported less perceived social support from peers and mothers than uninvolved students. Further, among bullies, victims, and bully-victims, individuals with moderate peer social support reported the least psychological distress. Future research should extend this work and examine specific components of social support that might buffer the impact of bullying involvement. Prevention and intervention approaches should consider the complex interplay between individuals’ utilization of social support and how peers might actually support bullying perpetration at least at a moderate level. Finally, anti-bullying programs should include components that encourage the development of healthy peer social support networks, and teach youth how to effectively utilize these networks.

References Achenbach TM (1991) Manual for the Youth Self-Report and 1991 Profile Bennett DS, Bates JE (1995) Prospective models of depressive symptoms in early adolescence: Attributional style, stress, and support. J Early Adolesc 15:299–315 Bond L, Carlin JB, Thomas L (2001) Does bullying cause emotional problems? A prospective study of young teenagers. Br Med J 323:480–484 Champion K, Vernberg E, Shipman K (2003) Nonbullying victims of bullies: Aggression, social skills, and friendship characteristics. Appl Dev Psychol 24:535–551 Cohen S, Wills TA (1985) Stress, social support, and the buffering hypothesis. Psychol Bull 98:310–357 Colarossi LG, Eccles JS (2003) Differential effects of support providers on adolescents’ mental health. Soc Work Res 27:19–30 Connell JP, Spenser MB, Aber JL (1994) Educational risk and resilience in African-American youth: Context, self, action, and outcomes in school. Child Dev 65:493–506 Crick NR, Grotpeter JK (1996) Children’s treatment by peers: Victims of relational and overt aggression. Dev Psychopathol 8:367– 380

993 Crick NR, Nelson DA (2002) Relational and physical aggression within friendships: Nobody told me there’d be friends like this. J Abnorm Child Psychol 30:599–607 Demaray MK, Malecki CK (2003) Perceptions of the frequency and importance of social support by students classified as victims, bullies, and bully/victims in an urban middle school. School Psychol Rev 32:471–489 Denny S, Clark TC, Fleming T, Wall M (2004) Emotional resilience: Risk and protective factors for depression among alternative education students in New Zealand. Am J Orthopsychiatry 74:137– 149 Dounay J (2005) State Anti-bullying Statutes. Retrieved March 1, 2006, from http://www.ecs.org/clearinghouse/60/41/6041.htm Espelage DL, Bosworth K, Simon TR (2000) Examining the social context of bullying behaviors in early adolescence. J Couns Dev 78:326–333 Espelage DL, Holt MK (2001) Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. In Geffner R, Loring M (eds) Bullying behaviors: Current issues, research, and interventions. The Haworth Press, Binghampton, NY Espelage DL, Holt MK, Henkel RR (2003) Examination of peer-group contextual effects on aggression during early adolescence. Child Dev 74:205–220 Furlong MJ, Chung A, Bates M, Morrison RL (1995) Who are the victims of school violence? A comparison of student non-victims and multi-victims. Educ Treat Child 18:282–299 Furman W (1996) The measurement of friendship perceptions: Conceptual and methodological issues. In Bukowski WM, Newcomb AF, Hartup WW (eds) The company they keep: Friendship in childhood and adolescence. Cambridge University Press, London, pp. 41–65 Furman W, Buhrmester D (1985) Children’s perceptions of the personal relationships in their social networks. Dev Psychol 21:1016–1024 Furman W, Buhrmester D (1992) Age and sex differences in perceptions of networks of personal relationships. Child Dev 63:103–115 Grant KE, O’Koon JH, Davis TH, Roache NA, Poindexter LM, Armstrong ML, et al (2000) Protective factors affecting low income urban African-American youth exposed to stress. J Early Adolesc 20:388–417 Hawker DSJ, Boulton MJ (2000) Twenty years’ research on peer victimization and psychosocial maladjustment: A meta-analytic review of cross-sectional studies. J Child Psychol Psychiatry Allied Discip 41:441–455 Haynie DL, Nansel T, Eitel P (2001) Bullies, victims, and bully/victims: Distinct groups of at-risk youth. J Early Adolesc 21:29–49 Hodges EVE, Boivin M, Vitaro F, Bukowski WM (1999) The power of friendship: Protection against an escalating cycle of peer victimization. Dev Psychol 35:94–101 Holt MK, Espelage DL (2005) Social Support as a Moderator Between Dating Violence Victimization and Depression/Anxiety Among African American and Caucasian Adolescents. School Psychol Rev 34:309–328 Hunter SC, Boyle JME (2004) Appraisal and coping strategy use in victims of school bullying. Br J Dev Psychol 74:83–107 Jessor R, Van Den Bos J, Vanderryn J (1995) Protective factors in adolescent problem behavior: Moderator effects and developmental change. Dev Psychol 31:923–933 Kaltiala-Heino R, Rimpelae M, Rantanen P (2001) Bullying at school: An indicator for adolescents at risk for mental disorders. J Adolesc 23:661–674 Kochenderfer BJ, Ladd GW (1997) Victimized children’s responses to peers’ aggression: Behaviors associated with reduced versus continued victimization. Dev Psychopathol 9:59–73 Kochenderfer-Ladd B, Skinner K (2002) Children’s coping strategies: Moderators of the effects of peer victimization? Dev Psychol 38(2):267–278 Springer

994 Kolbo JR (1996) Risk and reslience among children exposed to family violence. Violence Vict 11:113–128 Kumpulainen K, Rasanen E, Henttonen I (1998) Children involved in bullying: Psychological disturbance and the persistence of the involvement. Child Abuse Negl 23:1253–1262 Limber SP, Snell MA (2003) State laws and policies to address bullying in schools. School Psychol Rev 32:445–455 Luster T, Small SA (1997) Sexual abuse history and problems in adolescence: Exploring the effects of moderating variables. J Marriage Fam 59:131–142 Malecki CK, Elliott SN (1999) Adolescents’ ratings of perceived social support and its importance: Validation of the Student Social Support Scale. Psychol Schools 36:473–483 McCreary ML, Slavin LA, Berry EJ (1996) Predicting problem behavior and self-esteem among African-American adolescents. J Adolesc Res 11:216–234 Nansel TR, Haynie DL, Simons-Morton BG (2003) The association of bullying and victimization with middle school adjustment. J Appl School Psychol 19:45–61 Nansel TR, Overpeck M, Pilla RS, Ruan WJ, Simons-Morton BG, Scheidt P (2001) Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. J Am Med Assoc 285:2094–2100 Naylor P, Cowie H, del Rey R (2001) Coping strategies of secondary school children in response to being bullied. Child Psychol Psychiatry Rev 6:114–120 Neary A, Joseph S (1994) Peer victimization and its relation to selfconcept and depression among schoolgirls. Person Individ Diff 16:183–186 Olweus D (1995) Bullying or peer abuse at school: Intervention and prevention. In Davies G, Lloyd-Bostock S (eds) Psychology, law, and criminal justice: International developments in research and practice. Walter De Gruyter, Oxford, England Parker JG, Asher SR (1993) Friendship and friendship quality in middle childhood: Links with peer group acceptance and feelings of loneliness and social dissatisfaction. Dev Psychol 29:611–621 Parker JG, Low CM, Walker AR, Gamm BK (2005) Friendship jealousy in young adolescents: Individual differences and links to sex, selfesteem, aggression, and social adjustment. Dev Psychol 41:235– 250 Pellegrini AD, Bartini M, Brooks F (1999) School bullies, victims, and aggressive victims: Factors relating to group affiliation and victimization in early adolescence. J Educ Psychol 91:216– 224

Springer

J Youth Adolescence (2007) 36:984–994 Prinstein MJ, Boergers J, Vernberg EM (2001) Overt and relational aggression in adolescents: Social-psychological adjustment of aggressors and victims. J Clin Child Psychol 30:479–491 Rigby K (2000) Effects of peer victimization in schools and perceived social support on adolescent well-being. J Adolesc 23:57–68 Rigby K, Slee PT (1999) Suicidal ideation among adolescent school children, involvement in bully victim problems, and perceived low social support. Suicide Life-Threat Behav 29:119–130 Salmivalli C, Karhunen J, Lagerspetz K (1996) How do the victims respond to bullying? Aggressive Behav 22:99–109 Schwartz D, Dodge K, Pettit GS, Bates JE (2000) Friendship as a moderating factor in the pathway between early harsh home environment and later victimization in the peer group. Dev Psychol 36:646–662 Seals D, Young J (2003) Bullying and victimization: Prevalence and relationship to gender, grade level, ethnicity, self-esteem, and depression. Adolescence 38:735–747 Seidman E, Allen L, et al (1995) Development and validation of adolescent-perceived microsystem scales: Social support, daily hassles, and involvement. Am J Community Psychol 23:355–388 Smith C (2003) Theorizing religious effects among American adolescents. J Sci Study Religion 42:17–30 Smith PK, Talamelli L (2001) How pupils cope with bullying: A longitudinal study of successful and unsuccessful outcomes. Paper presented at the British Psychology Society Conference, Glasgow, Scotland Smith PK, Talamelli L, Cowie H, Naylor P, Chauhan P (2004) Profiles of non-victims, escaped victims, continuing victims and new victims of school bullying. Br J Educ Psychol 74:565–581 Swearer SM, Song SY, Cary PT (2001) Psychosocial correlates in bullying and victimization: The relationship between depression, anxiety, and bully/victim status. In Geffner RA, Loring M (eds) Bullying behavior: Current issues, research, and interventions. Haworth Maltreatment and Trauma Press/The Haworth Press, Inc., Binghamton, NY, pp. 95–121 Tajfel H, Turner JC (1979) An integrative theory of intergroup conflict. In Worchel S, Austin WG (eds) The social psychology of intergroup relations. Brooks/Cole Publishers, Monterey Visser JH, Van Der Ende J, Koot HM, Verhulst FC (1999) Continuity of psychopathology in youths referred to mental health services. J Am Acad Child Adolesc Psychiatry 38:1560–1568 Wasserstein SB, La Greca AM (1996) Can peer support buffer against behavioral consequences of parental discord? J Clin Child Psychol 25:177–182