You are on page 1of 22

Journal of Early Adolescence

Volume 27 Number 4
November 2007 457-478
2007 Sage Publications
Level of Bonding to School 10.1177/0272431607302940
http://jea.sagepub.com

and Perception of the School hosted at


http://online.sagepub.com

Environment by Bullies,
Victims, and Bully Victims
Nancy J. Cunningham
University of Louisville, Kentucky

Commitment and attachment to school and perception of school norms were


examined in a sample of sixth, seventh, and eighth graders to determine whether
bullies, victims, bully victims, and students who reported no or low levels of bul-
lying and victimization differed in their level of bonding to school and their per-
ceptions of standards and expectations for behavior in the school environment
(protective factors). Risk factors for bullying were also examined. Results of a
discriminant analysis demonstrated differences among the groups on the mea-
sures of risk and protective factors and perception of school norms. The group-
ing of variables differentiated between the comparison group and the bully,
victim, and bully victim groups on a dimension of healthy functioning indicated
by low risk for bullying and an investment in prosocial behaviors and beliefs.
The results have implications for schools in promoting prosocial bonding
through the development of academic, emotional, and social competence.

Keywords: bullying and victimization; bully victims; school bonding;


school norms; school commitment; school attachment

R esearch on bullying and victimization has increased in the United


States during the last decade, taking impetus from the work of
researchers in other countries (Ma, 2001, 2002; Ojala & Nesdale, 2004;
Olweus, 1994, 2003; Salmivalli & Nieminen, 2002; Salmivalli & Voeten,
2004; Sutton & Smith, 1999) and from reports issued after the series of
school shootings in the 1990s indicating that the shooters had been long-
term victims of peer bullying (Vossekuil, Fein, Reddy, Borum, &
Modzeleski, 2000, 2002). Bullying is now recognized as a serious mental

Authors Note: Please address correspondence to Nancy J. Cunningham, 317 Education,


University of Louisville, Louisville, KY 40292; e-mail: nancy.cunningham@louisville.edu.

457

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


458 Journal of Early Adolescence

health problem with harmful short- and long-term consequences for both
bullies and victims (Bond, Carlin, Thomas, Rubin, & Patton, 2001; Farrington,
1991; Fox & Boulton, 2005; Kumpulainen et al., 1998; Ladd & Ladd, 2001;
Olweus, 1993; Rigby, 2001; Williams, Chambers, Logan, & Robinson, 1996;
Wolke, Woods, Bloomfield, & Karstadt, 2001).
Research on school-based bullying has evolved from describing the
characteristics of bullies and victims to examining bullying behavior in the
context of the school environment. Much of the child and adolescent bully-
ing that occurs happens in schools, placing schools in a key position to help
us understand the dynamics of bullying and victimization and to test the
effectiveness of strategies for prevention and intervention. This study inves-
tigates the relationship between school-related factors and bullying and vic-
timization. Commitment and attachment to school and perception of school
norms were examined to determine whether bullies, victims, and bully vic-
tims (those students who report both bullying and being the victim of bul-
lying) differed in their level of bonding to school and their perceptions of
standards and expectations for behavior in the school environment. Risk
factors for bullying (history of bullying, having friends who bully, and
favorable attitudes toward bullying) were also examined.
Hawkins and Catalanos social development model (SDM) provided
the theoretical base for the study. According to the SDM, school bonding
and healthy beliefs serve as protective factors against the development of
antisocial behaviors by buffering against risk (Catalano, Haggerty,
Oesterle, Fleming, & Hawkins, 2004). Consistent with the model, it was
hypothesized that nonbully/nonvictim students or comparisons (students
who report no or low levels of bullying and victimization) would be
bonded more strongly to school and have more positive perceptions of
school norms than bullies, victims, or bully victims. Given their involve-
ment in bullying behavior, it was predicted that bullies and bully victims
would show greater risk than victims or comparisons. Furthermore,
because of research showing that bully victims appear to suffer from more
severe psychological and social problems than bullies or victims (Haynie
et al., 2001; Juvonen, Graham, & Shuster, 2003; Ma, 2001; Pellegrini,
2002; Pellegrini, Bartini, & Brooks, 1999; Salmivalli & Nieminen, 2002),
it was expected that bully victims would score lowest on the measures of
protective factors and highest on risk. Because sex differences have been
found for both level and type of bullying, sex differences were also exam-
ined in terms of protection and risk.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 459

The Nature of Early Adolescent Bullying and


Victimization

Bullying is an act of social aggression. A commonly used definition was


developed by Olweus (1994) from his pioneering research in the schools of
Norway. Bullying is defined as a repeated behavior, either verbal or physi-
cal, that occurs over time in a relationship characterized by an imbalance of
strength or power (Olweus, 1994). Bullying can be perpetrated by more
than one bully on more than one victim and can encompass a range of phys-
ical and psychological behaviors. As Pellegrini notes, bullying becomes
more sexual in nature when children move into puberty (Pellegrini, 2002).
Bullying can also include behaviors such as gossiping and exclusion, often
called indirect or relational bullying (Espelage & Swearer, 2003). Swearer
and Doll (2001) point out that behaviors must be understood within the
social context in order to be labeled bullying. Behaviors tolerated
among friends, such as teasing and name calling, would be labeled bully-
ing outside the friendship circle (McConnell & Odom, 1986). Bullying can
be proactive in nature, motivated to achieve some end, or reactive as an
emotional response to frustration or anger (Pellegrini, 2002).
Bullies, victims, and bully victims have different psychological and
social profiles. Adolescent bullies tend to have high emotionality and low
self-control (Haynie et al., 2001; Pellegrini et al., 1999). Although they are
both proactively and reactively aggressive, bullies appear to use proactive
aggression to establish dominance and leadership in their peer group
(Juvonen et al., 2003; Pellegrini et al., 1999). Bullies may enjoy social status
among their peers, perhaps partly because they challenge adult norms and
encourage exploration of new adolescent roles (Juvonen et al., 2003), but
their friendships are primarily with other bullies (Pellegrini et al., 1999).
They show little empathy for their peers (Bernstein & Watson, 1997) and low
social anxiety (Juvonen et al., 2003), and they view bullying as a way to get
what they want (Griffin & Gross, 2004). Bullies are more likely to hold anti-
social beliefs or values and show acceptance for antisocial behaviors (Haynie
et al., 2001; Pellegrini et al., 1999). They are also more inclined to be dis-
engaged from and do poorly in school (Juvonen et al., 2003).
Adolescent victims tend to be depressed and lonely with poor self-
esteem; they are socially anxious, hold low social status among their peers,
and have few friends (Haynie et al., 2001; Juvonen et al., 2003; Olweus,
1994). Victims have negative attitudes toward bullying and do not endorse
aggressiveness as means to an end, yet they report friendships with bullies
and bully victims (Pellegrini et al., 1999).

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


460 Journal of Early Adolescence

Bully victims are often described as the most troubled of the three groups.
They tend to be highly emotional, impulsive, and anxious, and they exhibit
poor social skills; consequently, their behavior makes them unpopular with
and disliked by their peers (Carney & Merrell, 2001; Olweus, 2003;
Pellegrini et al., 1999). Their bullying behaviors are often reactive in nature.
In response to being victimized by others who are more powerful than they
are, they strike out at others who they perceive to be weaker. They are often
identified as having conduct problems with poor school adjustment and lack
of bonding to school (Haynie et al., 2001; Juvonen et al., 2003).

Frequency of Early Adolescent


Bullying and Victimization

In a retrospective study of school bullying, Elsea and Rees (2001) found


that 73% of participants reported being bullied at school at some point in
time, most frequently between the ages of 11 and 13. Elsea and Rees sug-
gest that these early adolescent memories of bullying may be most memo-
rable because they are the most painful, reflecting the shift from physical to
psychological bullying that occurs at about this age. The 2001 Health
Behavior in School-Aged Children survey data collected on a nationally
representative sample of students in Grades 6 through 10 showed that bul-
lying in schools occurred most frequently in sixth through eighth grades
(Nansel et al., 2001). In a survey of 15,686 students sampled from public,
parochial, and other private schools, researchers found that 29% of the
students responding had been involved in some aspect of bullying as the
bully, the victim, or both. Thirteen percent said they had engaged in mod-
erate or frequent bullying of others, 10.6% had been bullied moderately or
frequently, and 6.3% had both bullied others and been bullied themselves
(bully victims). There was little variation among the responses of urban,
suburban, and rural participants. Males were more likely to bully others and
to be victims of bullying than females and more likely to indicate that they
had been bullied physically. Females admitted to being bullied more fre-
quently verbally and indirectly (rumors, exclusion, etc.) than males.
In a study of 454 rural seventh and eighth graders, Seals and Young
(2003) found similar percentages. Forty-five percent of the seventh graders
and 42% of the eighth graders reported that bullying occurred often in
school; 24% were directly involved in bullying one or more times per week
as bullies (10%), victims (13%), and bully victims (1%). Verbal bullying
was reported more frequently than physical bullying. Seals and Young

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 461

(2003) also looked at individual versus group bullying and found that both
males and females who bullied alone tended to target victims of the same sex;
however, more females than males were involved in mixed-sex group bullying.
Pellegrini has developed a body of research around bullying during the
transition from elementary to middle school (Pellegrini, 2001; Pellegrini &
Bartini, 2000; Pellegrini et al., 1999). During this developmental stage
when adolescents enter puberty and develop a greater orientation toward
their peer group, the quest for position in the peer group and initial explo-
ration of heterosexual relationships lead to increases in physical, verbal,
and sexual bullying (Pellegrini, 2001). Pellegrinis data show an increase in
bullying behavior from fifth to sixth grade (after the move to middle
school) followed by a decrease in bullying behavior by the end of sixth
grade. He attributes this rise in bullying by new sixth graders to an effort to
establish dominance in a new environment where they are the youngest and
often the smallest (Pellegrini & Bartini, 2000). Pellegrini (2001, 2002;
Pellegrini et al., 1999) found that bullying in early adolescence typically
involves boys bullying other boys and, less frequently, boys victimizing
girls either directly or with sexually aversive behaviors. Pellegrini et al.
(1999) found that males bullied more than females and that females held
more negative attitudes toward bullying than males.

The Social Development Model

The SDM guided the conceptualization of the study. The SDM inte-
grates social learning theory, social control theory, and differential associa-
tion theory to explain the process by which youth bond to prosocial systems
(systems with healthy norms) in order to achieve personal success
(Catalano et al., 2004; Catalano & Hawkins, 1996). Bonding is dependent
on an individual having the skills and opportunities as well as recognition
for engaging in successful behaviors in a system. The SDM emphasizes the
importance of environmental conditions (i.e., in the school environment)
that can support the development of positive (prosocial) behaviors or rein-
force the development of negative (antisocial) behaviors such as bullying.
According to the model, behavior will be prosocial or antisocial depending
on the predominant behaviors, norms, and values held by those individuals
or institutions to which/whom the individual is bonded (Catalano et al.,
2004, p. 252) and emphasizes the development of prosocial bonding as a
protective factor against risk. Once bonds are strongly established, they
inhibit behavior inconsistent with the norms and values of the institution to

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


462 Journal of Early Adolescence

which the individual is bonded. Catalano et al. (2004) claim that school
bonding plays a central role in inhibiting antisocial behavior and promoting
healthy development during childhood and adolescence.
The Social Development Research Group has tested the SDM through a
number of research projects including two longitudinal studies carried out
over a number of years in 18 Seattle-area public elementary schools
(Catalano et al., 2004). Both studies used a developmentally adjusted, mul-
ticomponent intervention to affect the primary systems that influence
child development during the elementary years. Herrenkohl and col-
leagues (2003), using this data, found that bonding to school at age 15
was related to a lower probability of violence among aggressive youth at
age 18. These studies, along with others (Catalano & Kosterman, 1996;
Hawkins, Guo, Hill, Battin-Pearson, & Abbott, 2001; Lonczak, Abbott,
Hawkins, Kosterman, & Catalano, 2002) provide evidence for the impor-
tance of school bonding for promoting healthy development (academic
success and social skills) and reducing risk for the development of problem
behaviors (substance abuse, delinquency, gang membership, violence, aca-
demic problems, and sexual activity).
In the present study, school bonding, school norms, and the risk factors
of favorable attitudes toward bullying, friends who bully, and past history
of bullying were examined for bullies, victims, bully victims, and compar-
isons. Based on the SDM, it was hypothesized that students who reported
little or no involvement in bullying and victimization (comparisons) would
be more strongly bonded to school, have more positive perceptions of gen-
eral school norms and norms related to bullying, and indicate less risk than
students who reported involvement in bullying and/or victimization (bul-
lies, victims, and bully victims).

Method

Data were collected on a sample of private (Catholic) school students (n =


517, including 287 girls and 230 boys) in sixth, seventh, and eighth grades
from 11 elementary schools in a midsouthern urban area. Participants ranged
in age from 11 to 15 years with a mean age of 12.5 years. Reported ethnicity
was 93.4% Caucasian, 0.6% Asian, 0.6% Hispanic, 1.4% Native American,
and 2.5% who categorized themselves as Other. Thirteen percent of the sam-
ple reported receiving free/reduced lunch. Students took the survey as part of
a schoolwide assessment to measure the level of bullying behaviors in the
upper elementary grades. All students who participated provided informed
consent.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 463

The survey was administered during a regular classroom period, and students
were monitored by teachers who received instructions in how to administer the
survey and how to respond to student questions during administration.

Survey
Items measuring risk factors for bullying and protective factors in the
school environment were developed based on descriptions of risk and pro-
tective factors in the work of Hawkins and Catalano (Catalano et al., 2004;
Catalano & Hawkins, 1996; Hawkins, Catalano, & Miller, 1992; Hawkins
et al., 2001) and Benard (1991). A principal component factor analysis
using varimax rotation was conducted on 72 items developed to measure
the variables studied in this investigation as well as additional variables
measured by the survey (N = 1,308 middle-school-aged public and private
school students). Most items loaded on factors as anticipated. Seven factors
related to risk and protective factors emerged that were used to develop
the scales employed in the present study. Internal consistency reliability
coefficients were computed for the subsets of items identified on the fac-
tor analysis. Items were chosen for each scale that maximized the internal
consistency of the scale. The following scales were used in the present
analysis to measure risk and protective factors in the school environment.
Participants responded to each item using a 4-point response scale indicat-
ing level of agreement with the content of the item (strongly agree, agree,
disagree, strongly disagree).

Bonding scales. The Attachment and Commitment scales were used to


measure bonding to school. Attachment and commitment have been identi-
fied in the research literature as two important components of bonding
(Catalano et al., 2004). The Commitment scale is comprised of six items
measuring commitment to school in terms of performing well academi-
cally, working hard, and following school rules (e.g., I work hard in my
classes, and I usually follow the school rules.). Internal consistency for
the Commitment scale is .86. The Attachment scale is comprised of eight
items and assesses emotional bonding to school reflected through items
measuring a sense of belonging at school with school viewed as a support-
ive and caring environment (e.g., People at my school care about me, and
I feel like I belong at this school.). Internal consistency for the
Commitment scale is .85. High scores on these scales indicate a higher
attachment and commitment to school.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


464 Journal of Early Adolescence

Norms scales. Two scales were used to measure student perception of


school norms. The General Norms scale is comprised of 10 items and mea-
sures student perceptions of general expectations for academic perfor-
mance and behavior in school (e.g., My teachers are clear about what they
expect me to do in my classes, and At this school, we are expected to treat
everyone with respect.). Internal consistency reliability for the General
Norms scale is .86. The Bully Norms scale includes four items assessing
student perception of whether adults view bullying as a problem and inter-
vene to stop bullying and other hurtful behaviors in the school (e.g., Kids
get away with hurting other kids at this school, and When a student gets
bullied, the teachers at this school do something to stop it.). Internal con-
sistency reliability for the Bully Norms scale is .70. Higher scores on these
scales indicate a healthier perception of school norms.

Risk factor scales. Three scales were used to assess risk for bullying
based on common risk factors identified through syntheses of the risk fac-
tor literature for other adolescent problem areas. Items were introduced
with a definition of bullying (including types of bullying behavior) so that
students understood that the items were intended to measure attitude toward
bullying, having friends who bully, and having a history of bullying behav-
ior. Attitude Toward Bullying consists of five items measuring positive atti-
tudes toward bullying and aggressive behaviors directed at others (e.g., Its
not so bad to tease other kids, and Most students who get bullied bring it
on themselves.). Internal consistency reliability for the Attitude Toward
Bullying scale is .82. The Bully Friends scale is comprised of four items
and measures whether the respondents friends engage in the four types of
bullying behavior measured by the survey (e.g., My friends sometimes
tease other kids or call them names, and My friends sometimes make sure
that certain kids are excluded from our group.). Internal consistency for the
Bully Friends scale is .69. The Bully History scale includes four items that
measure past history of fighting, rule violation, and bullying other students
(e.g., Ive always picked on other kids, and Others have always thought
of me as a bully.). The internal consistency coefficient for the Bully
History scale is .71. The items on these scales are reverse scored so that
higher scores indicate less risk.

Assignment of Students to Bully Role Groups


For the purpose of examining differential perceptions of the school envi-
ronment by students involved in bullying and victimization, the student

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 465

sample was divided into groups (bully role) based on scores on the Bully
and Victim scales. The Bully and Victim scales were developed from the
research literature identifying types of bullying behaviors manifested dur-
ing adolescence. The Bully scale includes four items that measure physical
bullying behavior (hitting, kicking, or pushing another student), direct ver-
bal bullying behavior (teasing, insulting, or calling other students names),
indirect/relational bullying behavior (gossiping, telling false stories about
others, and excluding others), and sexual bothering (inappropriate touching
or making sexual comments). The Victim scale includes four items that
measure being the victim of the four types of bullying behavior measured
by the Bully scale. Both sets of bully items and victim items were intro-
duced with a definition of bullying so that the student responded as either a
perpetrator or victim of bullying behavior. For example, in the case of the
bullying items, the following instructions were provided: The following
questions ask about whether you have bullied a student who was weaker or
less powerful than you. Remember that bullying is behavior that is directed
at someone who is weaker or less powerful in some way and can be any of
the following [with the four types of bullying defined]. Participants
responded on a 4-point continuum according to level of perpetration or vic-
timization (never, once in a while, pretty often, often). The internal consis-
tencies of the Bully and Victim scales are .75 and .72, respectively. Higher
scores indicate a higher level of self-report of bullying and victimization.
Scores one standard deviation above the means on the Bully and Victim
scales were used as cutoff scores to divide students into four groups: (1) com-
parison group (nonbully/nonvictim group), that is, students below the cutoff
scores on both the Bully and Victim scales; (2) bully group, that is, students
above the cutoff score on the Bully scale and below the cutoff score on the
Victim scale; (3) victim group, that is, students above the cutoff score on
the Victim scale and below the cutoff score on the Bully scale; and (4) bully-
victim group, that is, students above the cutoff scores on both the Bully and
Victim scales. Using these criteria, the sample consisted of the following:
comparison, n = 361 (75%); bully, n = 36 (7.5%); victim, n = 56 (11.5%); and
bully victim, n = 28 (6%). Means and standard deviations for each of the
groups on the Bully and Victim scales are reported in Table 1.

Results

A discriminant analysis was conducted using sex and the risk and protective
factor variables as predictors of membership in the four groupscomparison,

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


466 Journal of Early Adolescence

Table 1
Means and Standard Deviations for All Variables
by Bully Group and Sex

Variable X SD

Bully Comparison Male 1.74 1.05


Female 1.48 0.99
Bully Male 5.38 2.1
Female 4.63 1.15
Victim Male 1.77 1.07
Female 1.84 1.4
Bully victim Male 6.82 2.38
Female 6.27 3.07
Victim Comparison Male 2.79 1.44
Female 2.5 1.38
Bully Male 3.46 1.44
Female 3.94 1.06
Victim Male 6.88 1.4
Female 7.16 1.57
Bully victim Male 8.06 1.52
Female 8.91 2.3
Commitment Comparison Male 13.68 0.23
Female 15.22 0.19
Bully Male 12 0.6
Female 12.43 0.75
Victim Male 14.2 0.56
Female 14.68 0.5
Bully victim Male 10.88 0.68
Female 9.46 0.84
Attachment Comparison Male 16.96 0.32
Female 17.66 0.27
Bully Male 15.18 0.84
Female 15.5 1.05
Victim Male 13.68 0.79
Female 13.42 0.71
Bully victim Male 12.35 0.95
Female 12.36 1.18
General norms Comparison Male 22.95 0.35
Female 24.03 0.3
Bully Male 21.18 0.92
Female 19.43 1.16
Victim Male 22.56 0.87
Female 22.16 0.78
Bully victim Male 20.06 1.05
Female 16.64 1.3

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 467

Bully norms Comparison Male 6.47 0.19


Female 7.33 0.17
Bully Male 5.09 0.51
Female 5.29 0.64
Victim Male 4.56 0.48
Female 5 0.43
Bully victim Male 4.24 0.58
Female 3.55 0.72
Favorable attitudes Comparison Male 10.84 0.22
Female 12.14 0.19
Bully Male 6.36 0.57
Female 9.64 0.72
Victim Male 12 0.54
Female 11.61 0.48
Bully victim Male 7.59 0.65
Female 6.28 0.81
Bully friends Comparison Male 6.99 0.18
Female 7.98 0.16
Bully Male 4.55 0.48
Female 5.79 0.6
Victim Male 5.64 0.45
Female 5.48 0.41
Bully victim Male 4 0.55
Female 4.09 0.68
Bully history Comparison Male 8.7 0.17
Female 10.37 0.14
Bully Male 5.86 0.43
Female 8.36 0.54
Victim Male 8.76 0.41
Female 9.23 0.37
Bully victim Male 7.18 0.49
Female 6.64 0.61

bully, victim, and bully victim. All variables were entered simultaneously.
Table 1 includes means and standard deviations for males and females in each
bully role group on each dependent variable. Intercorrelations for the predic-
tor variables included in the discriminant analysis are shown in Table 2.
The first two functions were significant and will be interpreted. The first
function [Wilks of .566 with (24, N = 481) = 270.15, p < .001]
accounted for 61.8% of the variance associated with bully role. The second
function [Wilks of .796 with (14, N = 481) = 108.27, p < .001]
accounted for 36.4% of the variance associated with bully role. Table 3
shows the standardized discriminant coefficients for the two significant
functions. The relative magnitude of the coefficients provides information

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


468 Journal of Early Adolescence

Table 2
Intercorrelations for Predictor Variables Included
in the Discriminant Analysis
Predictor Variables 1 2 3 4 5 6 7 8

1. Sex 0.2 0.07 0.05 0.14 0.2 0.18 0.33


2. Commitment 0.58 0.6 0.26 0.53 0.28 0.46
3. Attachment 0.71 0.45 0.41 0.25 0.3
4. General norms 0.44 0.47 0.16 0.24
5. Bully norms 0.35 0.28 0.2
6. Attitudes toward bullying 0.28 0.45
7. Bully friends 0.35
8. Bully history

Note: All correlations are significant at .01.

Table 3
Standardized Discriminant Coefficients for Sex and Risk and
Protective Factor Variables for Predicting Group Membership for
Comparisons, Bullies, Victims, and Bully Victims
Standardized Discriminant
Coefficients

Variables df1 df2

Sex 0.1 0.028


Commitment 0.017 0.389
Attachment 0.169 0.991
General norms 0.048 0.382
Bully norms 0.274 0.284
Favorable attitudes toward bullying 0.308 0.652
Bully friends 0.485 0.214
Bully history 0.273 0.103

about the relative importance of each predictor variable in discriminating


between groups. As measures of the group means of the composite vari-
ables, the group centroids for the two significant discriminant functions are
plotted in Figure 1 and show the distance between the group means for the
four groups.
The canonical correlation between the first discriminant function and the
levels of bully role was .54, indicating that 28.9% of the variability between
groups could be accounted for by the first linear combination of the predictor

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 469

Figure 1
Plot of Group Centroids on Functions 1 and 2 Depicting the
Differences Between the Group Means for Comparisons, Bullies,
Victims, and Bully Victims

1.50

Victims
1.00
Function 2

0.50

0.00

Comparisons

Bully Victims
-0.50

Bullies
-1.00

-2.00 -1.50 -1.00 -0.50 0.00 0.50

Function 1

variables. The group centroids for the first function were .34 (comparison),
1.23 (bully), .54 (victim), and 1.72 (bully victim). As shown on the hor-
izontal axis of Figure 1, the discriminant function separated the comparison
group from the other three groups. The variables contributing most to dis-
criminating between groups were bully friends (.485), attitude toward bul-
lying (.308), bully norms (.274), and bully history (.273). These are the
three risk factor scales for bullying and the Norms scale that measures
student perceptions of whether teachers allow bullying to occur in the
school. High scores on this grouping of variables appear to measure a
healthy orientation with low risk for bullying and an investment in proso-
cial behaviors and beliefs.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


470 Journal of Early Adolescence

The canonical correlation between the second discriminant function and


the levels of bully role was .43, indicating that 18.7% of the variability
between groups could be accounted for by the second linear combination of
predictor variables. As shown on the vertical axis of Figure 1, the group
centroids for the second function were .09 (comparison), .78 (bully),
1.21 (victim), and .27 (bully victim). The discriminant function separated
the victim group from the other three groups. The variables contributing
most to discriminating between groups were attachment (.991), attitude
toward bullying (.652), commitment (.389), general norms (.382), and bully
norms (.284). This grouping includes the four protective factor variables;
however, attachment and bully norms are negatively weighted, indicating a
low attachment to school and the perception that teachers allow or support
bullying in the school environment. High scores on this grouping of vari-
ables appear to measure an orientation to school characterized by commit-
ment to and perception of high adult expectations regarding academic
achievement and rule following; however, this is accompanied by low emo-
tional bonding to school characterized by perceptions of lack of support and
tolerance for bullying by the adults in the school environment.
Classification rates were examined to determine how well the predictors
discriminated among the bully role groups. Overall, 66.3% of the partici-
pants were correctly classified, which included 68.1% of the comparisons,
55.6% of the bullies, 69.6% of the victims, and 50.0% of the bully victims.
Bullies were most often misclassified as bully victims (22.2%) or compar-
isons (19.4%); victims were most often misclassified as comparisons
(21.4%); and bully victims were most often misclassified as bullies
(25.0%). As can be seen in Figure 1, both discriminant functions were poor-
est in separating the bully and bully-victim groups from each other. Results
of ANOVAs with Tukey post hoc tests on the risk and protective factor vari-
ables (Table 4) indicate that bullies and bully victims scored similarly to
each other on the risk factor and norms variables but significantly differ-
ently from each other on the bonding variables.

Discussion

The study examined differences among bully role groups and school
bonding, perception of school norms, and risk for bullying behavior.
Consistent with the SDM, students who reported low levels of bullying and
victimization (comparisons) appeared to be the most strongly bonded to
school with an investment in prosocial behaviors and beliefs. The first

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 471

Table 4
Summary of Univariate Analyses of Variance and Post Hocs for the
Risk and Protective Factor Variables for Comparisons (C), Bullies
(B), Victims (V), and Bully Victims (BV)

Source df X2 F p Tukey Post Hoc

Commitment 3 193.62 24.76 .0001 BV < B < V, C


Attachment 3 404.69 26.26 .0001 BV, V < B, C
General norms 3 314.95 16.84 .0001 BV, B < V, C
Bully norms 3 150.49 26.42 .0001 BV, V, B < C
Attitudes toward bullying 3 270.81 37.63 .0001 BV, B < C, V
Bully friends 3 180.36 35.26 .0001 BV, B < V < C
Bully history 3 110.23 26.70 .0001 B, BV < V, C

Note: Error df = 473. Tukey post hocs are significant at the .05 level.

discriminant function differentiated between this group and the bully, vic-
tim, and bully-victim groups with the group centroids showing the bully-
victim group most distant from the comparison group. This finding is
consistent with that of Haynie et al. (2001), who found that measures of
school adjustment (defined similarly to commitment in this study) and
school bonding (combining aspects of both commitment and attachment as
defined in this study) differentiated between their comparison group and
bully, victim, and bully-victim groups with the comparisons and bully vic-
tims at opposite ends of the continuum.
This result is also consistent with the research literature that has iden-
tified bully victims as being the most troubled of the bully, victim, bully-
victim triad (Juvonen et al., 2003; Ma, 2001; Pellegrini, 2002; Pellegrini
et al., 1999; Salmivalli & Nieminen, 2002). Although the classification
data indicate that the group of variables studied did not perform well in dis-
criminating between bullies and bully victims (as shown in Figure 1),
Tukey post hoc tests on univariate ANOVAs of the risk and protective fac-
tors (Table 4) indicate differences between the two groups. Bullies and
bully victims scored similarly on the Risk Factor and Norms scales but dif-
ferently on the two measures of school bonding. These results perhaps
reflect a key difference between the two groups. Although they are similar
in terms of variables that define them as bulliesrisk (favorable attitudes
toward bullying, having friends who bully, and a history of bullying behav-
ior) and perception of school normsthey are different in terms of their
bonding to school, with bullies showing a higher level of commitment and

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


472 Journal of Early Adolescence

attachment to school than bully victims. Bully victims scored lowest


(unhealthiest) on all variables, indicating low bonding to school, higher risk
for bullying (having friends who bully, a history of bullying behavior, and
favorable attitudes toward bullying), and a perception of low expectations
among teachers for student behavior. It is interesting to note that bully-
victim females scored lowest of all groups on school bonding and percep-
tion of school norms, indicating an alienation from both the social and
academic aspects of school. Further research might look more closely at
this particular group of girls.
The second discriminant function differentiated between the victim
group and the comparison, bully victim, and bully groups. High scores on
this grouping of variables appear to measure an orientation to school char-
acterized by commitment to and perception of high adult expectations
regarding academic achievement and rules coupled with low emotional
attachment and a perception that adults in the school do not protect victims
from bullying. Victims appear to feel more comfortable with the academic
aspects and structure of the school environment than with the social and
emotional aspects of fitting in with adults or peers. They seem to be saying,
I fit here academically but not socially. This is compatible with the
research literature that has shown victims to be lonely, socially isolated, and
unpopular with their peers and teachers. Juvonen et al. (2003), for
example, found that victims reported more social anxiety and loneliness
than bullies, bully victims, and students uninvolved in bullying or victim-
ization. Peers rated victims as lowest in social status, and teachers rated
them lowest in popularity among their peers. Victims in this study showed
a strong rejection of attitudes favorable toward bullying, a finding that
makes great sense given their status as victims. This fits with the finding by
Pellegrini and Bartini (2000) that negative attitudes toward bullying were
negatively related to bullying and positively related to victimization.
The second discriminant function showed the greatest differentiation
between the victim and bully groups. Although bullies and bully victims
were similar on risk (friends who bully, favorable attitudes toward bullying,
a history of bullying behavior), bullies scored higher than both victims and
bully victims on attachment. Bullies appear to feel more comfortable in the
social environment of school than either victims or bully victims. This is
consistent with the research that shows that bullies are often popular and
hold high status with their peers (Juvonen et al., 2003; Pellegrini, 2002).
For example, Rodkin, Pearl, and Van Acker (2003) found that some aggres-
sive males were viewed as cool by their peers, particularly by girls.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 473

It is important to consider the nature of the bullys attachment to school


in order to make meaning of these findings. In this study, items on the
Attachment scale did not necessarily indicate an attachment to prosocial
adults and peers and could have been measuring attachment to an antisocial
peer group within the school (e.g., others who bully). Research by McNeely
and Falci (2004) on the relationships among teacher support, social belong-
ing, and health-risk behaviors indicated that school connectedness did not
necessarily reflect prosocial attachment. Their findings suggest that school
connectedness may serve as a protective factor against initiation of health-
risk behaviors when adolescents have a conventional connection (i.e., to
individuals who engage in prosocial behaviors and who regulate prosocial
behavior in others) rather than an unconventional connection (i.e., to others
who engage in behaviors that do not conform to prosocial norms) to school.
McNeely and Falci (2004) point out that when social belonging is an indi-
cator of an unconventional connection to a deviant peer group, as may be
true with the bullies in this study, it becomes a risk factor (rather than a pro-
tective factor) for engaging in health-risk behaviors such as bullying.
An alternate hypothesis is that bullies feel comfortable in the general
school environment despite their more antisocial beliefs. As noted above,
research has shown that bullies are often popular with both their peers and
adults in the school environment. Perhaps this is particularly so when they
perceive that adults are tolerant of their bullying behavior. Bullies in this
study generally scored higher than victims and bully victims and about
equal to the comparison group on Attachment scale items. They indicated
that they feel safe at school, that they feel they belong at school, and that
people at my school care about me. The SDM posits that healthy norms
as well as prosocial bonding must be present to support healthy behaviors
(Catalano et al., 2004; Catalano & Hawkins, 1996). Bullies, victims, and
bully victims in this study were more apt to view adults in the school as
more tolerant of bullying and less apt to intervene to stop bullying than
students in the comparison group. Perception of norms that tolerate bully-
ing may encourage bullies to feel a sense of belonging or attachment to
school while, at the same time, continuing their bullying behavior. Further
research might look more closely at the nature of the bullys attachment to
school and how tolerance of bullying behavior by both adults and peers
influences that attachment.
Limitations of this study include the restricted population of Catholic
school students who were predominantly White and middle class. In addi-
tion, they were enrolled in elementary school and had not made the transi-
tion to middle school characteristic of most public school students. The

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


474 Journal of Early Adolescence

data is cross-sectional and limited to self-report by the student. Given this,


historical antecedents that point to how bullying and bonding may be linked
developmentally cannot be assessed. Some research indicates that status as
a bully remains relatively stable from childhood into adolescence whereas
status as a victim does not (Schfer, Korn, Brodbeck, Wolke, & Schulz,
2005). Schfer et al. (2005) found that victimization in primary school was
generalized across children whereas in secondary school it tended to be
more embedded in peer social hierarchies and focused on fewer children
who were socially rejected by their peers. Their data show a shift in peer
perception of bullying roles as children maturefrom rejection of bullies
in the early grades to rejection of victims during adolescence. This change
in how bullying behaviors are valued as children mature socially supports
research cited earlier about the popularity of bullies and the social difficul-
ties that victims experience during adolescence. Further research might
explore how these changes occur and how bullying and victimization fit
into a broader picture of academic and social success in middle school.
Despite the limitations of the study, the findings are consistent with
other research that has explored differences among bullies, victims, and
bully victims and support research findings about the importance of school
bonding in protecting against the development of problem behaviors in
youth. The results of this study have important implications for schools.
Students who perceived the school environment as supportive, caring, and
promoting prosocial beliefs and behaviors (including both academic and
social competence) were more bonded to school. The comparison group,
those students who reported no or low levels of bullying and victimization,
reported the lowest levels of risk and the most positive perception of the
school environment.
The definition of bonding offered by Catalano et al. (2004), with its two
components of attachment and commitment, proved useful in differentiat-
ing among the bully, victim, bully victim, and comparison groups. Bullies
were high on Attachment and low on Commitment, victims scored low on
Attachment and high on Commitment, bully victims were low on both, and
comparisons scored high on both. These differences point to possible areas
of intervention to increase bonding to school for bullies, victims, and bully
victims. From their research on school bonding, Catalano et al. (2004)
stress the importance of developing social and emotional competence as
supports to academic competence in students. This makes sense in terms of
increasing bonding for bullies, victims, and bully victims. Bullies and bully
victims could benefit from interventions that aim at increasing their com-
mitment to academic and behavioral expectations because they indicate less

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 475

commitment to performing well academically and to following school


rules. Victims, bully victims, and bullies could benefit from interventions
that increase their feelings of social belonging in the mainstream peer cul-
ture. Both victims and bully victims feel a lack of attachment to school,
whereas bullies appear bonded but with attitudes that favor antisocial norms
and behaviors. Interventions that aim at increasing successful interactions
with prosocial peers for all three groups could increase their feelings of
belonging in the social environment of the school and help promote proso-
cial behaviors. Providing skills as well as opportunities to interact success-
fully in both academic and social situations can help bullies, victims, and
bully victims bond more strongly to school.
Developing skills for social and emotional as well as academic compe-
tence for bullies, victims, and bully victims should be done within a compre-
hensive approach to addressing bullying in the school. Research has shown
that bullying occurs in the social context of the school and involves more than
just those who bully and those who are victimized (Salmivalli, Lagerspetz,
Bjrkqvist, sterman, & Kaukiainen, 1996; Sutton & Smith, 1999). Olweus
recommends the adoption of a whole school approach that includes school-
wide and classroom components that attend to the overall ecology of the
school (Olweus, 2003). Creating environments that provide clear and consis-
tent norms for healthy behavior, opportunities for meaningful participation,
skills to participate effectively, and recognition for success can help students
feel physically and emotionally safe in school and support the development
of behaviors that promote academic and social success.

References
Benard, B. (1991). Fostering resiliency in kids: Protective factors in the family, school, and
community. Portland, OR: Northwest Regional Educational Laboratory.
Bernstein, J. Y., & Watson, M. W. (1997). Children who are targets of bullying. Journal of
Interpersonal Violence, 12(4), 483-498.
Bond, L., Carlin, J. B., Thomas, L., Rubin, K., & Patton, G. (2001). Does bullying cause emo-
tional problems? A prospective study of young teenagers. British Medical Journal,
323(7311), 480-484.
Carney, A. G., & Merrell, K. W. (2001). Bullying in schools. School Psychology International,
22(3), 364-382.
Catalano, R. F., Haggerty, K. P., Oesterle, S., Fleming, C. B., & Hawkins, J. D. (2004). The
importance of bonding to school for healthy development: Findings from the Social
Development Research Group. Journal of School Health, 74(7), 252-261.
Catalano, R. F., & Hawkins, J. D. (1996). The social development model: A theory of antiso-
cial behavior. In J. D. Hawkins (Ed.), Delinquency and crime: Current theories (pp. 149-
197). New York: Cambridge University Press.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


476 Journal of Early Adolescence

Catalano, R. F., & Kosterman, R. (1996). Modeling the etiology of substance abuse: A test of
the social development model. Journal of Drug Issues, 26(2), 429-455.
Elsea, M., & Rees, J. (2001). At what age are children likely to be bullied at school?
Aggressive Behavior, 27, 419-429.
Espelage, D. L., & Swearer, S. M. (2003). Research on school bullying: What have we learned
and where do we go from here? School Psychology Review, 32(3), 365-383.
Farrington, D. P. (1991). Childhood aggression and adult violence: Early precursors and later
life outcomes. In D. J. Peplar & K. H. Ruben (Eds.), The development and treatment of
childhood aggression (pp. 5-29). Hillsdale, NJ: Lawrence Erlbaum.
Fox, C. L., & Boulton, M. J. (2005). The social skills problems of victims of bullying: Self,
peer, and teacher perceptions. British Journal of Educational Psychology, 75, 313-328.
Griffin, R. S., & Gross, A. M. (2004). Childhood bullying: Current empirical findings and
future directions for research. Aggression and Violent Behavior, 9(4), 379-400.
Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol
and other drug problems in adolescence and early adulthood: Implications for substance
abuse prevention. Psychological Bulletin, 112, 64-105.
Hawkins, J. D., Guo, J., Hill, K. G., Battin-Pearson, S., & Abbott, R. D. (2001). Long-term
effects of the social development intervention on school bonding trajectories. Applied
Developmental Science, 5(4), 225-236.
Haynie, D. L., Nansel, T., Eitel, P., Crump, A. D., Saylor, K., Yu, K., et al. (2001). Bullies,
victims, and bully victims: Distinct groups of at-risk youth. Journal of Early Adolescence,
21(1), 29-50.
Herrenkohl, T. I., Hill, K. G., Chung, I., Guo, J., Abbott, R. D., & Hawkins, J. D. (2003).
Protective factors against serious violent behavior in adolescence: A prospective study of
aggressive children. Social Work Research, 27(3), 179-191.
Juvonen, J., Graham, S., & Shuster, M. A. (2003). Bullying among young adolescents: The
strong, the weak, and the troubled. Pediatrics, 112(6), 1231-1237.
Kumpulainen, K., Rasanan, E., Henttonen, I., Almqvist, F., Kresanov, K., & Linna, S. (1998).
Bullying and psychiatric symptoms among elementary-age children. Child Abuse and
Neglect, 22, 705-717.
Ladd B., & Ladd, G. W. (2001). Variations in peer victimization: Relations to childrens mal-
adjustment. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of
the vulnerable and the victimized (pp. 25-48). New York: Guilford Press.
Lonczak, H. S., Abbott, R. D., Hawkins, J. D., Kosterman, R., & Catalano, R. F. (2002).
Effects of the Seattle Social Development Project on sexual behavior, pregnancy, birth, and
sexually transmitted disease outcomes. Archives of Pediatrics and Adolescent Medicine,
156(5), 438-447.
Ma, X. (2001). Bullying and being bullied: To what extent are bullies also victims? American
Educational Research Journal, 38(2), 351-370.
Ma, X. (2002). Bullying in middle school: Individual and school characteristics of victims and
offenders. School Effectiveness and School Improvement, 13(1), 63-89.
McConnell, S. R., & Odom, S. L. (1986). Sociometrics: Peer-referenced measures and the
assessment of social competence. In P. S. Strain, M. J. Guralnic, & H. Walker (Eds.),
Childrens social behavior: Development, assessment, and modification (pp. 215-285).
New York: Academic Press.
McNeely, C., & Falci, C. (2004). School connectedness and the transition into and out of
health-risk behavior among adolescents: A comparison of social belonging and teacher
support. Journal of School Health, 74(7), 284-292.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


Cunningham / Level of Bonding to School 477

Nansel, T., Overpeck, M., Pilla, R., Ruan, W., Simons-Morton, B., & Scheidt, P. (2001).
Bullying behavior among U.S. youth: Prevalence and association with psychosocial
adjustment. Journal of the American Medical Association, 285(16), 2094-2100.
Ojala, K., & Nesdale, D. (2004). Bullying and social identity: The effects of group norms and
distinctiveness threat on attitudes towards bullying. Journal of Developmental Psychology,
22, 19-35.
Olweus, D. (1993). Bullying at school: What we know and what we can do. Cambridge, MA:
Blackwell.
Olweus, D. (1994). Bullying at school: Long-term outcomes for the victims and an effective
school-based intervention program. In L. R. Huesman (Ed.), Aggressive behavior: Current
perspectives (pp. 97-130). New York: Plenum.
Olweus, D. (2003). A profile of bullying at school. Educational Leadership, 60(6), 12-17.
Pellegrini, A. D. (2001). A longitudinal study of heterosexual relationships, aggression, and
sexual harassment during the transition from primary school through middle school.
Journal of Applied Developmental Psychology, 22, 9-15.
Pellegrini, A. D. (2002). Bullying, victimization, and sexual harassment during the transition
to middle school. Educational Psychologist, 37(3), 151-163.
Pellegrini, A. D., & Bartini, M. (2000). A longitudinal study of bullying, victimization, and
peer affiliation during the transition from primary school to middle school. American
Educational Research Journal, 37(3), 699-725.
Pellegrini, A. D., Bartini, M., & Brooks, F. (1999). School bullies, victims, and aggressive vic-
tims: Factors relating to group affiliation and victimization in early adolescence. Journal
of Educational Psychology, 91(2), 216-224.
Rigby, K. (2001). Health consequences of bullying and prevention in schools. In J. Juvonen &
S. Graham (Eds.), Peer harassment in schools: The plight of the vulnerable and victimized
(pp. 249-262). New York: Guilford Press.
Rodkin, P. C., Pearl, R., & Van Acker, R. (2003). Theyre cool: Ethnic and peer group supports
for aggressive boys and girls. In A. Rose & A. H. N. Cillessen (Chairs), New perspectives
on peer status: Advances in the study of perceived popularity. Symposium conducted at
the meeting of the Society for Research in Child Development, Tampa, FL.
Salmivalli, C., Lagerspetz, K., Bjrkqvist, K., sterman, K., & Kaukiainen, A. (1996).
Bullying as a group process: Participant roles and their relations to social status within the
group. Aggressive Behavior, 22, 1-15.
Salmivalli, C., & Nieminen, E. (2002). Proactive and reactive aggression among school bul-
lies, victims, and bully-victims. Aggressive Behavior, 28, 30-44.
Salmivalli, C., & Voeten, M. (2004). Connections between attitudes, group norms, and behaviour
in bullying situations. International Journal of Behavioral Development, 28(3), 246-258.
Schfer, M., Korn, S., Brodbeck, F. C., Wolke, D., & Schulz, H. (2005). Bullying roles in
changing contexts: The stability of victim and bully roles from primary to secondary
school. International Journal of Behavioral Development, 29(4), 323-335.
Seals, D., & Young, J. (2003). Bullying and victimization: Prevalence and relationship to
gender, grade level, ethnicity, self-esteem, and depression. Adolescence, 38(152), 735-747.
Sutton, J., & Smith, P. K. (1999). Bullying as a group process: An adaptation of the partici-
pant role approach. Aggressive Behavior, 25, 97-111.
Swearer, S., & Doll, B. (2001). Bullying in schools: An ecological framework. In R. A.
Geffner, M. Loring, & C. Young (Eds.), Bullying behavior: Current issues, research, and
interventions. New York: Haworth.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016


478 Journal of Early Adolescence

Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski, W. (2002). The final report and
findings of the safe school initiative: Implications for the prevention of school attacks in the
United States. Washington, DC: U.S. Secret Service and U.S. Department of Education.
Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski, W. (2000). U.S.S.S. Safe
School Initiative: An interim report on the prevention of targeted violence in schools.
Washington, DC: U.S. Secret Service, National Threat Assessment Center.
Williams, K., Chambers, M., Logan, S., & Robinson, D. (1996). Association of common
health symptoms with bullying in primary school children. British Medical Journal, 313,
17-19.
Wolke, D., Woods, S., Bloomfield, L., & Karstadt, L. (2001). Bullying involvement in primary
school and common health problems. Archives of Diseases in Childhood, 85, 197-201.

Nancy J. Cunningham, PhD, is a professor in the Department of Educational and Counseling


Psychology at the University of Louisville and a licensed counseling psychologist. Her pro-
fessional interests include prevention of mental disorders and problem behaviors, particularly
the prevention of substance abuse and violence in youth and the organization of communities
to prevent community mental health problems. Her present research interests include the study
of early adolescent bullying and victimization, including the development of sexual harass-
ment behaviors in youth.

Downloaded from jea.sagepub.com at PENNSYLVANIA STATE UNIV on September 18, 2016

You might also like