Using social norms to reduce bullying: A research intervention among adolescents in five middle schools

Abstract

Bullying attitudes and behaviors and perceptions of peers were assessed in a case study experiment employing a social norms intervention in five diverse public middle schools in the State of New Jersey (Grades 6 to 8). Data were collected using an anonymous online survey (baseline n = 2,589; postintervention n = 3,024). In the baseline survey, students substantially misperceived peer norms regarding bullying perpetration and support for probullying attitudes. As predicted by social norms theory, they thought bullying perpetration, victimization, and probullying attitudes were far more frequent than was the case. Also as predicted, variation in perceptions of the peer norm for bullying was significantly associated with personal bullying perpetration and attitudes. Using print media posters as the primary communication strategy, an intervention displaying accurate norms from survey results was conducted at each of the five school sites. A pre-/postintervention comparison of results revealed significant reductions overall in perceptions of peer bullying and probullying attitudes while personal bullying of others and victimization were also reduced and support for reporting bullying to adults at school and in one’s family increased. The extent of reductions across school sites was associated with the prevalence and extent of recall of seeing poster messages reporting actual peer norms drawn from the initial survey data. Rates of change in bullying measures were highest (from around 17% to 35%) for the school with the highest message recall by students after a one-and-a-half-year intervention. Results suggest that a social norms intervention may be a promising strategy to help reduce bullying in secondary school populations.

Full text

http://gpi.sagepub.com/
Relations
Group Processes & Intergroup
http://gpi.sagepub.com/content/14/5/703
The online version of this article can be found at:
DOI: 10.1177/1368430210398004
2011 14: 703 originally published online 7 April 2011Group Processes Intergroup Relations
H. Wesley Perkins, David W. Craig and Jessica M. Perkins
in five middle schools
Using social norms to reduce bullying: A research intervention among adolescents
Published by:
http://www.sagepublications.com
can be found at:Group Processes & Intergroup RelationsAdditional services and information for
Immediate free access via SAGE ChoiceOpen Access:
http://gpi.sagepub.com/cgi/alertsEmail Alerts:
http://gpi.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://gpi.sagepub.com/content/14/5/703.refs.htmlCitations:
What is This?
- Apr 7, 2011 OnlineFirst Version of Record
- Apr 8, 2011 OnlineFirst Version of Record
- Sep 1, 2011Version of Record >>
by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from by guest on October 11, 2013gpi.sagepub.comDownloaded from

Group Processes & Intergroup Relations
14(5) 703 –722
© The Author(s) 2011
Reprints and permission: sagepub.
co.uk/journalsPermissions.nav
DOI: 10.1177/1368430210398004
gpir.sagepub.com
Article
G
P
I
R
Group Processes &
Intergroup Relations
Article
398004
14510.1177/1368430210398004Perkins et al.Group Processes & Intergroup Relations
1
Hobart and William Smith Colleges, Geneva, New York
2
Harvard University, Cambridge, Massachusetts
Corresponding author:
H. Wesley Perkins, Hobart and William Smith Colleges, 214
Stern Hall, Pulteney Street, Geneva, New York 14456 USA
Email: perkins@hws.edu
Keywords
bullying, students, middle schools, violence,
social norms, peers, perceptions, misperceptions,
norms, adolescents
Paper received 16 June 2010; revised version
accepted 05 December 2010.
Using social norms to reduce
bullying: A research intervention
among adolescents in five middle
schools
H. Wesley Perkins,
1
David W. Craig
1
and
Jessica M. Perkins
2
Abstract
Bullying attitudes and behaviors and perceptions of peers were assessed in a case study experiment
employing a social norms intervention in five diverse public middle schools in the State of New
Jersey (Grades 6 to 8). Data were collected using an anonymous online survey (baseline n = 2,589;
postintervention n = 3,024). In the baseline survey, students substantially misperceived peer norms
regarding bullying perpetration and support for probullying attitudes. As predicted by social norms
theory, they thought bullying perpetration, victimization, and probullying attitudes were far more
frequent than was the case. Also as predicted, variation in perceptions of the peer norm for bullying
was significantly associated with personal bullying perpetration and attitudes. Using print media posters
as the primary communication strategy, an intervention displaying accurate norms from survey results
was conducted at each of the five school sites. A pre-/postintervention comparison of results revealed
significant reductions overall in perceptions of peer bullying and probullying attitudes while personal
bullying of others and victimization were also reduced and support for reporting bullying to adults at
school and in one’s family increased. The extent of reductions across school sites was associated with
the prevalence and extent of recall of seeing poster messages reporting actual peer norms drawn from
the initial survey data. Rates of change in bullying measures were highest (from around 17% to 35%)
for the school with the highest message recall by students after a one-and-a-half-year intervention.
Results suggest that a social norms intervention may be a promising strategy to help reduce bullying
in secondary school populations.
.

704 Group Processes & Intergroup Relations 14(5)
Introduction
Over the last decade, increased media attention to
the characteristics of school shooting and cyber-
bullying perpetrators and their victims has height-
ened research interest in bullying among
adolescents, particularly in school settings (Dake,
Price, & Telljohann, 2003; Elias & Zins, 2003;
Reuter-Rice, 2008; Srabstein, 2008). In the United
States, a nationally representative survey of youth
in Grades 6 to 12 showed that 9%, 9%, and 3%
were identified as bullies, victims, and both bullies
and victims, respectively, in 2001 (Spriggs, Iannotti,
Nansel, & Haynie, 2007). A 2002–2003 study on
the prevalence of various forms of victimization
in a nationally representative sample of young chil-
dren and adolescents found that emotional teasing
(one form of bullying) occurred among 20% of
the sample (Finkelhor, Ormrod, Turner, & Hamby,
2005). Other studies focusing on different areas
within the United States have shown a similar prev-
alence of bullying ranging from 20 to 30% (Carlyle
& Steinman, 2007; Juvonen, Graham, & Schuster,
2003; Sawyer, Bradshaw, & O’Brennan, 2008).
Given the potential psychosomatic, violent, and
other negative consequences of bullying (Brunstein
Klomek, Marrocco, Kleinman, Schonfeld, &
Gould, 2007; Klomek et al., 2008, 2009; Lund
et al., 2009; Nansel, Craig, Overpeck, Saluja, &
Ruan, 2004; Nansel et al., 2001; Nansel, Overpeck,
Haynie, Ruan, & Scheidt, 2003; Salmon, James, &
Smith, 1998; Sourander, Helstela, Helenius, &
Piha, 2000; Srabstein & Piazza, 2008), understand-
ing why some people are at risk of either bullying
perpetration or victimization is salient. Numerous
studies and reviews have shown many individual,
family, peer/social, community, and school risk
factors that contribute to bullying and youth vio-
lence such as low IQ, antisocial attitudes, weight
status, substance use, television viewing, exposure
to family violence, low parental involvement, poor
family functioning, social rejection by peers, poor
academic performance, diminished economic
opportunities, socially disorganized neighbor-
hoods, school social environment, school size, and
school policy (Bowes et al., 2009; Center for
Disease Control and Prevention [CDC], 2008;
Department of Health and Human Services
[DHHS], 2001; Janssen, Craig, Boyce, & Pickett,
2004; Johnson, 2009; Kuntsche, Knibbe, Engels, &
Gmel, 2007; Kuntsche et al., 2006; Lipsey &
Derzon, 1998; Resnick, Ireland, & Borowsky,
2004; Smith & Myron-Wilson, 1998; Spriggs et al.,
2007). Although several reports on youth violence
(American Psychological Association [APA], 1996;
DHHS, 2001; Hahn et al., 2007; Murray, Guerra, &
Williams, 1997; Srabstein et al., 2008) have signaled
the necessity of developing effective prevention
programs, many of the aforementioned studies do
not fully identify the mechanisms explaining why
youth may engage in bullying, knowledge that
would help to devise effective prevention.
Conformity to peer norms
Although sociodemographic and contextual fac-
tors represent an important consideration when
attempting to predict and prevent bullying perpe-
tration and victimization, another set of potential
risk factors—peer norms and the perception of
peer norms—deserves special attention. Decades
of research in social psychology going all the way
back to the classic experiments of Solomon Asch
(Asch, 1956) and Musafer Sherif (Sherif, 1936,
1937) have demonstrated the strong tendency of
people to conform to peer norms as they look to
others in their midst to help define the situation
and give guidance on expected behaviors in the
group or cultural setting. Although many people,
and especially adolescents, frequently think of
themselves as individuals in their actions, a con-
siderable degree of peer influence is consistently
documented in laboratory experiments, social
surveys, and observations of crowd behavior. In
studies on antecedents of personal health-related
behaviors, for example, extensive evidence has
supported the theory of reasoned action (Ajzen
& Fishbein, 1980) and its extension, the theory
of planned behavior, which posits norms as a
determinant of personal behavior along with per-
sonal attitudes and perceived behavioral control
(Ajzen, 2001, 2002; Ajzen & Madden, 1986).
Furthermore, research on adolescents’ health and
well-being has singled out peer influence as

Perkins et al. 705
crucial in regard to risk behaviors such as alcohol,
tobacco, and other drug use. Bullying and victimi-
zation in schools are inherently relational proc-
esses, relying on domination, subjugation, and
bystander apathy, all presumably shaped by peer
norms. This type of violence is a demonstration
of “peer group power” in which a whole peer
group participates in the bullying with individuals
fulfilling different roles and acting as moderators
of such behavior (Salmivalli, 1999).
Often, bullying occurs in academic settings, not
only because adolescents spend a significant por-
tion of every day in school, but also because
schools are such peer intensive social environments
where behaviors such as who sits with whom in the
lunchroom are rigidly defined by student norms
and pervasively communicated in the ways stu-
dents talk (or not talk) to each other (Eder, Evans,
& Parker, 1995). Thus, students form impressions,
be they correct or incorrect, about what is going on
in the school environment and who is involved in
peer social interaction from a context where peer
talk frequently dominates the milieu. In turn, these
impressions may lead students to participate in bul-
lying, to acquiesce to victimization, or to remain as
bystanders to the bullying of others.
Thus, widely shared practices or behaviors
(descriptive norms) and widely shared beliefs or
common attitudes (injunctive norms) serve as
social cues directing and constraining individuals’
behaviors and attitudes in educational environ-
ments at various stages of development. For
example, among 1,368 female sixth graders,
friends’ bullying perpetration or victimization was
associated with personal bully/victim status
(Mouttapa, Valente, Gallaher, Rohrbach, & Unger,
2004). Among college students, peer group ideo-
logical beliefs predicted individual members prej-
udiced attitudes (Poteat & Spanierman, 2010).
Misperceived norms and the
social norms approach to
reducing problem behavior
Since its introduction in an initial study of univer-
sity student drinking (H. W. Perkins & Berkowitz,
1986), the examination of the degree of discrep-
ancy between actual and perceived norms as well
as the potential influence of both has received a
great deal of theoretical and empirical examina-
tion as applied to adolescent and young-adult
consumption of alcohol, tobacco, and other
drugs (H. W. Perkins, 2003a). Indeed, a consistent
and dramatic pattern of misperceptions about
peer norms for substance use has been docu-
mented in studies conducted in several nations
(Hughes, Julian, Richman, Mason, & Long, 2008;
Kilmer et al., 2006; Linkenbach & Perkins, 2003;
Lintonen & Konu, 2004; McAlaney & McMahon,
2007; Page, Ihasz, Hantiu, Simonek, & Klarova,
2008; Page, Ihasz, Simonek, Klarova, & Hantiu,
2006; H. W. Perkins, 2007; H. W. Perkins & Craig,
2003; H. W. Perkins, Haines, & Rice, 2005;
H. W. Perkins, Meilman, Leichliter, Cashin, &
Presley, 1999) where the tendency is to overesti-
mate the permissiveness of peers and the extent
or prevalence of use, even in peer contexts where
use is relatively high. Similarly, adolescent and
young adult misperceptions of norms have been
identified for other concerns regarding health
and well-being including body weight and image
(Clemens, Thombs, Olds, & Gordon, 2008; J. M.
Perkins, Perkins, & Craig, 2010a), consumption
of sugar-sweetened drinks (J. M. Perkins, Perkins,
& Craig, 2010b), violence against women
(Fabiano, Perkins, Berkowitz, Linkenbach, &
Stark, 2003; Neighbors et al., 2010), and sexual
behavior (Martens et al., 2006).
Thus, the various studies consistently show
that positive attitudes and behaviors, though
most often the norm among young people in
schools and communities, are often not perceived
to be the peer norm. Adolescents and young
adults tend to believe that risky or problem
behaviors and attitudes are most common among
peers and think protective responsible action is
rare. These exaggerated or erroneous perceptions
may be the result of (a) attribution error where
behavior occasionally observed in others is
thought to be typical of them when only incom-
plete or superficial information about peers is
available, (b) social conversation among youth
about the most extreme behavior in their midst

706 Group Processes & Intergroup Relations 14(5)
getting disproportionate attention, thus creating a
sense that the extreme behavior is common, and
(c) entertainment and news media further ampli-
fying misperceptions by focusing almost entirely
on images and stories of the risky or problem
behavior (H. W. Perkins, 1997, 2002, 2003a).
It is argued that these misperceptions then
contribute to or exacerbate the problem behavior
as more youth begin to support and engage in the
behavior than would otherwise be the case if
norms were accurately perceived. Amidst these
widely held misperceptions of problem behavior
as “normal” among peers, those who regularly
engage in the problem behavior freely do so
thinking they are just like most others and are
likely to have the greatest commitment to the
misperception. Those who are ambivalent about
joining in the behavior, nonetheless, by misper-
ceiving the norm, may occasionally do so mistak-
enly feeling a false majority pressure. Finally, most
of those who oppose the behavior (the real
majority) remain silent as bystanders to the prob-
lem behavior believing that they, as bystanders,
are alone in their opposition. They may hold less
extreme misperceptions of the problem as the
norm and thus feel least pressured to actually
engage in the behavior. However, the mispercep-
tion that does exist among them is still harmful as
it spawns apathy and withdrawal from interaction
with peers (H. W. Perkins, 2007).
Thus, the strategy of the social norms approach
to preventing problem behavior, put simply, is to
dispel the myths about the problem being the
norm among peers. Social norms interventions
seek to turn the process around by intensively
communicating the truth about positive norms
based on credible data drawn from the target
population. In short, social norms theory
(H. W. Perkins, 1997, 2003b) predicts that by
reducing misperceptions and increasing the pro-
portion of students with more accurate informa-
tion about existing healthy norms, occurrences of
unhealthy or problem behavior will decrease.
Several intervention studies regarding alcohol,
tobacco, and other drug use have shown that when
students are intensively exposed to actual norms,
their misperceptions and actual problem behavior
can be reduced (Bewick, Trusler, Mulhern,
Barkham, & Hill, 2008; DeJong et al., 2006; Haines
& Spear, 1996; Haines, Barker, & Rice, 2003;
Hansen & Graham, 1991; Linkenbach & Perkins,
2003; Mattern & Neighbors, 2004; Neighbors,
Larimer, & Lewis, 2004; H. W. Perkins & Craig,
2006; Turner, Perkins, & Bauerle, 2008).
Interventions using social norms feedback about
peer and community attitudes and behavior for
other topic areas such as conservation and recy-
cling have demonstrated positive effects as well
(Cialdini, Reno, & Kallgren, 1990; Schultz, 1999;
Schultz, Khazian, & Zaleski, 2008).
Although limited research has examined the
relationship between bullying norms and personal
involvement in bullying, and some studies have
found norms to be important predictors of other
health-related behavior among adolescents, no
studies have examined the accuracy of students’
perceptions of bullying norms (personal percep-
tion of the bullying norm in a given group versus
the actual extent of bullying behavior and attitu-
dinal support for it in the group). Furthermore,
no study has reported an intervention to chal-
lenge misperceptions as a means to reduce bully-
ing. Thus, the current study introduces research
examining three important questions related to
perceived norms of bullying. Specifically, we con-
sider: (a) the extent and direction of mispercep-
tions about bullying as well as how much variation
in perceived norms exists, (b) the degree of asso-
ciation between perceptions of the peer bullying
norm and personal involvement in bullying, and
(c) the impact that might be produced by dissemi-
nating actual norms about bullying in adolescent
populations. Thus, the objective of the study was
to address these three questions with action-ori-
ented research on bullying conducted in five mid-
dle-school populations.
Method
Participants
Students in five middle schools located through-
out the State of New Jersey in the United States
provided the data for this research. Each school

Perkins et al. 707
contained Grades 6 through 8 and almost all stu-
dents were between the ages of 11 and 14. The five
schools were from an initial group of seven middle
schools in the state that had chosen to participate
in an online survey of their students regarding bul-
lying in late spring of 2006. The five sites provid-
ing data for this study were all of the schools from
the initial survey group that fulfilled the following
criteria: (a) the entire school population served as
the sampling frame for the survey; (b) the school
subsequently conducted an intervention to chal-
lenge misperceptions with data-based messages
about actual peer norms in the local school; (c) the
intervention campaign included at least the posting
of print media in the school with messages that
had been created by the research team and dis-
played with supporting images created or approved
by the researchers (additional communication
venues were also used by local schools in some
instances); (d) the same survey of bullying was
again administered as a postintervention assess-
ment with all students as the sampling frame; and
(e) demographically comparable pre- and postin-
tervention samples were obtained from the school
as a result of the surveys. The two other schools
that participated initially were excluded from the
study because their response rates for the baseline
survey were very low (17% and 22%) and resulted
in samples that did not adequately represent the
school populations. No schools conducted the
survey at two time points without conducting
the intervention so no overtime control compari-
son sites were available. Thus, this study provides
five case studies of the intervention based on rep-
resentative cross-sectional data collected at each
school site at pre- and postintervention time
points.
Four schools in this study were very large mid-
dle schools (populations between 900 and 1,300
students) and one was midsize (300–400 stu-
dents). Three were located in suburban settings,
one was in a combined urban and suburban area,
and one was rural. Three schools were largely
homogeneous in racial composition (85% or
higher White) and two schools reflected substan-
tial diversity (about 50% minority races). The
average response rate across schools from the
school populations was 59%. Table 1 provides
the pre- and postintervention respondent charac-
teristics for each school.
Survey procedures
Data were collected using the “Survey of Bullying
at Your School” (Social Norms Surveys Online).
The Institutional Review Board of the academic
institution hosting the online survey approved
the survey procedures and local schools obtained
parental consent for student participation. From
class sessions or other group assignments in
school, students who had parental consent were
instructed to go in groups to rooms where a com-
puter was available for each student. Each group
was given general information about the online
survey and told that the survey was voluntary and
anonymous. A student could leave all questions
Table 1. Pre- and postintervention sample demographics for five school sites
School A
a
School B
a
School C
a
School D
b
School E
b
Pre Post Pre Post Pre Post Pre Post Pre Post
N 180 225 759 681 578 799 484 592 588 727
Response rate (%) 50 70 80 71 47 69 43 50 50 61
Female (%) 58 56 53 53 55 52 53 50 50 52
Mean age 12.5 12.3 12.8 12.3 12.7 12.4 12.8 12.5 12.6 12.5
(SD) (.9) (1.0) (1.0) (1.0) (.9) (1.0) (.9) (1.0) (1.0) (1.0)
Minority (%) 27 24 21 25 29 29 48 57 57 59
Note:
a
Schools with 1.5 academic-year intervention;
b
Schools with 1.0 academic-year intervention.

708 Group Processes & Intergroup Relations 14(5)
blank if they did not want to participate. No per-
sonal computing accounts were used. To access
the survey all students in a specific group session
were publicly given the same password and URL
address in order to assure students of their ano-
nymity in completing the survey. However, the
password was changed between sessions so that
no student could access the survey and submit
multiple responses after leaving his or her survey
session. There was a teacher or other adult moni-
tor present simply to make sure that students did
not speak with each other while taking the survey.
The survey data were subsequently checked to
screen out submissions with intentionally pro-
vided erroneous or random answers. The small
number of respondents who submitted multiple
answers that were clearly outside of possible
ranges or who answered sets of questions with
contradictory responses was eliminated.
Measures
Bullying perpetration The survey instrument
included a series of questions about what are
commonly identified as bullying behaviors in
schools including: (a) pushing, shoving, hitting,
kicking, hair pulling, or tripping; (b) teasing in
an unfriendly way; (c) calling hurtful names;
(d) excluding someone from a group to make
them feel bad; (e) taking or damaging someone
else’s belongings; (f) spreading unkind stories or
rumors about someone else; (g) threatening to
hurt someone; and (h) making someone do some-
thing they did not want to do. Specifically,
respondents were asked how often in the last 30
days they had done each of these eight behaviors
to another student using the response categories
of “Not in the last 30 days” (coded 0), “Once”
(coded 1), “2–3 times” (coded 2), and “4 or more
times” (coded 3). We refer to these behaviors as
personal bullying behaviors. An index measure of
personal bullying perpetration was subsequently
computed by summing scores for responses to all
eight items. Scale reliability analyses (Cronbach’s
alpha) indicated high inter-item reliability for
responses in both the preintervention (.82) and
postintervention (.83) surveys.
Using the same behavioral items and response
categories, respondents were also asked how
often they thought most other students had done
these things at their school. We refer to these
responses as perceived peer norms for bullying
behaviors. An index measure of the perceived
norm for bullying perpetration was subsequently
computed by summing scores of responses to all
eight items. Scale reliability analyses (Cronbach’s
alpha) indicated high inter-item reliability for the
perceived norm measure in both the preinterven-
tion (.91) and postintervention (.91) surveys.
Bullying victimization Seven items comprised
measures of personal bullying victimization and
the perceived norm for bullying victimization that
paralleled the first seven items of the perpetration
measures. Respondents were asked how often in
the last 30 days each of the following things had
happened to them, and also how often in the same
30 days they thought these things had happened
to most other students at their school: (a) being
pushed, shoved, hit, kicked, hair pulled, or tripped;
(b) being teased in an unfriendly way; (c) being
called hurtful names; (d) being excluded from a
group to hurt feelings; (e) belongings being taken
or damaged; (f) unkind story or rumor spread; and
(g) threatened to be hurt. The eighth item in the
list of perpetration measures—making someone
do something they did not want to do—was not
converted to a victimization item in the construc-
tion of the survey because it was judged that
“being made to do something you did not want to
do” might be confused with being required to per-
form legitimate or positive behaviors under the
direction of peers or others at school. Thus, the
survey only presented seven victimization items.
The same response categories and scores as used
in the bullying perpetration questions were
employed, and again, indices were created by sum-
ming responses to the items. Scale reliability analyses
(Cronbach’s alpha) also indicated high inter-item
reliability for personal bullying victimization in
both the preintervention (.82) and postintervention
(.81) surveys and for the perceived norm for
bullying victimization at both times (.86 and .88).

Perkins et al. 709
Probullying attitudes A second set of ques-
tions measured personal pro-bullying attitudes
and the perceived norm for pro-bullying attitudes
by providing four statements and asking respond-
ents to what extent they agreed or disagreed and
to what extent they thought most other students
agreed or disagreed. Statements were as follows:
(a) “Students should NOT tease in a mean way,
call others hurtful names, or spread unkind sto-
ries about other students”; (b) “Students should
NOT shove, kick, hit, trip, or hair pull another
student”; (c) “Students should NOT threaten to
hit another student even if they don’t actually hit
the other student”; (d) “Students should always
try to be friendly with students who are different
from themselves.” Response categories for per-
sonal beliefs and for what respondents thought
most others would say were strongly agree, agree,
disagree, and strongly disagree coded 0 to 3,
respectively. Indices for personal attitude and the
perceived norm were created by summing the
response scores for all four items. Scale reliability
analyses (Cronbach’s alpha) indicated high
inter-item reliability for personal probullying
attitudes (.82 and .84 for pre- and postsurveys,
respectively) and for the perceived norm for
pro-bullying attitudes (.82 at each survey time).
Reporting bullying In a third set of questions
students were asked: “Who do you think students
should tell if they or someone else are being bul-
lied at school? And what would most other stu-
dents say?” Respondents could indicate any of
several different types of people for their per-
sonal opinion with a separate listing of the same
types for their perceptions about what would be
most others’ response (perceived peer norm). Of
specific interest for the intervention and for
assessment in this study were the three categories:
(a) principal, (b) teacher or counselor, and (c) par-
ent or other adult relative.
Poster campaign message exposure One
final measure used in this study was drawn from a
set of questions added to the postintervention
survey at the end of the survey instrument. The
questions asked how often and where during the
school year respondents had seen or heard infor-
mation about what most students or the majority
do or think about bullying and unfriendly behav-
iors based on survey results from students at their
school. Given the nature of the intervention
described below, the item focusing on how often,
if ever, students had seen such material on posters
at school was of specific interest for this study.
Response categories were “never,” “once,” and
“more than once” (scored 0, 1, or 2, respectively).
Intervention
The basic strategy of the social norms interven-
tion was to provide students in each local school
with feedback about the results of the initial sur-
vey by conveying actual positive norms, which
were widely misperceived. Social norm messages
about the prevalence of positive behavior and
opinion were created for each school based on
their data. Messages indicated the recent survey
on bullying at their school as the source of infor-
mation and noted the large number of students
participating. Examples of norm messages
(school name deleted) are:
Most ____ Middle School students (9 out of
10) agree that students should always try to be
friendly with students who are different from
themselves.
95% of ____ Middle School students say stu-
dents should NOT tease in a mean way, call
others hurtful names, or spread unkind stories
about other students.
94% of ____ Middle School students believe
students should NOT shove, kick, hit, trip, or
hair pull another student.
9 out of 10 ____ Middle School students
agree that students should NOT threaten to
hit another student even if they don’t actually
hit the other student.
Most ____ Middle School students (3 out of
4) do NOT exclude someone from a group to
make them feel bad.

710 Group Processes & Intergroup Relations 14(5)
Most ____ School students (9 out of 10) do
NOT take or damage other’s belongings.
Most ____ Middle School students (8 out of
10) think that students should tell a teacher or
counselor if they or someone else are being
bullied at school.
94% of ____ students say they are encour-
aged to help and respect other students.
Most (4 out of 5) ____ students do NOT
spread unkind rumors or stories about other
students.
2 out of 3 ____ students think that students
should tell a parent or other adult relative if
they or someone else are being bullied at school.
7 out of 10 ____ students do NOT get
involved in any pushing, shoving, kicking,
pulling hair or tripping any other students.
Poster images containing these messages did not
display the negative behavior. Rather, they pre-
sented scenes of positive student interaction or
simply scenes or emblems associated with the
local school. Figures 1 and 2 provide examples
(with actual school names and survey dates
changed or removed). These posters were printed
as large (3 by 4 feet) wall posters and as smaller
posters for display in the school. In three schools,
the campaign was carried out over one and one
half academic years before the postintervention
assessment and in two schools the campaign ran
for one academic year before the postinterven-
tion assessment.
Analytic approach
The first hypothesis we examined was that stu-
dents tend to overestimate the prevalence of and
Figure 1. Examples of social norms intervention posters.

Perkins et al. 711
support for bullying behaviors among peers. We
assessed this initial assumption by comparing the
mean reported personal bullying perpetration,
bullying victimization, and probullying attitudes
with the corresponding perceived peer-norm
mean for perpetration, victimization, and atti-
tudes at each school using a paired sample t test
of significance for each set of measures in each
preintervention school sample. We then tested
our second hypothesis that students’ personal
behaviors and attitudes are associated with what
they perceive to be the norm. That is, even though
most individuals may overestimate problem
behaviors and attitudes, we further hypothesized
that those who tend to think bullying behaviors
and attitudes are more pervasive than not will
also be the ones who engage in more of the
actual behavior and more often believe it is
acceptable. To test this assumption we calculated
the correlation (Pearson r) between the perceived
norm and personal behavior or attitude in each
initial sample.
Our third hypothesis was that an intervention
providing students with information about actual
norms to challenge misperceptions of the peer
norm would reduce misperceptions (i.e., lower
students’ estimates of the prevalence and support
for bullying behavior and raise estimates of peer
support for reporting bullying to principals, teach-
ers, and parents), and, in turn, reduce the actual
levels of bullying and support for bullying and
raise levels of personal support for reporting bul-
lying. This prediction was tested for each school
site by (a) comparing the pre- and postinterven-
tion levels on each of the perception of peer
measures of this study, and (b) comparing the pre-
and postintervention levels on each of the per-
sonal measures. For each of these comparisons,
we conducted an independent sample t test for a
significant pre/postintervention difference in
means or proportions in the predicted direction.
It is important to note that although the social
norms intervention model predicts that percep-
tions of peers can be shifted in the direction of
the actual norms by providing accurate normative
information, it does not necessarily predict that
the gap between perceived and actual norms will
be lessened. Rather, the model predicts that
Figure 2. Example of social norms intervention poster.

712 Group Processes & Intergroup Relations 14(5)
changing what is perceived as normative will lead
to a change in personal attitudes and behaviors.
Thus, an equally large gap may still exist in the
wake of an effective intervention if both percep-
tions of norms and personal attitudes or behav-
iors have shifted toward the actual and
subsequently more positive norm. Assessing the
pre/post differences in perception and in per-
sonal attitude or behavior, and not a change in the
gap between perceived peer and personal attitudes
or behaviors, is the critical measure of impact.
Furthermore, it is most appropriate to use an
independent samples test of the difference of
pre/post means (proportions) in this instance
(and not a paired sample test). First, the data were
collected anonymously so there are no links
between the pre- and postintervention respond-
ent records. Second, the samples measured at
baseline preintervention represented mostly dif-
ferent respondents than the respondents in the
postintervention survey. The postintervention
survey, which took place 12 to 18 months later,
was intended to measure the impact of the inter-
vention on perceptions and experiences of bully-
ing for all grade levels in the school including new
cohorts that moved into the middle school after
such an intervention was underway. Thus, one or
two new grade cohorts had moved in (and out) of
each school by the time the postsurvey was given.
Therefore, posttest samples inherently included
from one third to two thirds different respond-
ents who were new in each school. Moreover,
some additional students move in and out of the
school district every year as families move. Finally,
each year a significant portion of the student
body did not participate (response rates reported
in Table 1) due to absences on the survey days or
parental permission slips for participation not
being available. The net result is that only a rela-
tively small percentage of students were the same
in the pre- and postsamples, and thus very little
distortion is created in using the independent
samples t test. Also, it is important to note that
the significance test is fundamentally conserva-
tive in identifying true Time 1/Time 2 differences
in the population because the sampling frame
used was the entire population itself, not a ran-
dom sample drawn from the population.
An additional prediction associated with our
third hypothesis was that the degree of success
of a social norms intervention is based on achiev-
ing widespread and intensive exposure to the
campaign messages. Thus, variation among
school sites in the prevalences of recalling multi-
ple exposures and no exposure to social norms
poster media about bullying at school were exam-
ined. The mean exposure level for each school
was calculated and then correlated (Pearson r)
with the mean rate of pre/post intervention
change in all bullying measures at each site.
Results
Baseline findings
Table 2 presents the initial mean scores in all five
school samples (preintervention surveys) for per-
sonal bullying perpetration compared to the per-
ceived norm for bullying perpetration, for personal
bullying victimization compared to the perceived
norm for bullying victimization, and for personal
probullying attitudes compared to the perceived
norm for probullying attitudes. Here, the mean of
each personal measure provides an estimate of the
actual norm existing in each school based on the
sample. The mean of each perceived-norm index
provides an estimate of students’ average subjec-
tive perception of how much bullying perpetra-
tion, victimization, and attitudinal support for
bullying is the peer norm. Perceptions of the peer
norm for bullying behavior are three to four times
higher than the estimates of the actual norm
based on the aggregate of personal behaviors in
each sample. Perceived levels of victimization are
more than twice as large as what are found in the
anonymous personal reports, and peers are per-
ceived to be about twice as supportive of probul-
lying attitudes as what is actually found among
students at each school. Statistically significant
differences (p < .001) were found in the predicted
direction in every instance of comparing means in
each set of measures at each site.

Perkins et al. 713
Table 3 presents correlation coefficients
examining the association between students’ per-
sonal bullying perpetration and their perceived
norm for bullying perpetration as well as the
association between the perceived norm for
probullying attitudes and personal probullying
attitude. Clearly one’s personal behavior and atti-
tudes regarding bullying are highly linked to how
commonplace one thinks such behaviors are and
how much support one believes exists for these
actions among peers, regardless of the accuracy
(or more often inaccuracy) of these perceptions.
All correlations are positive as predicted, ranging
from .33 to .56 in strength, and all are statistically
significant at p < .001.
Pre/postintervention comparisons
Table 4 presents the results comparing pre- and
postintervention data on all perceived norm and
personal measures used in this study (all bullying
indices as well as the measures of respondents’
perceived norms and personal attitudes regard-
ing reporting bullying perpetration). School A
demonstrated significant change in the pre-
dicted direction on all measures. There was less
perception of bullying perpetration and victimi-
zation, less personal reporting of being a perpe-
trator or victim, and less personal and perceived
peer support for bullying and more personal and
perceived peer willingness to report bullying to
Table 2. Preintervention bullying perpetration, victimization, and attitude norms compared to perceived peer
norms by school site
School A School B School C School D School E
BULLYING PERPETRATION
Personal bullying perpetration mean (SD) 2.3 (2.8) 2.9 (3.8) 2.7 (3.7) 3.0 (3.6) 2.4 (3.3)
Perceived norm for bullying perpetration
mean (SD)
10.6 (5.7) 10.7 (6.6) 13.0 (6.8) 11.5 (6.3) 9.4 (6.2)
BULLYING VICTIMIZATION
Personal bullying victimization mean (SD) 5.0 (4.6) 4.4 (4.4) 4.7 (5.0) 4.5 (4.3) 4.1 (4.2)
Perceived norm for bullying victimization
mean (SD)
10.7 (4.9) 10.8 (5.4) 12.7 (5.6) 11.4 (5.1) 10.2 (5.2)
PROBULLYING ATTITUDES
Personal bullying attitudes mean (SD) 1.7 (2.0) 1.8 (2.2) 1.8 (2.2) 1.8 (2.1) 1.7 (2.0)
Perceived norm for bullying attitudes
mean (SD)
3.3 (2.1) 3.5 (2.6) 4.5 (2.7) 3.9 (2.4) 3.5 (2.2)
Note: All personal index means are significantly different from the corresponding perception index means at p < .001.
Table 3. Preintervention correlations of personal bullying perpetration and attitudes with the corresponding
perceived peer norm index by school site
School A School B School C School D School E
BULLYING PERPETRATION
Personal bullying perpetration by 0.356 0.407 0.330 0.412 0.474
perceived norm for bullying perpetration
PROBULLYING ATTITUDES
Personal bullying attitudes by 0.465 0.563 0.395 0.503 0.529
perceived norm for bullying attitudes
Note: All correlations are significant at p < .001.

714 Group Processes & Intergroup Relations 14(5)
Table 4. Pre- and postintervention perceived norms and personal bullying perpetration, victimization, probullying attitudes, and beliefs about reporting by
school site
School A School B School C School D School E
Pre Post Rate
(%)
change
Pre Post Rate
(%)
change
Pre Post Rate
(%)
change
Pre Post Rate
(%)
change
Pre Post Rate
(%)
change
BULLYING PERPETRATION
Perceived norm for bullying
perpetration mean
10.6 8.1 −24*** 10.7 8.4 −21*** 13.0 10.7 −18*** 11.5 9.2 −20*** 9.4 9.4 0
Personal bullying perpetration
mean
2.3 1.5 −35** 2.9 2.5 −14* 2.7 2.4 −11 3.0 2.0 −33*** 2.4 2.3 −4
BULLYING VICTIMIZATION
Perceived norm for bullying
victimization mean
10.7 8.8 −18*** 10.8 8.9 −18*** 12.7 10.7 −16*** 11.4 9.7 −15*** 10.2 10.0 −2
Personal bullying victimization
mean
5.0 3.3 −34*** 4.4 4.0 −9 4.7 4.2 −11* 4.5 4.0 −11* 4.1 3.7 −10
PROBULLYING ATTITUDES
Perceived norm for bullying
attitudes mean
3.3 2.4 −27*** 3.5 3.0 −14** 4.5 3.4 −24*** 3.9 3.2 −18*** 3.5 3.2 −9*
Personal bullying attitudes mean 1.7 1.1 −35** 1.8 1.5 −17* 1.8 1.7 −6 1.8 1.4 −22*** 1.7 1.5 −12
REPORTING BULLYING
Perceived most peers would say
students should tell _____ if they
or others are being bullied (%)
A principal 46 60 30** 36 48 33*** 39 48 23*** 31 37 19* 30 36 20**
A teacher or counselor 59 69 17* 54 63 17*** 48 54 13* 57 63 11* 51 61 20***
A parent/other adult relative 52 61 17* 48 54 13*** 41 48 17** 48 53 10 38 48 26***
Personally thinks students
should tell _____ if they or
others are being bullied (%)
A principal 55 73 33*** 40 53 33*** 58 67 16*** 37 43 16* 44 44 0
A teacher or counselor 77 89 16** 72 79 10*** 74 80 8* 81 83 2 78 80 3
A parent/other adult relative 71 81 14* 68 73 7* 68 76 12** 73 75 3 62 69 11**
Note: *p < .05; **p < .01: ***p < .001 in the predicted direction.

Perkins et al. 715
principals, teachers and counselors, and adult
relatives. Rates of change ranged from 17% to
35% in the expected direction. Postintervention
samples from schools B, C, and D, likewise, dis-
played change as predicted on all 12 pre/
postcomparisons with 11, 10, and 9 of them
producing statistically significant results, respec-
tively. Rates of change on the significant items
ranged between 7% and 33% in the expected
direction. Finally, School E demonstrated the
least change after the intervention; only 5 of the
12 measures showed a statistically significant
difference in the predicted direction and two
items showed no change. Nevertheless, observed
differences between the pre- and postinterven-
tion samples of School E remained in the pre-
dicted direction and there were appreciable rates
of change (between 9% and 26%) in the
expected direction on the five statistically sig-
nificant measures for this school with the weak-
est impact. Thus, overall, four schools provided
strong support for the intervention having a
positive impact while one school showed a more
mixed or weaker positive result.
Exposure to poster message intervention
Figure 3 presents the prevalence of respondents
recalling having seen multiple poster messages at
school about what the majority of peers think
and do regarding bullying based on survey data
and the prevalence of respondents not recalling
seeing any poster with this type of message.
Prevalence of exposure and lack of exposure to
the poster campaign is broken down by school.
The school achieving the greatest postinterven-
tion change in the expected direction with signifi-
cant results on all bullying items (School A) is also
the school demonstrating the highest exposure
level to the poster campaign with 72% reporting
multiple exposures and only 13% reporting no
recall. Schools B, C, and D with 11, 10, and 9
measures demonstrating significant pre/post-
change, respectively, exhibited multiple exposures
for 66%, 62%, and 62% of their students, respec-
tively, and no recall for 20%, 20%, and 23%,
respectively. Finally, the school that revealed the
least change in perceptions of norms and per-
sonal attitudes and behaviors regarding bullying
Figure 3. Recollection of seeing posters reporting survey results during the school year by intervention site.

716 Group Processes & Intergroup Relations 14(5)
with significant differences on just five measures
(School E), also demonstrated the lowest cam-
paign exposure level. Only about half (52%)
recalled seeing multiple poster messages and
almost one third (30%) recalled no messages.
As a further assessment of the relationship
between poster message exposure levels achieved
at each school and pre/postchange in school bul-
lying climate, the mean score on the exposure
question (scored as 0 for never recalling seeing a
poster, 1 for once, and 2 for recalling seeing two
or more poster messages) at each school was
compared with the mean percentage pre/post-
intervention rate of change in all attitudinal and
behavioral measures of perceived and personal
bullying and victimization. That is, the average of
absolute values of all percentage rates of change
reported in Table 4 was calculated for each
school. Absolute values of rates of change were
used because declines were expected based on
item wording and scoring for the measures of
bullying perpetration, victimization, and probul-
lying attitudes whereas increases were expected
for the measures of attitudes about reporting bul-
lying to school officials and familial adults.
Absolute values could be used for this computa-
tion because all observed change at each school
was in the predicted direction. Table 5 reports
these means by school along with the correlation
between mean exposure and mean rate of change.
Near perfect association is demonstrated here
with a Pearson r of .96, which is highly significant
in the expected direction at p < .005 even taking
into consideration the small number of schools
in the correlation.
Thus, overall school exposure levels to the
campaign were strongly related to the overall
change experienced at the school. More social
norm message exposure was associated with
greater change in the school in an almost perfect
correlation. This pattern of results provides
further evidence supporting the claim that the
intervention to reduce misperceptions about
bullying was the crucial factor accounting for the
postintervention change in bullying.
Discussion
This study expands the realm of research on mis-
perceptions of peer norms among youth to the
phenomenon of bullying in middle schools.
Without doubt, participation in bullying behav-
iors is a serious problem in schools. However, this
research finds that middle school students grossly
Table 5. Mean exposure to poster media and mean percentage pre-/postintervention rate of change in all
bullying measures by school site
School site Mean exposure to poster media
a
Mean percentage rate of change in
all bullying measures
b
A 1.59 25.0
B 1.46 17.1
C 1.42 14.4
D 1.40 15.1
E 1.22 9.7
Pearson r .96
c
Note:
a
Mean score was computed from respondent recall of the frequency of seeing posters displaying survey information
about what most students think and do about bullying during the school year where 0 = never, 1 = once, and 2 = 2 or more
times.
b
Mean percentage pre/postintervention rate of change was computed from the absolute values of the rates of change
in all survey measures of perceived and personal bullying as reported in Table 4. Absolute values of rates of change were used
because declines were expected for the measures of bullying perpetration, victimization, and probullying attitudes whereas
increases were expected for the measures of attitudes about reporting bullying. Absolute values could be used for this
computation because all observed change at each school was in the predicted direction.
c
p < .005 in the predicted direction.

Perkins et al. 717
overestimate the prevalence of bullying, and also
overestimate support for it in their perceptions
of the norm for peer attitudes (the first hypoth-
esis). The pattern replicates what has been found
for other youth risk behaviors, most notably con-
cerning alcohol, tobacco, and other drug use.
Furthermore, variation in personal attitudes
and behaviors observed among individual stu-
dents was highly correlated with variation in their
perceptions of the peer norm (the second
hypothesis), again a pattern commonly found in
research of substance abuse. Although variation
in personal behavior may be, in part, a determi-
nant of one’s perception (presuming some stu-
dents will rely on themselves as a referent for
establishing a sense of the peer norm), much
research as previously discussed, has also indi-
cated that peer norms, and more importantly, the
perception of peer norms, are strong determi-
nants of personal attitudes and behaviors. This
suggests that interventions to reduce mispercep-
tions can help reduce problem behaviors.
The third hypothesis predicted that an inter-
vention exposing students to accurate positive
norms based on local data through a print media
campaign at each of the five research sites would
reduce perceptions of bullying attitudes and
behaviors as the peer norm and concomitantly
reduce personal bullying behaviors and attitudes.
The predicted result of the intervention was pre-
cisely what was observed in the comparison of
the pre- and postintervention data. That is, results
showed significant reductions in problematic
misperceptions of the prevalence of bullying and
of peer support for bullying and simultaneous
reductions in personal bullying behaviors and
experiences of victimization. Students were also
more supportive of reporting bullying to school
authorities and parents and they came to believe
that peers more often supported this behavior
than was previously thought to be the case.
Without the availability of control site com-
parisons, one must be cautious in attributing the
change observed to the intervention that was
conducted. Other local events or newly intro-
duced programs or policies might have contrib-
uted to the observed changes. However, all five
schools with differing demographic characteris-
tics and drawn from different areas within the
state exhibited significant changes in the pre-
dicted direction and none of the schools experi-
enced any changes in the opposite direction.
Moreover, variation in the extent of change from
pre- to postintervention across sites corre-
sponded highly to the level of message exposure
achieved at each school. These facts provide a
greater confidence that the results are likely due
to the effect of the intervention, but further
research employing simultaneous control sites are
no doubt needed to more rigorously test our
third hypothesis. Use of a multiple baseline
design (multiple pretest assessments over time
before introducing the intervention) in future
research might be considered to strengthen the
evidence of any intervention effect if control
sites cannot be enlisted.
Another limitation of this study is the reliance
on self-report measures for an estimate of actual
bullying norms. It may be that some students did
not fully recall all of the bullying behaviors in
which they engaged over the last month or they
may not want to reveal the full extent of their
actions, and therefore, actual norms might be
higher than estimated based on self-reports. If so,
then the gap between the actual norm for per-
sonal behavior and perceived norms in reality
may not be as great as the results of this study
suggested. However, at least three points argue
against this possibility as significantly accounting
for the difference. First, the method employed
made clear that the surveys were anonymous for
students. Thus, they did not need to hide their
behavior for fear of punishment. Second, given
that students most often believed that the norm
for bullying behavior among other students was
so much more than they did, even among those
engaging in bullying, then it could be argued that
there should be little shame or fear in reporting
more behavior if they in fact were doing more
bullying. And, if they thought most others were
engaging in extensive bullying one might argue
that their perceived norm could even encourage
them to say they were personally doing more bul-
lying than was actually the case, meaning that the

718 Group Processes & Intergroup Relations 14(5)
actual gap could be larger than observed. Third,
the suggestion that recall error—the possibility
that the respondent would tend not to remember
all of the bullying behaviors over several weeks
time—is not an issue for the measures of atti-
tudes about bullying. One does not forget what
one’s attitude is, and yet the gap between personal
attitudes and perceived attitudinal norms of peers
was substantial as well. Although one may not
always act in accordance with one’s attitudes—
and here that may occur precisely because of the
pressure one feels to behave in bullying ways
because of misperceptions of the peer norm—
one’s attitude is still presumably what one states
unless one is intentionally being evasive.
The question about the accuracy of self-
reporting personal behavior may also arise in the
context of assessing the pre- to postintervention
change. It is possible that exposure to messages
indicating that engaging in bullying is not norma-
tive might lead some respondents to simply say
they are doing less than what they reported in the
initial survey given the new information.
However, there were no messages about the prev-
alence rates of victimization, only messages
about volitional behavior. And yet, being a victim
of bullying also declined in the wake of the inter-
vention, which strengthens the conclusion that
actual bullying had declined.
To conclude, this research suggests that a
social norms intervention may be a promising
strategy to help reduce bullying. Future research
should examine this approach in a broader range
of school settings, consider how misperceptions
emerge and are transmitted from cohort to
cohort in the school context, determine which
groups are most vulnerable to acting in accord-
ance with the erroneously perceived norm, and
explore other ways of delivering accurate norm
messages that would effectively reduce
misperceptions.
Acknowledgments
The authors wish to thank the staff of the Rowan Uni-
versity Center for Addiction Studies for their essential
role in the recruitment of schools in this research
and for their work at local school sites promoting the
implementation of the social norms media interven-
tion. The authors also wish to thank Bernadette van der
Vliet of BMT Design for her work designing the poster
media and Deborah Herry for administrative assistance
in producing print media. Funding for survey develop-
ment and administration and for poster media produc-
tion was provided by Rowan University Center for
Addiction Studies with support the Center received
from the New Jersey Department of Education.
References
Ajzen, I. (2001). Nature and operation of attitudes.
Annual Review of Psychology, 52(1), 27–58.
Ajzen, I. (2002). Perceived behavioral control, self-
efficacy, locus of control, and the theory of
planned behavior. Journal of Applied Social Psychology,
32, 665–683.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes
and predicting social behavior. Englewood Cliffs, NJ:
Prentice-Hall.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal-
directed behavior: Attitudes, intentions, and per-
ceived behavioral control. Journal of Experimental
Social Psychology, 22(5), 453–474.
American Psychological Association (APA). (1996).
Reducing violence: A research agenda. A Human
Capital Initiative report. Washington, DC: American
Psychological Association.
Asch, S. E. (1956). Studies of independence and con-
formity: A minority of one against a unanimous
majority. Psychological monographs, 70(9), 1–70.
Bewick, B. M., Trusler, K., Mulhern, B., Barkham, M.,
& Hill, A. J. (2008). The feasibility and effective-
ness of a web-based personalised feedback and
social norms alcohol intervention in UK univer-
sity students: A randomised control trial. Addictive
Behaviors, 33(9), 1192–1198.
Bowes, L., Arseneault, L., Maughan, B., Taylor, A., Caspi,
A., & Moffitt, T. E. (2009). School, neighborhood,
and family factors are associated with children’s bul-
lying involvement: A nationally representative longi-
tudinal study. Journal of American Academic Child and
Adolescent Psychiatry, 48(5), 545–553.
Brunstein Klomek, A., Marrocco, F., Kleinman, M.,
Schonfeld, I. S., & Gould, M. S. (2007). Bullying,
depression, and suicidality in adolescents. Journal
of American Academic Child and Adolescent Psychiatry,
46(1), 40–49.
Carlyle, K. E., & Steinman, K. J. (2007). Demographic
differences in the prevalence, co-occurrence, and

Perkins et al. 719
correlates of adolescent bullying at school. Journal
of School Health, 77(9), 623–629.
Center for Disease Control and Prevention (CDC).
(2008, August 11). Youth violence: Risk and protective
factors. Retrieved June 10, 2010, from http://www.
cdc.gov/ViolencePrevention/youthviolence/
riskprotectivefactors.html
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990).
A focus theory of normative conduct: Recycling
the concept of norms to reduce littering in pub-
lic places. Journal of Personality and Social Psychology,
58(6), 1015–1026.
Clemens, H., Thombs, D., Olds, R. S., & Gordon,
K. L. (2008). Normative beliefs as risk factors for
involvement in unhealthy weight control behavior.
Journal of American College Health, 56(6), 635–641.
Dake, J. A., Price, J. H., & Telljohann, S. K. (2003). The
nature and extent of bullying at school. Journal of
School Health, 73(5), 173–180.
DeJong, W., Schneider, S. K., Towvim, L. G., Murphy,
M. J., Doerr, E. E., Simonsen, N. R., …Scribner, R.
(2006). A multisite randomized trial of social norms
marketing campaigns to reduce college student
drinking. Journal of Studies on Alcohol, 67(6), 868–879.
Department of Health and Human Services (DHHS).
(2001). Youth violence: A report of the surgeon general.
Rockville, MD: U.S. Department of Health and
Human Services.
Eder, D., Evans, C. C., & Parker, S. (1995). School talk:
Gender and adolescent culture. New Brunswick, NJ:
Rutgers University Press.
Elias, M. J., & Zins, J. E. (Eds.). (2003). Bullying, peer
harassment, and victimization in the schools. Binghamton,
NY: Haworth Press.
Fabiano, P. M., Perkins, H. W., Berkowitz, A.,
Linkenbach, J., & Stark, C. (2003). Engaging men
as social justice allies in ending violence against
women: Evidence for a social norms approach.
Journal American College Health, 52(3), 105–112.
Finkelhor, D., Ormrod, R., Turner, H., & Hamby, S. L.
(2005). The victimization of children and youth: A
comprehensive, national survey. Child Maltreatment,
10(1), 5–25.
Hahn, R., Fuqua-Whitley, D., Wethington, H., Lowy, J.,
Crosby, A., Fullilove, M., …Liberman, A. (2007).
Effectiveness of universal school-based programs
to prevent violent and aggressive behavior: A sys-
tematic review. American Journal of Preventive Medicine,
33(2), S114–S129.
Haines, M. P., & Spear, S. F. (1996). Changing the per-
ception of the norm: A strategy to decrease binge
drinking among college students. Journal of American
College Health, 45(3), 134–140.
Haines, M. P., Barker, G. P., & Rice, R. (2003). Using
social norms to reduce alcohol and tobacco use in
two Midwestern high schools. In H. W. Perkins (Ed.),
The social norms approach to preventing school and college age
substance abuse: A handbook for educators, counselors, and cli-
nicians (pp. 235–244). San Francisco, CA: Jossey-Bass.
Hansen, W. B., & Graham, J. W. (1991). Preventing
alcohol, marijuana, and cigarette use among ado-
lescents: Peer pressure resistance training versus
establishing conservative norms. Preventive Medicine,
20(3), 414–430.
Hughes, C., Julian, R., Richman, M., Mason, R., &
Long, G. (2008). Harnessing the power of percep-
tion: Exploring the potential of peer group proc-
esses to reduce alcohol-related harm among rural
youth. Youth Studies Australia, 27(2), 26–35.
Janssen, I., Craig, W. M., Boyce, W. F., & Pickett, W.
(2004). Associations between overweight and obes-
ity with bullying behaviors in school-aged children.
Pediatrics, 113(5), 1187–1194.
Johnson, S. L. (2009). Improving the school environ-
ment to reduce school violence: A review of the
literature. Journal of School Health, 79(10), 451–465.
Juvonen, J., Graham, S., & Schuster, M. A. (2003).
Bullying among young adolescents: The strong, the
weak, and the troubled. Pediatrics, 112(6), 1231–1237.
Kilmer, J. R., Walker, D. D., Lee, C. M., Palmer, R. S.,
Mallett, K. A., Fabiano, P., & Larimer, M. (2006).
Misperceptions of college student marijuana use:
Implications for prevention. Journal of Studies on
Alcohol, 67(2), 277–281.
Klomek, A. B., Sourander, A., Kumpulainen, K., Piha,
J., Tamminen, T., Moilanen, I.…Gould, M.S. (2008).
Childhood bullying as a risk for later depression
and suicidal ideation among Finnish males. Journal
of Affective Disorders, 109(1–2), 47–55.
Klomek, A. B., Sourander, A., Niemela, S.,
Kumpulainen, K., Piha, J., Tamminen, T. …Gould,
M. S. (2009). Childhood bullying behaviors as a
risk for suicide attempts and completed suicides:
A population-based birth cohort study. Journal of
American Academic Child and Adolescent Psychiatry,
48(3), 254–261.
Kuntsche, E., Knibbe, R., Engels, R., & Gmel, G.
(2007). Bullying and fighting among adolescents
– Do drinking motives and alcohol use matter?
Addictive Behaviors, 32(12), 3131–3135.
Kuntsche, E., Pickett, W., Overpeck, M., Craig, W.,
Boyce, W., & de Matos, M. G. (2006). Television

720 Group Processes & Intergroup Relations 14(5)
viewing and forms of bullying among adolescents
from eight countries. Journal of Adolescent Health,
39(6), 908–915.
Linkenbach, J., & Perkins, H. W. (2003). Most of us
are tobacco free: An eight-month social norms
campaign reducing youth initiation of smoking in
Montana. In H. W. Perkins (Ed.), The social norms
approach to preventing school and college age substance
abuse: A handbook for educators, counselors, and clinicians
(pp. 224–244). San Francisco, CA: Jossey-Bass.
Lintonen, T. P., & Konu, A. I. (2004). The misper-
ceived social norm of drunkenness among early
adolescents in Finland. Health Education Research,
19(1), 64–70.
Lipsey, M. W., & Derzon, J. H. (1998). Predictors of vio-
lent or serious delinquency in adolescence and early
adulthood: A synthesis of longitudinal research. In
R. Loeber & D. P. Farrington (Eds.), Serious and
violent juvenile offenders: Risk factors and successful
interventions (pp. 86–105). Thousand Oaks, CA:
Sage Publications, Inc.
Lund, R., Nielsen, K. K., Hansen, D. H., Kriegbaum,
M., Molbo, D., Due, P., & Christensen, U. (2009).
Exposure to bullying at school and depression in
adulthood: A study of Danish men born in 1953.
European Journal of Public Health, 19(1), 111–116.
Martens, M. P., Page, J. C., Mowry, E. S., Damann,
K. M., Taylor, K. K., & Cimini, M. D. (2006).
Differences between actual and perceived student
norms: An examination of alcohol use, drug use,
and sexual behavior. Journal of American College
Health, 54(5), 295–300.
Mattern, J. L., & Neighbors, C. (2004). Social norms
campaigns: Examining the relationship between
changes in perceived norms and changes in drinking
levels. Journal of Studies on Alcohol, 65(4), 489–493.
McAlaney, J., & McMahon, J. (2007). Normative
beliefs, misperceptions, and heavy episodic drink-
ing in a British student sample. Journal of Studies on
Alcohol and Drugs, 68(3), 385–392.
Mouttapa, M., Valente, T., Gallaher, P., Rohrbach,
L. A., & Unger, J. B. (2004). Social network pre-
dictors of bullying and victimization. Adolescence,
39(154), 315–335.
Murray, M. E., Guerra, N. G., & Williams, K. R.
(1997). Violence prevention for the 21st century. In
R. P. Weissberg & T. P. Gulotta (Eds.), Healthy chil-
dren 2010: Enhancing children’s wellness (pp. 105–128).
Thousand Oaks, CA: Sage Publications.
Nansel, T. R., Craig, W., Overpeck, M. D., Saluja, G.,
& Ruan, W. J. (2004). Cross-national consistency
in the relationship between bullying behaviors and
psychosocial adjustment. Archives of Pediatric and
Adolescent Medicine, 158(8), 730–736.
Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan,
W. J., Simons-Morton, B., & Scheidt, P. (2001).
Bullying behaviors among US youth: Prevalence
and association with psychosocial adjustment.
Journal of the American Medical Association,
285(16), 2094–2100.
Nansel, T. R., Overpeck, M. D., Haynie, D. L., Ruan,
W. J., & Scheidt, P. C. (2003). Relationships between
bullying and violence among US youth. Archives of
Pediatric and Adolescent Medicine, 157(4), 348–353.
Neighbors, C., Larimer, M. E., & Lewis, M. A. (2004).
Targeting misperceptions of descriptive drinking
norms: Efficacy of a computer-delivered person-
alized normative feedback intervention. Journal of
Consulting and Clinical Psychology, 72(3), 434–447.
Neighbors, C., Walker, D. D., Mbilinyi, L. F., O’Rourke,
A., Edleson, J. L., Zegree, J., & Roffman, R. (2010).
Normative misperceptions of abuse among perpe-
trators of intimate partner violence. Violence Against
Women, 16(4), 370–386.
Page, R. M., Ihasz, F., Hantiu, I., Simonek, J., & Klarova,
R. (2008). Social normative perceptions of alcohol
use and episodic heavy drinking among Central and
Eastern European adolescents. Substance Use and
Misuse, 43(3–4), 361–373.
Page, R. M., Ihasz, F., Simonek, J., Klarova, R., &
Hantiu, I. (2006). Cigarette smoking, friendship
factors, and social norm perceptions among Central
and Eastern European high school students. Journal
of Drug Education, 36(3), 213–231.
Perkins, H. W. (1997). College student misperceptions of
alcohol and other drug norms among peers: Exploring
causes, consequences, and implications for prevention
programs. In Designing alcohol and other drug prevention
programs in higher education: Bringing theory into practice (pp.
177–206). Newton, MA: Higher Education Center for
Alcohol and Other Drug Prevention.
Perkins, H. W. (2002). Social norms and the prevention
of alcohol misuse in collegiate contexts. Journal of
Studies on Alcohol, Supplement 14, 164–172.
Perkins, H. W. (2003a). The emergence and evolution
of the social norms approach to substance abuse
prevention. In H. W. Perkins (Ed.), The social norms
approach to preventing school and college age substance
abuse: A handbook for educators, counselors, and clinicians
(pp. 3–18). San Francisco, CA: Jossey-Bass.
Perkins, H. W. (2003b). The social norms approach to pre-
venting school and college age substance abuse: A handbook

Perkins et al. 721
for educators, counselors, and clinicians. San Francisco,
CA: Jossey-Bass.
Perkins, H. W. (2007). Misperceptions of peer drink-
ing norms in Canada: Another look at the “reign
of error” and its consequences among college stu-
dents. Addictive Behaviors, 32(11), 2645–2656.
Perkins, H. W., & Berkowitz, A. D. (1986). Perceiving
the community norms of alcohol use among stu-
dents: Some research implications for campus alco-
hol education programming. International Journal of
Addiction, 21(9–10), 961–976.
Perkins, H. W., & Craig, D. W. (2003). The imaginary
lives of peers: Patterns of substance use and mis-
perceptions of norms among secondary school
students. In H. W. Perkins (Ed.), The social norms
approach to preventing school and college age substance
abuse: A handbook for educators, counselors, and clinicians.
pp. 209–223. San Francisco, CA: Jossey-Bass.
Perkins, H. W., & Craig, D. W. (2006). A successful
social norms campaign to reduce alcohol misuse
among college student-athletes. Journal of Studies on
Alcohol, 67(6), 880–889.
Perkins, H. W., Haines, M. P., & Rice, R. (2005).
Misperceiving the college drinking norm and related
problems: A nationwide study of exposure to preven-
tion information, perceived norms and student alco-
hol misuse. Journal of Studies on Alcohol, 66(4), 470–478.
Perkins, H. W., Meilman, P. W., Leichliter, J. S., Cashin,
J. R., & Presley, C. A. (1999). Misperceptions of the
norms for the frequency of alcohol and other drug
use on college campuses. Journal of American College
Health, 47(6), 253–258.
Perkins, J. M., Perkins, H. W., & Craig, D. W. (2010a).
Peer weight norm misperception as a risk factor
for being over- and underweight among UK sec-
ondary school students. European Journal of Clinical
Nutrition, 64(9), 965–971.
Perkins, J. M., Perkins, H. W., & Craig, D. W. (2010b).
Misperceptions of peer norms as a risk factor for
sugar-sweetened beverage consumption among
secondary school students. Journal of the American
Dietetic Association, 110(12), 1916–1921.
Poteat, V. P., & Spanierman, L. B. (2010). Do the
ideological beliefs of peers predict the prejudiced
attitudes of other individuals in the group? Group
Processes & Intergroup Relations, 13(4), 495–514.
Resnick, M. D., Ireland, M., & Borowsky, I. (2004).
Youth violence perpetration: What protects? What
predicts? Findings from the National Longitudinal
Study of Adolescent Health. Journal of Adolescent
Health, 35(5), 424.e1–424.e10.
Reuter-Rice, K. (2008). Male adolescent bullying and
the school shooter. Journal of School Nursing, 24(6),
350–359.
Salmivalli, C. (1999). Participant role approach to
school bullying: Implications for interventions.
Journal of Adolescence, 22(4), 453–459.
Salmon, G., James, A., & Smith, D. M. (1998). Bullying
in schools: Self reported anxiety, depression, and
self esteem in secondary school children. British
Medical Journal, 317, 924–925.
Sawyer, A. L., Bradshaw, C. P., & O’Brennan, L. M.
(2008). Examining ethnic, gender, and develop-
mental differences in the way children report being
a victim of “bullying” on self-report measures.
Journal of Adolescent Health, 43(2), 106–114.
Schultz, P. W. (1999). Changing behavior with normative
feedback interventions: A field experiment on curbside
recycling. Basic and Applied Social Psychology, 21(1), 25–36.
Schultz, P. W., Khazian, A. M., & Zaleski, A. C. (2008).
Using normative social influence to promote conser-
vation among hotel guests. Social Influence, 3(1), 4–23.
Sherif, M. (1936). The psychology of social norms. New
York, NY: Harper.
Sherif, M. (1937). An experimental approach to the
study of attitudes. Sociometry, 1(1/2), 90–98.
Smith, P. K., & Myron-Wilson, R. (1998). Parenting
and school bullying. Clinical Child Psychology and
Psychiatry, 3(3), 405–417.
Social Norms Surveys Online. Retrieved January 17,
2011, from http://www.socialnormsurveys.org/
Sourander, A., Helstela, L., Helenius, H., & Piha, J.
(2000). Persistence of bullying from childhood
to adolescence – A longitudinal 8-year follow-up
study. Child Abuse and Neglect, 24(7), 873–881.
Spriggs, A. L., Iannotti, R. J., Nansel, T. R., & Haynie,
D. L. (2007). Adolescent bullying involvement
and perceived family, peer and school relations:
Commonalities and differences across race/ethnicity.
Journal of Adolescent Health, 41(3), 283–293.
Srabstein, J. (2008). Deaths linked to bullying and haz-
ing. International Journal of Adolescent Medicine and
Health, 20(2), 235–239.
Srabstein, J., Joshi, P., Due, P., Wright, J., Leventhal,
B., Merrick, J., …Riibner, K. (2008). Prevention of
public health risks linked to bullying: A need for a
whole community approach. International Journal of
Adolescent Medicine and Health, 20(2), 185–199.
Srabstein, J., & Piazza, T. (2008). Public health, safety
and educational risks associated with bullying behav-
iors in American adolescents. International Journal of
Adolescent Medicine and Health, 20(2), 223–233.

722 Group Processes & Intergroup Relations 14(5)
Turner, J., Perkins, H. W., & Bauerle, J. (2008).
Declining negative consequences related to alcohol
misuse among students exposed to a social norms
marketing intervention on a college campus. Journal
of American College Health, 57(1), 85–94.
Biographical notes
h. wesley perkins is Professor of Sociology in
the Department of Anthropology and Sociology
at Hobart and William Smith Colleges. His
research interests include the social psychology
of norm misperception, normative influence on
health and well-being and evaluation of alcohol
abuse and violence prevention strategies for
adolescents and young adults in school and
community settings.
david w. craig is Professor of Biochemistry at
Hobart and William Smith Colleges. His research
interests include alcohol, tobacco, and other drug
use and violence among youth with an emphasis
on how perceptions of peers drive behavior and
how educational programs directed toward
correcting inaccurate perceptions can reduce
harmful behaviors.
jessica m. perkins is a doctoral candidate in
Health Policy at Harvard University. Her
research interests include exploring the influ-
ence of social capital, social networks, and
social norms in developed and developing
countries with an emphasis on how percep-
tions of peers drive behavior as well as exam-
ining social determinants of health across a
variety of outcomes. (e.g., weight status,
depression, alcohol use, and other health and
development topics).