Short Paper: Aggression in Childhood

E M P I R I C A L R E S E A R C H

The Role of the School Environment in Relational Aggression and Victimization

Caitlin Elsaesser • Deborah Gorman-Smith •

David Henry

Received: 21 July 2012 / Accepted: 5 October 2012 / Published online: 31 October 2012

� Springer Science+Business Media New York 2012

Abstract Research conducted over the last decade has

documented both the high rates of and serious consequences

associated with both victimization and perpetration of rela-

tional aggression. This study examines risk for involvement

in relational aggression and victimization among middle

school youth, evaluating both individual beliefs about vio-

lence, as well as aspects of the school environment, including

interpersonal school climate and school responsiveness to

violence. A sample of 5,625 primarily urban minority middle

school youth (49.2 % female) participating in a violence

prevention project completed measures of relational

aggression and victimization as well as indicators of indi-

vidual beliefs about aggression, school norms for aggression,

student–teacher and student–student interpersonal climate,

and school responsiveness to violence. Unlike results pre-

viously found for physical aggression, no school-level

indicator of climate was related to relational aggression or

victimization. However, individual beliefs about aggression

and individual perceptions of the school environment were

related strongly to both the perpetration of and victimization

by relational aggression. These results suggest not only that

individual beliefs and perceptions of the school environment

are important in understanding perpetration and victimiza-

tion of relational aggression, but also that risk for involve-

ment in relational aggression is distinct from that of physical

aggression. Implications for school interventions are dis-

cussed, as well as suggestions for future research.

Keywords Relational aggression � Relational victimization � School climate

Introduction: Risk and Protective Factors

for Relational Aggression

Research focused on relational aggression has increased

significantly in the last decade. Studies have documented

the serious consequences associated with both victimiza-

tion by and perpetration of relational aggression (Crick and

Grotpeter 1995; Crick et al. 1999; Preddy and Fite 2012).

This same research has shown that, while highly correlated

with physical aggression (Card et al. 2008), relational

aggression has a developmental course distinct from that of

physical aggression. Girls are more likely to report expe-

riences of both relational aggression and victimization,

though high rates are found across genders (Xie et al. 2003;

Waasdorp and Bradshaw 2009). Children who exhibit

relational aggression are at heightened risk for a host of

poor outcomes including peer rejection, loneliness,

depression and isolation (Card et al. 2008; Galen and

Underwood 1997), as well as substance abuse, externaliz-

ing and other behavior problems (Crick et al. 1996; Cul-

lerton-Sen and Crick 2005; Galen and Underwood 1997;

Sullivan et al. 2006). Given the serious problems linked to

both relational aggression perpetration and victimization,

understanding associated risk and protective factors is

critical to informing the development and implementation

of intervention and prevention programs focused on this

distinct form of aggression.

Much of the previous research on the correlates of

relational aggression has focused on individual-level pre-

dictors of involvement (e.g. Preddy and Fite 2012; Her-

renkohl et al. 2009). While these studies have provided

C. Elsaesser (&) � D. Gorman-Smith University of Chicago, 969 E. 60th St., Chicago, IL 60637, USA

e-mail: [email protected]

D. Henry

University of Illinois at Chicago, 1747 West Roosevelt Road,

Chicago, IL 60608, USA

123

J Youth Adolescence (2013) 42:235–249

DOI 10.1007/s10964-012-9839-7

important direction, relatively little research has expanded

beyond individual risk factors to examine environmental or

contextual factors related to relational aggression. Eco-

logical theories point to environmental and contextual

factors as important influences on child development

(Bronfenbrenner 1986). For children and youth of school

age, the school setting is a particularly important devel-

opmental context (Jessor 1993). School factors, including

norms about behavior and climate, consistently have been

shown to be important in understanding behavioral and

academic outcomes (Henry et al. 2011; Kuperminc et al.

1997). These findings suggest that the school environment

also may play an important role in risk for relational

aggression. This study is intended to expand on existing

research by examining the relationship between both

individual and school-level factors and risk for involve-

ment in relational aggression.

Of early research on relational aggression, much has

focused on preschool and early childhood samples, with

surprisingly less attention on middle school and early

adolescent youth. However, understanding the factors

related to relational aggression and victimization may be

especially important for early adolescents. Close peer

relationships are important at this developmental stage, and

risk for both perpetration and victimization may be par-

ticularly high (Savin-Williams and Berndt 1990). Given the

importance of peer relations at this time, actions intended

to harm those connections may be especially hurtful.

Contextual factors, particularly school factors, also may be

of special importance. Middle school youth are no longer

with one teacher for the full day, classroom constellations

of students change, and students often receive higher levels

of responsibility at school and spend more time away from

home, making the school and peer contexts especially

influential (Crockett and Petersen 1993). These develop-

mental changes indicate that the school environment—

particularly norms regarding aggression, relations between

students and teachers and school response to aggression

and violence—may be especially important in risk for

relational aggression.

The Impact of School Climate on Relational

Aggression and Victimization

Previous research points to three aspects of the school

context that may be important for understanding what

supports or decreases the risk for relational aggression

among middle school youth: (1) interpersonal school cli-

mate (i.e., the quality of student–teacher relationships and

student–student relationships), (2) norms for aggression,

and (3) school responsiveness to violence. Because the

literature on contextual factors related to risk more gen-

erally and school climate more specifically and relational

aggression is in its early stages, in cases where little

research exists on relational aggression, we draw on evi-

dence for physical aggression to inform our hypotheses.

Interpersonal Climate: Student–Student

and Student–Teacher Relationships

Social learning theorists have long posited that individuals

learn from observing the behavior of others, and that

behavior is more likely to be imitated when it is observed

repeatedly (Bandura 1986). Among early adolescents,

interactions between students and teachers and among

students are perhaps those most frequently observed in the

school environment. A robust body of research links

positive school interpersonal climate—relations between

students and teachers and among students—to lower levels

of emotional and behavioral problems, including exter-

nalizing problems (Roeser and Eccles 1998; Kuperminc

et al. 2001). The broader peer context may be particularly

influential for relationally aggressive youth, who have a

high degree of network centrality (Xie et al. 2002).

There is also evidence that more positive student–tea-

cher relationships promote lower levels of aggression

(Henry et al. 2011). Teacher warmth and quality of rela-

tionships with students have been associated with greater

student rejection of aggressive behavior (Chang 2003) and

lower levels of student aggression (Hughes et al. 1999).

Student–teacher relationships also may influence student

victimization. For example, anti-social student behavior

predicts conflict with teachers (Birch and Ladd 1998), and

these conflictive student–teacher relations may model a

lack of acceptance of the student to peers, who may

respond similarly. These data suggest that interpersonal

school climate impacts physical aggression. However, no

studies to our knowledge have evaluated the relationship

between interpersonal school climate and relational

aggression.

School Responsiveness to Violence: Perceptions of School

Safety and Awareness and Reporting

Data suggest that perceptions of school safety problems

(e.g., fighting, destruction of property and carrying weap-

ons at school) may cause emotional stress and increase

problem behaviors, especially for middle school students,

who are more likely than their younger peers to view the

school environment as unsafe (Astor et al. 2001). It has

been well documented that adolescents who are fearful in

school face an array of health risk behaviors, including

alcohol use, carrying weapons, and poor sleep habits

(Dowdell and Santucci 2003). Researchers also have found

positive associations between higher perceived school

safety problems—such as fighting and destruction of

236 J Youth Adolescence (2013) 42:235–249

123

property—and bullying perpetration, victimization, and

problem behaviors (Bowen and Bowen 1999). While it is

clear that school safety concerns influence physical

aggression, less research has explored how school safety

may impact relational aggression and victimization.

Related to students’ experience of school safety is the

way teachers and administrators respond to aggression and

violence. Schools that have unclear and inconsistent poli-

cies and that ignore student misbehavior have been found

to have higher rates of student victimization (Gottfredson

and Gottfredson 1985). Henry et al. (2011) found that

higher school-levels of awareness and reporting of

aggression and violence predicted lower individual-levels

of physical aggression. Policy evaluations support this

connection: interventions designed to raise student and

teacher awareness about school violence and create clear

rules and policies have been found to reduce aggression

and violence within schools (Astor et al. 2005).

It is not clear if and how these same associations may hold

for relational aggression. Research has found that adults’

response to relational aggression may differ from that of

physical aggression, with teachers responding inconsistently

to acts of relational aggression (Yoon and Kerber 2003).

Unlike physical aggression, there may be more acceptance

and belief that this type of behavior is ‘‘normal’’ and

expected of youth (Casey-Cannon et al. 2001). Additionally,

relationally aggressive acts, such as peer manipulation and

spreading rumors, tend to be covert (Crick and Grotpeter

1995), making them less visible to those outside the peer

group and more difficult for teachers and administrators to

respond. Given the findings in the literature on physical

aggression, however, we expect that students who perceive

higher levels of school safety problems and higher respon-

siveness to violence will be less likely to experience rela-

tional aggression and victimization.

School-Level Norms Regarding Violence

Social norm theorists highlight the powerful influence

collective norms have on individual behavior (Cialdini

et al. 1990). This body of literature distinguishes between

descriptive norms (mean levels of behavior) and injunctive

norms (the perception of what should be) (Cialdini et al.

1991). While both descriptive and injunctive norms have

been linked to physical aggression (Henry et al. 2011;

Barth et al. 2004), research has shown that injunctive

norms about the acceptability of aggression exert an

especially strong influence on individual-level aggression

(Henry et al. 2000). Injunctive norms may motivate

behavior by promising social rewards or punishment for

particular actions (Cialdini et al. 1991).

Few studies have evaluated the relationship between

school-level norms and relational aggression. Onishi et al.

(2011) found that perceptions of classroom injunctive

norms regarding abusive behavior were related to relational

aggression among a sample of 5 and 6th grade adolescents.

However, their study was cross-sectional and drew on a

sample of children in Japan, which may have limited

generalizability to American youth. Kuppens et al. (2008)

found in a longitudinal study that classroom-level norms

supporting aggression were associated positively with

relational aggression among 3rd through 5th grade youth.

This study conceptualized norms as the mean level of

behavior in the class and did not examine the impact of

injunctive norms. The present study builds on previous

research by examining the impact of school-level injunc-

tive norms regarding aggression on relational aggression.

Individual Beliefs

Cognition plays a powerful role in the perpetration of

aggressive acts, according to learning theories of aggres-

sion (Bandura 1986; Huesmann 1988). Beliefs may have an

especially salient influence on aggression as youth enter

middle school. Whereas aggressive behavior has been

found to be predictive of normative beliefs about aggres-

sion among elementary school children, as children age, it

is beliefs supporting aggression that predict aggressive

behavior (Huesmann and Guerra 1997). Research suggests

that as children approach adolescence, beliefs solidify and

become more predictive of behavior; these beliefs may

affect aggression through heightening the perception of

hostility from others or cuing aggressive scripts for social

behavior (Guerra et al. 1994; Huesmann 1988). Empirical

evidence backs the position that normative beliefs have a

powerful influence on aggression. Numerous studies dem-

onstrate that individual beliefs exert a strong influence on

individual-level behavior (e.g., perpetration of aggression)

(Zelli et al. 1999). More recently, studies have shown that

individual beliefs supporting aggression are correlated

positively with relational aggression (Werner and Hill

2010). Researchers also recently have found evidence that

norms supporting nonviolence may impact aggression

differently than beliefs supporting nonviolence (Henry

et al. 2011). These findings suggest that individual-level

beliefs impact physical aggression; however, little is

known about how beliefs supporting nonviolence impact

relational aggression.

Current Research and Hypotheses

The current study expands on previous research by

exploring factors in the school environment that may

contribute risk for relational aggression and victimization.

Specifically, this study is designed to evaluate whether

J Youth Adolescence (2013) 42:235–249 237

123

individual beliefs about aggression and individual percep-

tions of the school environment (school interpersonal cli-

mate and school responsiveness to violence) are predictive

of relational aggression and victimization. This study also

explores whether aggregated school-levels of interpersonal

climate, school responsiveness to violence, and school

norms favoring nonviolence are predictive of relational

aggression and victimization after controlling for individ-

ual-levels of these constructs. Given gender differences

that have been found both in the rates of relational

aggression (Crick and Grotpeter 1995; Xie et al. 2003) and

the intimacy of personal relationships (Maccoby 1998),

potential gender differences are explored.

Two broad sets of hypotheses guide the present study. We

expect individual beliefs and perceptions of the school

environment, controlling for the aggregated school-level

indicators, to be associated with relational aggression and

victimization. We similarly expect school-level variables,

controlling for individual-level variables, to be associated

with relational aggression and victimization. For both the

individual perceptions and aggregated school predictors, we

anticipate that a more positive interpersonal climate, norms

more supportive of nonviolence and less supportive of vio-

lence, and higher levels of awareness and reporting and

lower levels of school safety problems will be associated

negatively with relational aggression and victimization.

Second, we hypothesize that individual perceptions of

the school environment and their aggregated school levels

will be predictive of relational aggression for both females

and males, but that the strength of these relationships will

vary by gender. The broader literature has indicated that

female peer relations may be characterized by higher

intimacy and than those of males (Maccoby 1998; Parker

and Asher 1993), and that females may be more responsive

than males to the input of teachers (Underwood 2003).

Based on these findings, we expect that the effect of

interpersonal school climate on relational aggression and

victimization will be stronger for females.

Method

Procedure

This study uses data collected in the context of a multisite

randomized trial. Investigators from four universities—

Duke University, The University of Georgia, University of

Illinois at Chicago and Virginia Commonwealth Univer-

sity—recruited area schools serving middle school stu-

dents. Data were collected from students who returned both

assent and parental consent forms; telephone follow-ups

and home visits were made to increase study participation.

Of 7,364 eligible students, 5,625 returned active parental

consent and student assent, for a mean recruitment rate of

76 %. Students received a $5 gift card for returning consent

forms at the 3 sites it was permitted. Nine-tenths of the

sample had two waves of the outcome variables, resulting

in a sample size of 5,106 for the present study. The insti-

tutional review boards at the four participating universities

and the Centers for Disease Control and Prevention Insti-

tutional Review Board approved all study procedures.

Two to three schools from each site were assigned

randomly to one of four intervention conditions: (1) a

universal condition including a social-cognitive student

intervention (Meyer et al. 2004) and a teacher training and

support intervention (Orpinas et al. 2004; n = 9 schools

and 1,305 youth); (2) a selective intervention condition,

which included a group-based family intervention (Phil-

lips-Smith et al. 2004; n = 10 schools and 1,349 youth);

(3) a condition that included both universal and selective

interventions (n = 9 schools and 1,241 youth); and (4) a

no-intervention control condition (n = 9 schools and 1,211

youth).

Graduate students administered a 20 session social-

cognitive problem-solving curriculum focused on avoiding

danger, ignoring teasing, and asking for help in the uni-

versal condition (Meyer et al. 2000). Teacher training

included a two-day workshop and 10 support group

meetings for teachers; meetings aimed to augment teacher

awareness of various forms of aggression, increase ability

to prevent aggression, and develop better classroom man-

agement skills (Orpinas et al. 2004). The selective inter-

vention was designed to increase home-school

partnerships, teacher efficacy, parent monitoring and

communication strategies, as well as improve parent and

child coping and problem solving skills. Students who were

identified by teachers as exhibiting high levels of aggres-

sion and influence with peers were recruited, along with

their parents. A 15-week intervention was conducted for

four to eight youth and their families. Sessions began with

a meal and introduced new topics through interactive

activities, including role-plays. Beginning in 2001, data

from a representative sample of students and teacher rat-

ings of students was collected for two successive cohorts of

6th grades. Data was collected from students and teacher

ratings of students in each cohort during the fall and spring

of the 6th grade and in the spring of 7 and 8th grade years.

Participants

This study uses six waves of data collected over three years

(6 through 8th grade) from 5,106 middle school students.

This sample was recruited from 37 schools at four sites

(northeastern GA, n = 9 schools; Chicago, IL, n = 12

schools; Durham, NC, n = 8 schools; and Richmond, VA,

n = 8 schools) that participated in the Multisite Violence

238 J Youth Adolescence (2013) 42:235–249

123

Prevention Project (Multisite Violence Prevention Project

(MVPP) 2004). In Chicago, where schools include grades

K-8, we selected schools for size (over 1,100 students) and

those with at least 75 % of students residing within school

district boundaries. Participating schools in Durham and

Richmond represented almost all middle schools in the

public school systems, while middle schools in Georgia

represented six school districts in Northeastern Georgia.

All included schools had a high proportion of low-income

students, based on eligibility for the federal free and

reduced lunch program (percentage ranged from 42 to

96 % across sites). See Table 1 for demographic charac-

teristics by site.

Participants consisted of a random sample of about 80

students per cohort from the rosters of each of the larger

middle schools in three of the sites and all eligible students

at the smaller Chicago schools. Students who had been

selected for Cohort 1 but repeated the 6th grade were not

included in Cohort 2. The sample was approximately 50 %

male and 50 % female, and 52.1 % reported African

American ethnic identification, 21.2 % Hispanic, 17.3 %

Caucasian, with remaining students identifying as Ameri-

can Indian or Asian. Of participants, 70 % reported the

presence of an adult male in the home (see Table 1).

Measures

This study utilized two individual-level outcome variables

(relational aggression and relational victimization) and

three individual-level and school-level constructs drawn

from six predictor variables: individual beliefs about

aggression, individual perceptions of school norms for

aggression, interpersonal climate (positive student–teacher

relationships and positive student–student relationships),

and school responsiveness to violence (awareness and

reporting of violence and school safety problems). School-

level scores on the predictors were constructed by taking

the mean score of all students within each cohort at each of

the 37 schools for each wave, resulting in 74 school-level

scores for each wave. Thus, each student in a particular

cohort had identical scores on each school-level measure.

Relational Aggression

Relational aggression was assessed with a 6-item scale

drawn from the Problem Behavior Frequency Scales

(Farrell et al. 2000). The items were based on Crick and

Grotpeter’s (1995) measure of relational aggression. All

items were preceded by the stem, ‘‘In the last 30 days, how

many times have you?’’ Items included ‘‘spread a false

rumor about someone,’’ and ‘‘left another kid out on pur-

pose when it was time to do an activity’’ and were rated on

a 6-point scale ranging from ‘‘Never’’ to ‘‘20 or more

times.’’ Scores were calculated based on the mean value for

items in the scale; higher values represent a higher level of

aggression. The internal consistency of the relational

aggression scale was .72.

Relational Victimization

Relational victimization was assessed using the same items

as the relational aggression scale, but with the stem, ‘‘In the

last 30 days, how many times has this happened to you?’’

Items included ‘‘been left out on purpose by other kids

when it was time to do an activity,’’ and ‘‘had a told tell lies

about you to make other kids not like you anymore.’’ Items

Table 1 Multi-site violence prevention project school and

sample characteristics by site

Chicago Durham Georgia Richmond

School characteristics

No. of participating schools 12 8 9 8

Average no. of 6th graders 70 241 239 236

Grade levels K-8 6–8 6–8 6–8

% of Eligible for federal lunch program 96 42 47 75

Overall sample

Black non Hispanic (%) 37 56 29 75

Hispanic (%) 53 9 12 5

White non Hispanic (%) 2 23 46 6

Multiracial (%) 6 8 9 11

Gender (% male) (%) 50 47 50 49

Adult male in the home (%) 74 68 77 61

Two-parent families (%) 47 44 53 30

Single parent families (%) 35 36 27 44

Single parent & stepparent families (%) 11 12 15 15

J Youth Adolescence (2013) 42:235–249 239

123

were rated on a 6-point scale, ranging from ‘‘Never’’ to ‘‘20

or more times.’’ Scores were calculated based on the mean

value for items in the scale; higher values represent a

higher level of victimization. The scale has a Cronbach’s

alpha of .84.

Individual Beliefs About Aggression and Nonviolence

Individual approval of aggression and nonviolence were

assessed using two scales, one (10 items) assessed beliefs

about verbal and physical aggression, and the other (8

items) assessed beliefs about nonviolent solutions to con-

flict. Items were rated on a 3-point scale (approval, dis-

approve, or neutral). Questions about aggression included,

‘‘How would you feel if a kid at your school threatened

someone who said something mean’’ and questions about

nonviolence included, ‘‘How would you feel if a kid in

your school avoided a fight by walking down a different

hall to class?’’ The scale of individual norms measure had

internal consistency reliability of .73 for aggression and .74

for nonviolence (Miller-Johnson et al. 2004). Previous

research has found that at the individual-level these scales

are only moderately correlated with each other and have

independent effects on measures of aggressive behavior

(Henry et al. 2011; Henry et al. (in press)).

School-Level Norms Regarding Aggression

School level norms were measured by creating a single

scale out of two scales, one assessing perception of

schoolmate’s approval of aggression (10 items), and the

other schoolmate’s approval of nonviolent solutions to

conflict (8 items). Items were rated using a 3-point scale

(approval, disapprove, or neutral). Questions about

aggression included ‘‘How would other students feel if a

kid at your school hit someone for no reason,’’ and ques-

tions about nonviolence included, ‘‘How would students at

your school feel if a kid in your school avoided a fight by

walking down a different hall to class?’’ The scale of

perceived school norms measure had internal consistency

reliability of .80 for aggression and .70 for nonviolence

(Miller-Johnson et al. 2004). The school-level scale was

created by reverse scoring the 10 aggression items and then

aggregating 18 questions from the two individual-level

scales for perceived school norms described above by

taking the mean of each class (cohort within school).

Multilevel factor analysis had revealed that a single factor

at the school-level, representing support for nonviolence

and disapproval of aggression, fit the data well (6th grade

fall, v2(135) = 110.31, ns, CFI = 1.0, RMSEA = 0.000; 8th grade spring, v2(135) = 58.85, ns, CFI = 1.0, RMSEA = 0.000; Henry et al. 2011) but that two factors

were required to model the data at the individual level.

Moreover, the school-level scale had effects on physical

aggression with the individual-level scores in the model.

Interpersonal Climate

Interpersonal climate was assessed through two measures:

positive student–student relationships and positive student–

teacher relationships.

Positive student–student relationships Student–student

relationships were assessed using the Vessels’ School

Climate Survey (Vessels 1998). Students are asked to

report on the degree to which students in the school

get along with each other. Items included ‘‘students are

kind and supportive of one another’’ and ‘‘students stop

other students who are unfair or disruptive.’’ Responses

were rated on a 4-point scale ranging from ‘‘strongly dis-

agree’’ to ‘‘strongly agree,’’ with higher scores representing

more positive relationships. Internal consistency was .61.

Positive student–teacher relationships Student–teacher

relationships were also assessed using the Vessels’ School

Climate Survey (Vessels 1998). Items included ‘‘teachers

treat students fairly’’ and ‘‘teachers praise students more

often than they criticize them.’’ Responses are rated using a

4-point scale ranging from ‘‘strongly disagree’’ to

‘‘strongly agree,’’ with higher scores representing more

positive student–teacher relationships. The internal con-

sistency was .66.

School Responsiveness to Violence

School responsiveness to violence was assessed with two

measures: awareness and reporting of violence and school

safety concerns.

Awareness and reporting of violence Awareness and

reporting of violence was measured using a scale from the

Vessels School Climate Scale (Vessels 1998). The scale is

intended to measure teachers’ response to hearing about or

witnessing bullying and violence and includes items such

as ‘‘teachers know when students are being picked on or

being bullied’’ and ‘‘students report it when one student

hits another.’’ Responses are rated using a 4-point scale

ranging from ‘‘strongly disagree’’ to ‘‘strongly agree.’’

Internal consistency for the scale was .63. Higher scores

indicate higher levels of awareness and reporting of

violence.

School safety concerns Concerns about school safety were

measured with a 10-item scale based on the Department of

Education Schools and Staffing Survey (US Department of

Education 1999–2000). The scale is intended to measure the

extent to which students perceive their school to be unsafe, with

higher scores indicating higher levels of concern regarding

240 J Youth Adolescence (2013) 42:235–249

123

school safety. Students were asked questions representing

potential problems or issues within the school and included

items such as ‘‘fighting (hitting and kicking) among students’’

and ‘‘students carrying weapons.’’ Students responded using a

4-point scale, with responses ranging from ‘‘not a problem’’ to

‘‘serious problem.’’ Internal consistency was .89.

Demographics

Demographic information was collected from student

reports and included gender, race, ethnicity and family

structure (presence of adult male in the home).

Data Analytic Approach

Henry et al. (2011) report acceptable school-level reliability

for each of the school-level predictors used in this investi-

gation. Three mixed effects regression models were run for

each outcome to assess the impact of individual- and school-

level predictors: (1) a base model with the non-climate pre-

dictors (e.g. time, gender, intervention condition, and

school); (2) a model adding all individual-level predictors to

the base model; and (3) a third model adding all school-level

predictors to Model 2. Mixed effects regression models can

include cases with missing data at one or more waves of

assessment without requiring multiple imputation since each

data point within every individual is used to estimate the

group trend at the time the measure was gathered. The

additional explanatory power of each model was assessed by

finding the difference in -2 log-likelihoods between the

models and evaluating the result using the Chi-square dis-

tribution with degrees of freedom equal to the difference in

free parameters between models. The model with the smaller

-2 log likelihood value was considered a better fit if the

difference in -2 log-likelihoods between the models was

significant. Additionally, a pseudo R 2

was calculated for the

additional explanatory power of each model by dividing the

difference in -2 log-likelihoods by the -2 log-likelihood of

the base model.

Mixed effects regression models were then used to

evaluate the effects of the school-level predictors on rela-

tional aggression and victimization. First, a set of mixed

effects regression models were used to examine the effect

of each school-level predictor on each outcome variable. A

fixed quadratic time variable was included in the initial

models to explore whether the growth over time was cur-

vilinear. Only predictors that emerged as significantly

associated with the outcomes were included in the final

models for parsimony.

The models included terms to measure the effect of eth-

nicity and adult male in the home. Time was centered at

pretest so all effects on the school intercepts could be inter-

preted as effects at the fall of 6th grade. Each model also

included dummy codes for the intervention condition with

the control group as a comparison, so that all effects could be

interpreted as expected effects in the control group. Henry

et al. (2011) tested this procedure, finding that the parameter

estimates for the full sample closely approximated the

parameter estimates that were obtained in analysis using the

control group only. School level terms that were significant

in these initial models were entered into a final model.

We tested moderation by gender for all outcome vari-

ables by entering a binary variable representing gender as

well as an interaction between gender and each predictor in

each model. If the interaction was significant effects were

estimated separately by gender to facilitate interpretation.

Results

Preliminary Analyses

Descriptive statistics and correlations for individual-level

outcome variables at baseline (6th grade fall) and 8th grade

spring are reported in Table 2. The correlations were

moderate in magnitude (.34–.37) and were in the expected

direction. These correlations were significant, which was

expected given the large sample size.

Individual and school-level predictor descriptive statis-

tics and correlations are presented in Table 2. Baseline

correlations are reported on the subdiagonal and 8th grade

spring correlations are reported on the superdiagonal. Most

correlations were small to moderate. Student–student

relationships and student–teacher relationships were mod-

erately correlated (.36 at baseline and .43 at spring 8th

grade for individual and .48 at baseline and .56 at spring

8th grade for school-level)—not an unexpected result given

that these predictors are conceptually related. Student–

teacher relationships and awareness and reporting were

also moderately correlated (.48 at baseline and .61 at spring

8th grade for individual and .53 at baseline and .76 at

spring 8th grade for school).

Main Effects and Changes in Effects on Relational

Aggression over the Course of Middle School

Main effects did not differ by intervention condition for

models predicting relational aggression perpetration. The

quadratic term added in the initial models was not significant

and was therefore dropped from the final models. Consistent

with our hypothesis, individual beliefs and individual per-

ceptions of school climate added significant explanatory

power to the model both in the fall of 6th grade and over time.

The individual-level variables added approximately

R 2

= .10 for at the cross sectional level and R 2

= .07 in the

longitudinal model. However, contrary to our hypothesis, the

J Youth Adolescence (2013) 42:235–249 241

123

addition of school-level indicators of school climate did not

have a significant impact on relational aggression in the 6th

grade. Over time, these variables together added marginal

explanatory power to the model (see Table 3), but the con-

tribution in terms of pseudo R 2

was miniscule, and none of

the school-level predictors explained significant unique

variance in either outcome.

Effects of Individual Beliefs

Individual beliefs favoring aggression were significantly

and positively associated with relational aggression

perpetration (B = .62, p \ .001), as was hypothesized. The strength of this association varied by gender, as expected:

Compared to females, males showed a somewhat weaker

positive association between beliefs favoring aggression

and relational aggression (B = .48, p \ .001) (see Table 4). The positive effect of individual beliefs favoring

aggression on relational aggression became stronger over

the course of middle school (B = .40, p \ .001), support- ing our hypothesis; but contrary to expectations, change

over time did not vary by gender. Individual beliefs

favoring nonviolence was significantly and negatively

associated with relational aggression at baseline (B =

Table 2 Descriptives, outcomes and predictors

6th grade fall 8th grade spring Correlations

Mean SD Mean SD 1 2 3 4 5 6

Outcomes

1. Relational aggression 1.54 .64 1.61 .73 1 .34***

2. Relational victimization 1.73 .85 1.49 .73 .37*** 1

Individual-level predictors

Norms

1. Beliefs favoring nonviolence (NL) 2.32 .43 2.35 .42 1 -.46*** .13*** .24*** .19*** -.13***

2. Beliefs favoring aggression (NA) 1.38 .35 1.65 .41 -.13*** 1 -.19** -.30*** -.30*** .03***

Interpersonal climate

3. Student–student relationships (SS) 2.66 .46 2.64 .41 .00 -.21*** 1 .43*** .47*** -.12***

4. Student–teacher relationships (ST) 3.12 .58 2.80 .64 .10*** -.25*** .36*** 1 .61*** -.06***

School responsiveness to violence

5. Awareness and reporting (AR) 2.96 .50 2.59 .56 .04*** -.26*** .44*** .48*** 1 -.05***

6. School safety problems (SP) 1.83 .82 1.56 .79 -.17*** -.02 -.02 -.03*** .02 1

School-level predictors

1. School norms favoring nonviolence (SN) -.08 .09 -.31 .12 1 .44** .50** .65** -.31**

Interpersonal climate

2. Student–student relationships (SS) 2.67 .08 2.64 .10 .49** 1 .56** .53** .00

3. Student–teacher relationships (ST) 3.12 .10 2.80 .18 .39** .48** 1 .76** .18**

School responsiveness to violence

4. Awareness and reporting (AR) 2.96 .10 2.59 .15 .57** .69** .53** 1 .00

5. School safety problems (SP) 1.83 .28 1.56 .29 -.14** .21** -.11** .19** 1

Units of analysis consist of one cohort in one school. 6th-grade fall correlations are on the sub-diagonal, 8th-grade spring correlations are on the

superdiagonal

* p \ .05, ** p \ .01, *** p \ .001

Table 3 Significance tests between models

Relational aggression Relational victimization

Cross sectional Longitudinal Cross sectional Longitudinal

Model 1: demographics and time 10,819.4 39,804.3 11,568.4 39,742.6

Model 2: individual-level effects 1,044.5*** 2,913.6*** 405.5*** 1,164.8***

Change in pseudo R 2

.10 .07 .04 .03

Model 3: school-level effects 9.6 34.2* 8 33.4*

Change in pseudo R 2

0 0 0 0

* p \ .05, ** p \ .01, *** p \ .001

242 J Youth Adolescence (2013) 42:235–249

123

-.08, p \ .001), consistent with our hypothesis, and this relationship became stronger over time (B = -.14,

p \ .001). However, contrary to predictions, this relation- ship did not vary by gender.

Effects of Interpersonal Climate

Individual perceptions of positive student–student rela-

tionships was negatively associated with relational

aggression perpetration in the fall of 6th grade (B = -.16,

p \ .001), consistent with our hypothesis. This relationship became increasingly negative over time (B = -.13,

p \ .0001) but did not vary by gender (see Table 4). Also supporting our hypothesis, students who perceived more

positive student–teacher relationships also exhibited lower

levels of relational aggression (B = -.10, p \ .001); this relationship became stronger over the course of middle

school (B = -.11, p = .001) but did not vary by gender.

Effects of Responsiveness to Violence

Individual perceptions of awareness and reporting were not

significantly associated with relational aggression and there

was no significant interaction with gender. Although the

interaction between the time slope and this predictor was

significant (B = -.07, p \ .001), post hoc analysis revealed that by 8th grade, awareness and reporting was not signifi-

cantly associated with relational aggression. Students who

perceived higher levels of school safety problems exhibited

significantly higher levels of relational aggression in the 6th

grade (B = .04, p \ .001), findings in line with expectations. This relationship became stronger during middle school

(B = .03, p \ .05) (see Table 4) and did not vary by gender.

Main Effects and Changes in Effects on Relational

Aggression over the Course of Middle School

In models predicting relational victimization, none of the

intervention effects were significant and the quadratic term

was not significant. Individual level indicators of school

climate added significant explanatory power to the model

both in the fall of 6th grade and over time, in line with our

expectations. Individual-level predictors explained an

additional R2 = .04 in relational victimization in the cross

sectional model, and an additional R2 = .03 in the longi-

tudinal model. As with relational aggression, the school-

level indicators of school climate did not have a significant

impact on relational victimization in the 6th grade, con-

tradicting our hypothesis. Over time, these variables toge-

ther added marginal explanatory power to the model but

none of the specific school-level climate variables

explained unique additional variance (see Table 3).

Table 4 Effects of individual-level predictors, controlling for school- level predictors

Cross

sectional

Longitudinal

Relational aggression

Individual beliefs about aggression

Individual beliefs favoring aggression .62*** .40***

Time*beliefs favoring aggression – -.12***

Gender*beliefs favoring aggression -.14** -.02

Individual beliefs favoring nonviolence

aggression

-.08** -.14***

Time*beliefs favoring nonviolence – -.02

Gender*beliefs favoring nonviolence .02 .05*

Interpersonal climate

Student–student relationships -.16*** -.13***

Time*student–student relationships – -.02

Gender*student–student relationships .03 .01

Student–teacher relationships -.10*** -.11***

Time*student–teacher relationships – .00

Gender*student–teacher relationships -.01 .00

Responsiveness to Violence

Awareness and Reporting -.01 -.07***

Time*awareness and reporting – .00

Gender*awareness and reporting -.04 .01

School safety problems .04** .03*

Time*school safety problems – .01

Gender*school safety problems -.01 .02

Relational victimization

Individual beliefs about aggression

Individual beliefs favoring aggression -.06 -.07***

Time*beliefs favoring aggression – -.02

Gender*beliefs favoring aggression .07 .08**

Individual beliefs favoring nonviolence

aggression

-.01 .02

Time*beliefs favoring nonviolence – .02

Gender*beliefs favoring nonviolence .00 .01

Interpersonal Climate

Student–student relationships -.23*** -.18***

Time*student–student relationships – .03

Gender*student–student relationships .02 -.01

Student–teacher relationships -.05 -.03*

Time*student–teacher relationships – -.02

Gender*student–teacher relationships -.04 -.02

Responsiveness to violence

Awareness and reporting -.02 -.02

Time*awareness and reporting – .01

Gender*awareness and reporting .02 .02

School safety problems .08*** .08***

Time*school safety problems – -.01

Gender*school safety problems .01 .01

* p \ .05, ** p \ .01, *** p \ .001

J Youth Adolescence (2013) 42:235–249 243

123

Effects of Individual Beliefs

Individual beliefs favoring aggression were not signifi-

cantly associated with relational victimization and did not

vary by gender; although the interaction between time and

perceived norms indicated significant change in the asso-

ciation over time (B = -.07, p \ .001), post hoc analyses revealed that this construct was not significantly associated

with relational victimization by the spring of 8th grade.

Individual beliefs favoring nonviolence were not signifi-

cantly associated with relational victimization and did not

have a significant interaction by gender (see Table 4).

These results were contrary to our hypotheses.

Effects of Interpersonal Climate

Students who perceived better relations with other students

reported lower levels of relational victimization (B = -.23,

p \ .001), and this became increasingly true over the course of middle school (B = -.18, p \ .001), supporting our hypothesis. This relationship did not vary by gender. Student

perception of student–teacher relationships was not signifi-

cantly associated with relational victimization and there was

no significant moderation by gender, also contrary to our

hypotheses (see Table 4).

Effects of Responsiveness to Violence

Perception of awareness and reporting at the individual

level was not associated with relational victimization and

did not vary by gender. Perceptions of school safety

problems were positively associated with relational vic-

timization (B = .08, p \ .001), in line with our hypothesis, and this relationship became increasingly strong over time

(B = .08, p \ .001). This relationship did not vary by gender, contrary to expectations (see Table 4).

Discussion

Although the literature has highlighted important individ-

ual risk factors for relational aggression (Preddy and Fite

2012; Herrenkohl et al., 2009), few studies have examined

how contextual factors—particularly norms around the use

of aggression and violence, interpersonal relationships

within the school environment, and school response to

aggression and violence—may relate to relational aggres-

sion. Schools are among the most important developmental

contexts for adolescents (Jessor 1993), and the literature

has provided strong evidence that the school environment

has a powerful effect on physical aggression (Henry et al.

2011). However, little is known about how factors in the

school environment may influence relational aggression.

This study was one of the first to examine the impact of

both individual- and school-level risk factors related to the

perpetration and victimization of relational aggression over

the course of middle school.

Consistent with the literature on physical aggression, we

find that perceptions of interpersonal school climate and

school safety were related to both relational aggression

perpetration and victimization. Individual beliefs regarding

the use of aggression and nonviolence alternatives were

also, not surprisingly, related to relational aggression per-

petration. Thus, as expected, individual beliefs and per-

ceptions of safety and the interpersonal environment were

related to involvement in relational aggression.

However, unlike results found for physical aggression

(Henry et al. 2011), school-level indicators of school cli-

mate were unrelated to youth victimization or perpetration

of relational aggression above and beyond individual per-

ceptions of the same school climate constructs. Social

ecological theory has long emphasized that perceptions of

the social environment are critical to understanding how

individuals adapt to their social environment (Bronfen-

brenner 1986). These results suggest that relational

aggression appears to be more strongly related to individ-

ual beliefs and individual perceptions of the environment

than aggregated indicators of the school environment.

Beliefs Favoring Aggression and Nonviolence

Consistent with previous findings (Kuppens et al. 2008;

Onishi et al. 2011), we found that individual beliefs were

associated with higher levels of relational aggression per-

petration. These findings also extend previous research by

suggesting that it is not just norms supporting violence, but

also norms supporting nonviolence that are important for

understanding risk. Having strong beliefs that there are

nonviolent alternatives to violence is related to reduced risk

for relational aggression. These findings suggest a poten-

tially important point of intervention for programs aiming

to reduce relational aggression.

Notably, we found that the relationship between beliefs

about aggression and relational aggression was moderated

by gender. The effect of individual beliefs supporting

aggression on relational aggression was stronger for

females than males. It is possible that beliefs supporting

aggression have stronger effects on relational aggression

for females compared to males because the behavior is

more common among middle school females (Xie et al.

2003). Data have indicated that engaging in gender non-

normative forms of aggression (relational aggression for

males and physical aggression for females) is associated

with higher levels of psychological stress (Crick 1997). It

244 J Youth Adolescence (2013) 42:235–249

123

may be that the social sanctions against relational aggres-

sion are stronger for males than for females, and that

beliefs supporting aggression must be stronger for males in

order for them to act out relationally. This is an area that

further research should explore.

Interpersonal Climate

Previous studies suggest that a more positive interpersonal

climate, indicated here by student–student relationships

and student–teacher relationships, are predictive of lower

levels of physical aggression (Hartup and Stevens 1997;

Hughes 1999). As expected, both student–student rela-

tionships and student–teacher relationships were predictive

of lower levels of relational aggression. The relationship to

victimization, however, diverged from our expectations.

While positive student–student relationships were associated

negatively with relational victimization, student–teacher

relationships were not related to relational victimization.

One possible explanation for why positive teacher

relations did not impact relational victimization is that

teachers may view acts of relational aggression as norma-

tive (Casey-Cannon et al. 2001) and have more lenient

responses to such acts (Bauman and Del Rio 2006). The

differential effects of student–teacher relationships on

relational aggression and relational victimization also may

be due to these being individual-level predictors. School-

levels of interpersonal climate measures could be expected

to have similar effects on relational aggression and vic-

timization levels because they reflect a system where

aggression and victimization levels should correspond. At

the individual-level, however, a student’s perpetration of

relational aggression could be influenced more powerfully

by teacher relationships than the experience of relational

victimization.

School Responsiveness to Violence

We hypothesized that higher levels of school responsive-

ness to violence, in this study conceptualized as awareness

and reporting of violence and school safety concerns,

would predict lower levels of relational aggression and

victimization. Surprisingly, awareness and reporting were

not related to relational aggression or victimization. One

explanation for this lack of relationship is that teachers and

students may not be as aware of and likely to report rela-

tional aggression as other forms of aggression. Studies

have suggested that teachers often see relational aggression

as normative and are less likely to intervene than with

physical aggression (Craig et al. 2000). Additionally, many

teachers may not be aware of relational aggression due to

its covert nature (Crick and Grotpeter 1995). Students may

perceive teachers as responding to other forms aggression

that may be unrelated to their acts of relational aggression;

in these cases, higher perceptions of awareness and

reporting would not impact relational aggression.

In regard to school safety, as predicted, youth who

perceived higher levels of school safety concerns exhibited

higher levels of relational aggression and higher levels of

relational victimization. These results support the argument

that, parallel to physical aggression, students who perceive

greater school safety problems are more likely to be vic-

timized by and perpetrate acts of relational aggression

(Astor et al. 2005). This is in line with research that shows

that students who are exposed to violence in school expe-

rience higher psychological problems (Flannery et al.

2004), which may lead to increased behavioral problems.

Limitations and Future Directions

Although the present study demonstrates robust results

concerning the relationship between perceptions of the

school environment and the experience of relational

aggression, these findings must be interpreted with con-

sideration of the study’s limitations. One limitation is

related to the measures used in this study. This study relied

on self-reported experiences of relational aggression and

victimization, and some researchers have argued that self-

reports of these acts may result in bias due to a desire to

hide these experiences from adults (Bjorkqvist et al. 1992).

However, studies have shown that teachers may not be

aware of relationally aggressive acts when they are

occurring (Card et al. 2008); for this reason, self-reports

may be more valid than teacher reports of relational

aggression and victimization. Future research should

explore the effect of the school environment on relational

aggression drawing on multiple informants for relational

aggression. Additionally, our measure of norms favoring

aggression is in reference to verbal and physical aggres-

sion, yet research has indicated that relational aggression

may be particularly sensitive to norms related to specifi-

cally relational aggression (Werner and Nixon 2005;

Werner and Hill 2010). Because of this, the present study

may provide a conservative estimate of the effects of

individual norms on relational aggression. Finally, a

number of scales used in the present study had compara-

tively low levels of internal consistency. This would tend

to increase Type II error, or the likelihood that our study

would detect real effects.

Another limitation is that we drew on data from a pre-

vention trial in which most youth in our study participated.

Both the predictors and outcomes in this study may have

been affected by the interventions. We attempted to

address this limitation by including dummy variables for

the interventions with the control condition as the

J Youth Adolescence (2013) 42:235–249 245

123

comparison. The fact that the direction of our effects did

not substantially differ by intervention condition provides

some evidence that our sample is not skewed by inter-

vention participation. Related to the prevention trial, our

sample was drawn from schools serving low-income pop-

ulations and may not be representative of the broader

population. Therefore, our results must be interpreted with

caution and not assumed to be applicable to more affluent

samples.

In spite of these limitations, our results collectively

present evidence that unlike physical aggression, school-

level indicators of school climate do not add to the

explanation of involvement in relational aggression and

victimization. Rather, it is perceptions of the school envi-

ronment and individual beliefs that are related to risk.

Students who hold beliefs favoring nonviolence, who per-

ceive positive interpersonal relations in school, and who

perceive their school to be safe are less at risk for vic-

timization by and perpetration of relational aggression. Our

findings that responsiveness to violence and school-level

indicators of school climate did not affect relational

aggression notably differed from the literature on physical

aggression (Henry et al. 2011).

These results have implications for the prevention of

relational aggression. While a recent review of programs

targeting relational aggression (Leff et al. 2010) found

that none met the high standards for efficacious set forth

by the Society for Prevention Research (2006), some

interventions have shown promise. A relatively brief

school-based bullying prevention intervention, Steps to

Respect, was correlated with a reduction in the prevalence

of relationally aggressive acts (Low et al. 2010). Steps to

Respect: A Bullying Prevention Program, a universal

school-wide intervention specifically targeting relational

aggression (Frey et al. 2005), aims to create positive

norms through implementing school policies about bul-

lying, to increase teacher responsiveness to relational

aggression through teacher training, and to enhance

healthy interactions among youth through teaching social

skills. Our findings support interventions such as Steps to

Respect that focus on creating a safer school climate

through fair policies as well as improving the relation-

ships between teachers and students and among students.

With studies showing that from an early age, youth see

relational aggression as more acceptable than physical

aggression (Goldstein et al. 2002), targeting individual

beliefs, particularly those that support the use of nonvi-

olent alternatives to aggression, is important to changing

rates of relational aggression.

The present study highlights the important role that

perceptions of the school environment and individual

beliefs about aggression play for relational aggression and

victimization. Researchers in adolescent development have

established the powerful influence of school climate on

physical aggression (Henry et al. 2011; Kuperminc et al.

1997). This study provides evidence that, for relational

aggression, it is perceptions of—and not school-level

measures of—the school environment that matter: stu-

dents’ perceptions of school safety and interpersonal cli-

mate influence both victimization by and perpetration of

relational aggression. Our finding that beliefs supporting

nonviolence predicted lower levels of relational aggression

over and above the influence of beliefs favoring aggression

builds on previous work linking beliefs supporting

aggression to relational aggression (Kuppens et al. 2008;

Onishi et al. 2011) and represents a potentially new point

of intervention. Together, these results underscore the

promise of research that explores how perceptions of the

social environment impact relational aggression and

victimization.

Acknowledgments This research was funded by Grant # U49CE001296 from the Centers for Disease control and Prevention.

We are grateful for the contributions of the originators of the Multisite

Violence Prevention Project (MVPP), the talented and dedicated team

that made this project possible. The members of the MVPP project

include: Centers for Disease Control and Prevention, Atlanta GA:

Thomas R. Simon, PhD; Robin M. Ikeda, MD, MPH (National Center

for Injury Prevention and Control; Emilie Phillips Smith, PhD (Penn

State University); Le’Roy E. Reese, PhD (Morehouse School of

Medicine); Duke University, Durham NC: David L. Rabiner, PhD;

Shari Miller-Johnson, PhD; Donna-Marie Winn, PhD (University of

North Carolina—Chapel Hill); Kenneth A. Dodge, PhD (Center for

Child and Family Policy); Steven R. Asher, PhD (Department of

Psychology and Neuroscience); University of Georgia, Athens GA:

Arthur M. Horne, PhD (Department of Counseling and Human

Development Services); Pamela Orpinas, PhD (Department of Health

Promotion and Behavior); Roy Martin, PhD (Dept. of Educational

Psychology and Instructional Technology); William H. Quinn, PhD

(Clemson University); University of Illinois at Chicago, Chicago IL:

Patrick H. Tolan, PhD (University of Virginia); Deborah Gorman-

Smith, PhD (University of Chicago); David B. Henry, PhD; Franklin

N. Gay, MPH, Michael Schoeny, PhD; Virginia Commonwealth

University, Richmond VA: Albert D. Farrell, PhD; Aleta L Meyer,

PhD (National Institute on Drug Abuse); Terri N. Sullivan, PhD;

Kevin W. Allison, PhD (all Department of Psychology).

Author Contributions CE conceived of the study, performed the statistical analysis, and drafted the manuscript; DGS participated in

the design and coordination of the study and helped conceive of the

study and draft the manuscript; DH participated in the design and

coordination of the study, helped with the statistical analysis, and

helped draft the manuscript. All authors read and approved the final

manuscript.

References

Astor, R. A., Meyer, H. A., Benbenishty, R., Marachi, R., &

Rosemond, M. (2005). School safety interventions: Best prac-

tices and programs. Children and Schools, 27, 17–32. doi:10.1093/cs/27.1.17.

246 J Youth Adolescence (2013) 42:235–249

123

Astor, R. A., Meyer, H. A., & Pitner, R. O. (2001). Elementary and

middle school students’ perceptions of violence-prone school

subcontexts. The Elementary School Journal, 101(5), 511–528. Bandura, A. (1986). Social foundations of thought and action: A

social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Barth, J. M., Dunlap, S. T., Dane, H., Lochman, J. E., & Wells, K. C.

(2004). Classroom environment influences on aggression, peer

relations, and academic focus. Journal of School Psychology, 42, 115–133. doi:10.1016/j.jsp.2003.11.004.

Bauman, S., & Del Rio, A. (2006). Preservice teachers’ responses to

bully scenarios: Comparing physical, verbal, and relational

bullying. Journal of Educational Psychology, 98, 219–231. doi:10.1037/0022-0663.98.1.219.

Birch, S. H., & Ladd, G. W. (1998). Children’s interpersonal

behaviors and the teacher-child relationship. Developmental Psychology, 34, 934–946. doi:10.1037/0012-1649.34.5.934.

Bjorkqvist, K., Osterman, K., & Kaukiainen, A. (1992). The

development of direct and indirect aggressive strategies in males

and females. In K. Bjorkqvist, & P. Niemela (Eds.), Of mice and women. Aspects of female aggression (pp. 51–64). San Diego: Academic Press.

Bowen, N. K., & Bowen, G. L. (1999). Effects of crime and violence

in neighborhoods and schools on the school behavior and

performance of adolescents. Journal of Adolescent Research, 14(3), 319–342. doi:10.1177/0743558499143003.

Bronfenbrenner, U. (1986). Ecology of the family as a context for

human development. Developmental Psychology, 22, 521–530. Card, N. A., Stucky, B. D., Sawalani, G. M., & Little, T. D. (2008).

Direct and indirect aggression during childhood and adoles-

cence: A meta-analytic review of gender differences, intercor-

relations, and relations to maladjustment. Child Development, 79, 1185–1229. doi:10.1111/j.1467-8624.2008.01184.x.

Casey-Cannon, S., Hayward, C., & Gowen, K. (2001). Middle-school

girls’ reports of peer victimization: Concerns, consequences, and

implications. Professional School Counseling, 5, 138–147. Chang, L. (2003). Variable effects of children’s aggression, social

withdrawal, and prosocial leadership as functions of teacher

beliefs and behaviors. Child Development, 74(2), 535–548. doi:10.1111/1467-8624.7402014.

Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory

of normative conduct: A theoretical refinement and reevaluation

of the role of norms in human behavior. In M. P. Zanna (Ed.),

Advances in experimental social psychology (pp. 201–234). New York: Academic Press.

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 public places. Journal of Personality and Social Psychology, 58, 305–313.

Craig, W. M., Henderson, K., & Murphy, J. G. (2000). Prospective

teachers’ attitudes toward bullying and victimization. School Psychology International, 21, 5–21. doi:10.1177/01430343 00211001.

Crick, N. R. (1997). Engagement in gender normative versus

nonnormative forms of aggression: Links to social-psychological

adjustment. Developmental Psychology, 33(4), 610–617. doi:10.1037/0012-1649.33.4.610.

Crick, N. R., Bigbee, M. A., & Howes, C. (1996). Gender differences

in children’s normative beliefs about aggression: How do I hurt

thee? Let me count the ways. Child Development, 67, 1003–1014. doi:10.1111/j.1467-8624.1996.tb01779.x.

Crick, N. R., Casas, J. F., & Ku, H.-C. (1999a). Relational and

physical forms of peer victimization in preschool. Develop- mental Psychology, 35, 376–385. doi:10.1037/0012-1649.35. 2.376.

Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender,

and social-psychological adjustment. Child Development, 66, 710–722. doi:10.1111/j.1467-8624.1995.tb00900.x.

Crick, N. R., Werner, N. E., Casas, J. F., O’Brien, K. M., Nelson, D.

A., Grotpeter, J. K., et al. (1999b). Childhood aggression and

gender: A new look at an old problem. In D. Bernstein (Ed.), The Nebraska symposium on motivation, gender and motivation (Vol. 45, pp. 75–141). Omaha: University of Nebraska Press.

Crockett, L. J., & Petersen, A. C. (1993). Adolescent development:

Health risks and opportunities for health promotion. In S. G. Mill-

stein, A. C. Petersen, & E. O. Nightingale (Eds.), Promoting the health of adolescents: New directions for the twenty-first century (pp. 13–37). New York: Oxford University Press.

Cullerton-Sen, C., & Crick, N. R. (2005). Understanding the effects of

physical and relational victimization: The utility of multiple

perspectives in predicting social–emotional adjustment. School Psychology Review, 34, 147–160.

Dowdell, E. B., & Santucci, M. E. (2003). The relationship between health

risk behaviors and fear in one urban seventh grade class. Journal of Pediatric Nursing, 18, 187–194. doi:10.1053/jpdn.2003.37.

Farrell, A. D., Kung, E. M., White, K. S., & Valois, R. (2000). The

structure of self-reported aggression, drug use, and delinquent

behaviors during early adolescence. Journal of Clinical Child Psychology, 29, 282–292. doi:10.1207/S15374424jccp2902_13.

Flannery, D. J., Wester, K. L., & Singer, M. I. (2004). Impact of

exposure to violence in school on child and adolescent mental

health and behavior. Journal of Community Psychology, 32, 559–573. doi:10.1002/jcop.20019.

Frey, K. S., Hirschstein, M. K., Snell, J. L., Edstrom, L. V.,

MacKenzie, E. P., & Broderick, C. J. (2005). Reducing

playground bullying and supporting beliefs: An experimental

trial of the steps to respect program. Developmental Psychology, 41, 479–491. doi:10.1037/0012-1649.41.3.479.

Galen, B. R., & Underwood, M. K. (1997). A developmental

investigation of social aggression among children. Developmen- tal Psychology, 33, 589–600. doi:10.1037/0012-1649.33.4.589.

Goldstein, S. E., Tisak, M. S., & Boxer, P. (2002). Preschoolers’

normative and prescriptive judgments about relational and overt

aggression. Early Education and Development, 13, 23–39. doi:10.1207/s15566935eed1301_2.

Gottfredson, G. D., & Gottfredson, D. C. (1985). Victimization in schools. New York: Plenum.

Guerra, N. G., Huesmann, L. R., & Hanish, L. (1994). The role of

normative beliefs in children’s social behavior. In N. Eisenberg

(Ed.), Review of personality and social psychology, development and social psychology: The interface (pp. 140–158). London: Sage.

Hartup, W., & Stevens, N. (1997). Friendships and adaptation in the

life course. Psychological Bulletin, 121(3), 355–370. doi:10.1037/0033-2909.121.3.355.

Henry, D. B., Dymnicki, A. B., Schoeny, M. E., Meyer, A. C., Martin,

N. C., & the Multisite Violence Prevention Project (in press).

Middle school students overestimate normative support for

aggression and underestimate normative support for non-violent

problem-solving strategies. Journal of Applied Social Psychology.

Henry, D. B., Farrell, A. D., Schoeny, M. E., Tolan, P. H., & Dymnicki, A. B. (2011). Influence of school-level variables on

aggression and associated attitudes of middle school students.

Journal of School Psychology, 49, 481–503. doi:10.1016/j. jsp.2011.04.007.

Henry, D., Guerra, N. G., Huesmann, L. R., Tolan, P. H., VanAcker, R.,

& Eron, L. D. (2000). Normative influences on aggression in urban

elementary school classrooms. American Journal of Community Psychology, 28, 59–81. doi:10.1023/A:1005142429725.

J Youth Adolescence (2013) 42:235–249 247

123

Herrenkohl, T. I., Catalano, R. F., Hemphill, S. A., & Toumbourou, J.

W. (2009). Longitudinal examination of physical and relational

aggression as precursors to later problem behaviors in adoles-

cents. Violence and Victims, 24(1), 3–19. doi:10.1891/0886- 6708.24.1.3.

Huesmann, L. R. (1988). An information processing model for the

development of aggression. Aggressive Behavior, 14, 13–24. doi: 10.1002/1098-2337(1988)14:1\13:AID-AB2480140104[3.0. CO;2-J.

Huesmann, R. L., & Guerra, N. G. (1997). Children’s normative beliefs

about aggression and aggressive behavior. Journal of Personality and Social Psychology, 72(2), 408–419. doi:10.1037/0022- 3514.72.2.408.

Hughes, J. N., Cavell, T. A., & Jackson, T. (1999). Influence of the

teacher–student relationship on childhood conduct problems: A

prospective study. Journal of Clinical Child Psychology, 28(2), 173–184. doi:10.1207/s15374424jccp2802_5.

Jessor, R. (1993). Successful adolescent development among youth in

high-risk settings. American Psychologist, 48, 117–126. doi:10.1037/0003-066X.48.2.117.

Kuperminc, G. P., Leadbeater, B. J., & Blatt, S. J. (2001). School

social climate and individual differences in vulnerability to

psychopathology among middle school students. Journal of School Psychology, 39, 141–159. doi:10.1016/S0022-4405(01) 00059-0.

Kuperminc, G. P., Leadbeater, B. J., Emmons, C., & Blatt, S. J.

(1997). Perceived school climate and difficulties in the social

adjustment of middle school students. Applied Developmental Science, 1(2), 76–88. doi:10.1207/s1532480xads0102_2.

Kuppens, S., Grietens, H., Onghena, P., Michiels, D., & Subramanian,

S. V. (2008). Individual and classroom variables associated with

relational aggression in elementary-school aged children: A

multilevel analysis. Journal of School Psychology, 46, 639–660. Leff, S. S., Waasdorp, T. E., & Crick, N. R. (2010). A review of existing

relational aggression programs: Strengths, limitations, and future

directions. School Psychology Review, 39(4), 508–535. Low, S., Frey, K. S., & Brockman, C. J. (2010). Gossip on the

playground: Changes associated with universal intervention,

retaliation beliefs, and supportive friends. School Psychology Review, 39(4), 536–551.

Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge: Belknap Press of Harvard University Press.

Meyer, A. L., Allison, K. W., Reese, L. E., Gay, F. N., & The Multisite

Violence Prevention Project. (2004). Choosing to be violence free

in middle school: The student component of the GREAT schools

and families universal program. American Journal of Preventive Medicine, 26, 20–29. doi:10.1016/j.amepre.2003.09.014.

Meyer, A. L., Farrell, A. D., Northup, W. B., Kung, E. M., & Plybon,

L. (2000). Promoting nonviolence in early adolescence: Responding in peaceful and positive ways. Dordrecht: Kluwer Academic.

Miller-Johnson, S., Sullivan, T. N., Simon, T. R., & The Multisite

Violence Prevention Project. (2004). Evaluating the impact of

interventions in the multisite violence prevention study: Sam-

ples, procedures, and measures. American Journal of Preventive Medicine, 26, 48–61. doi:10.1016/j.amepre.2003.09.015.

Multisite Violence Prevention Project (MVPP). (2004). The multisite

violence prevention project: Background and overview. Amer- ican Journal of Preventive Medicine, 26, 3–11. doi:10.1016/ j.amepre.2003.09.017.

Onishi, A. Kawabata, Y. Kurokawa, M., & Yoshida, T. (2011). A

mediating model of relational aggression, narcissistic orientations,

guilt feelings, and perceived classroom norms. School Psychology International, 0(0), 1–24. doi:10.1177/0143034311421433.

Orpinas, P., Horne, A. M., & The Multisite Violence Prevention

Project. (2004). A teacher-focused approach to prevent and

reduce students? Aggressive behavior: The GREAT teacher

program. American Journal of Preventive Medicine, 26, 29–38. doi:10.1016/j.amepre.2003.09.016.

Parker, J. G., & Asher, S. R. (1993). Friendship and friendship quality in

middle school childhood: Links with peer group acceptance and

feelings of loneliness and social dissatisfaction. Developmental Psychology, 29, 621–622. doi:10.1037/0012-1649.29.4.611.

Preddy, T. M., & Fite, P. J. (2012). Differential associations between

relational and overt aggression and children’s psychosocial

adjustment. Journal of Psychopathology and Behavioral Assess- ment, 34(1), 182–190. doi:10.1007/s10862-011-9274-1.

Roeser, R. W., & Eccles, J. S. (1998). Adolescents’ perceptions of

middle school: Relation to longitudinal changes in academic and

psychological adjustment. Journal of Research on Adolescence, 8, 123–158. doi:10.1207/s15327795jra0801_6.

Savin-Williams, R. C., & Berndt, T. J. (1990). Friendship and peer

relationships. In S. S. Feldman & G. R. Elliott (Eds.), At the threshold: The developing adolescent (pp. 277–307). Cambridge, MA: Harvard University Press.

Smith, E. P., Gorman-Smith, D., Quinn, W. H., Rabiner, D. L., Tolan, P.

H., Winn, D., et al. (2004). Community-based multiple family

groups to prevent and reduce violent and aggressive behavior: The

GREAT families program. American Journal of Preventive Med- icine, 26(Supplement 1), 39–47. doi:10.1016/j.amepre.2003.09.018.

Society for Prevention Research. Standards of evidence: Criteria for

efficacy, effectiveness and dissemination. (2006). Retrieved

from http://www.preventionscience.org/StandardsofEvidencebook.

pdf.

Sullivan, T. N., Farrell, A. D., & Kliewer, W. (2006). Peer

victimization in early adolescence: Associations between phys-

ical and relational victimization and drug use, aggression, and

delinquent behaviors among urban middle school students.

Development and Psychopathology, 18, 119–137. doi:10.1017/ S095457940606007X.

Underwood, M. K. (2003). Social aggression among girls. New York: Guilford Press.

US Department of Education. (1999–2000). Public school teacher questionnaire: Schools and staffing survey—1999–2000 school year. Web site: http://nces.ed.gov/surveys/sass/.

Vessels, G. (1998). Character and community development: A school planning and teacher training handbook. Westport, CT: Praeger.

Waasdorp, T. E., & Bradshaw, C. P. (2009). Child and parent

perceptions of relational aggression within urban predominantly

African American children’s friendships: Examining patterns of

concordance. Journal of Child and Family Studies, 18(6), 731–745. doi:10.1007/s10826-009-9279-5.

Werner, N., & Hill, L. (2010). Individual and peer group normative

beliefs about relational aggression. Child Development, 81, 826–836. doi:10.1111/j.1467-8624.2010.01436.x.

Werner, N. E., & Nixon, C. L. (2005). Normative beliefs and

relational aggression: An investigation of the cognitive bases of

adolescent aggressive behavior. Journal of Youth and Adoles- cence, 34, 229–243. doi:10.1007/s10964-005-4306-3.

Xie, H., Cairns, R. B., & Cairns, B. D. (2002). The development of

social aggression and physical aggression: A narrative analysis

of interpersonal conflicts. Aggressive Behavior, 28, 341–355. Xie, H., Farmer, T. W., & Cairns, B. D. (2003). Different forms of

aggression among inner-city African-American children: Gen-

der, configurations, and school social networks. Journal of School Psychology, 41, 355–375. doi:10.1016/S0022-4405 (03)00086-4.

Yoon, J., & Kerber, K. (2003). Bullying: Elementary teachers’

attitudes and intervention strategies. Research in Education, 69, 27–35.

Zelli, A., Dodge, K. A., Lochman, J. E., & Laird, L. D. (1999). The

distinction between beliefs legitimizing aggression and deviant

248 J Youth Adolescence (2013) 42:235–249

123

processing of social cues: Testing measurement validity and the

hypothesis that biased processing mediates the effects of beliefs

on aggression. Journal of Personality and Social Psychology, 77, 150–166. doi:10.1037/0022-3514.77.1.150.

Author Biographies

Caitlin Elsaesser is a doctoral student at the School of Social Service Administration, University of Chicago. She received a Masters of

Arts from the School of Social Service Administration, University of

Chicago and a Master of Arts in Teaching from Dominican

University. Her research interests include community, school and

family-level influences on adolescent victimization, aggression and

school success.

Deborah Gorman-Smith is a Professor at the School of Social Service Administration, University of Chicago. She received her

Ph.D. in Clinical-Developmental Psychology from the University of

Illinois at Chicago. Her research is focused on advancing knowledge

about development, risk, and prevention of aggression and violence,

with specific focus on minority youth living in poor urban settings.

David Henry is a Professor of Health Policy and Administration at the Institute for Health Research and Policy and the University of

Illinois Chicago. He received his Ph.D from University of Illinois at

Chicago in Psychology and studies contextual processes that influence

individual behavior, child and adolescent development and psycho-

pathology, and prevention.

J Youth Adolescence (2013) 42:235–249 249

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.