ArticlePDF Available

Authoritative School Discipline: High School Practices Associated With Lower Bullying and Victimization

May 2010
Journal of Educational Psychology 102(2):483-496

DOI:10.1037/a0018562

Authors:

Anne Gregory

Rutgers, The State University of New Jersey

Dewey Cornell

University of Virginia

Peter Sheras

University of Virginia

Show all 6 authorsHide

In this study we examined authoritative discipline theory, which posits that 2 complementary aspects of school climate—structure and support—are important for adolescents' safety in school. Using a statewide sample of over 7,300 ninth-grade students and 2,900 teachers randomly selected from 290 high schools, we showed, using hierarchical linear modeling, that consistent enforcement of school discipline (structure) and availability of caring adults (support) were associated with school safety. Structure and support were associated with less bullying and victimization after we controlled for size of school enrollment and the proportion of ethnic minority and low-income students. These findings suggest that discipline practices should not be polarized into a “get tough” versus “give support” debate because both structure and support contribute to school safety for adolescents. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Content uploaded by Anne Gregory

Content may be subject to copyright.

Authoritative School Discipline:

High School Practices Associated With Lower Bullying and Victimization

Anne Gregory, Dewey Cornell, Xitao Fan, Peter Sheras, Tse-Hua Shih, and Francis Huang

University of Virginia

In this study we examined authoritative discipline theory, which posits that 2 complementary aspects of

school climate—structure and support—are important for adolescents’ safety in school. Using a statewide

sample of over 7,300 ninth-grade students and 2,900 teachers randomly selected from 290 high schools,

we showed, using hierarchical linear modeling, that consistent enforcement of school discipline (struc-

ture) and availability of caring adults (support) were associated with school safety. Structure and support

were associated with less bullying and victimization after we controlled for size of school enrollment and

the proportion of ethnic minority and low-income students. These findings suggest that discipline

practices should not be polarized into a “get tough” versus “give support” debate because both structure

and support contribute to school safety for adolescents.

Keywords: adolescence, high schools, at-risk students, learning environments, classroom management

There is currently a wide disparity in high school discipline

practices, ranging from schools that demand behavioral conformity

and compliance to those that emphasize student autonomy and

independent decision making (Stronach & Piper, 2008). Reforms

in discipline policies range from the systematic reinforcement of

positive behavior (Bohanon et al., 2006) to automatic expulsion for

an ever-widening list of offenses (American Psychological Asso-

ciation Zero Tolerance Task Force, 2006). Despite the variety of

approaches and reforms, there is little research identifying the

characteristics of a safe high school environment.

Safety Problems in Schools

School safety is not a problem confined to a few troubled

schools. Data from multiple sources using different methods indi-

cate the pervasiveness of the problem. According to principal

reports on the national School Survey on Crime and Safety (Na-

tional Center for Educational Statistics [NCES], 2007), 95% of

American high schools experienced at least one violent crime in

2005–2006. According to victim reports from the National Crime

Victimization Survey (NCES, 2007), approximately 1.5 million

crimes were committed against students (ages 12–18) at school in

2005, including 136,500 serious violent crimes. Although school

crime rates have fallen over the past 10 years, the overall rate at

school (57 per 1,000 students) remains higher than the rate away

from school (47 per 1,000; NCES, 2007). Bullying and fighting,

which typically are not counted in crime statistics, are even more

pervasive. Approximately 28% of ninth-grade students reported

being victims of bullying at school in the past 6 months, including

21% who reported a physical injury (NCES, 2007). In 2005, 14%

of Grades 9 –12 students reported being in a physical fight on

school property (NCES, 2007). The consequences of student vic-

timization include depression (Hawker & Boulton, 2000), low

academic performance (Holt, Finkelhor, & Kantor, 2007), and a

diminished sense of academic belonging (Holt & Espelage, 2003).

Teacher surveys also document the extent of the problem.

Thirty-five percent of teachers reported that student misbehavior

interfered with their teaching (NCES, 2007). Approximately 7.5%

of secondary teachers reported being threatened with physical

injury, and 2.3% reported being physically attacked by a student in

2003–2004 (NCES, 2007).

School discipline sanctions are another, albeit imprecise, indi-

cation of school safety (Heaviside, Rowand, Williams, & Farris,

1998; Morrison, Redding, Fisher, & Peterson, 2006). According to

the Indicators of School Crime and Safety (NCES, 2007), 48% of

public schools took serious disciplinary action (ranging from sus-

pensions of at least 5 days to expulsion from school) for incidents

such as fights and possession of a weapon. Approximately three

million students receive school suspensions, and 97,000 students

are expelled each year (U.S. Department of Education, 2000).

High schools differ in the degree to which they are safe envi-

ronments for students (D. C. Gottfredson, 2001). Several demo-

graphic risk factors have been identified in previous research.

Larger schools tend to have more fights and suspensions (Stewart,

Anne Gregory, Dewey Cornell, Xitao Fan, Peter Sheras, Tse-Hua Shih,

and Francis Huang, Curry School of Education, University of Virginia.

Francis L. Huang is now at the Phonological Awareness Literacy

Screening Office, University of Virginia.

This project was supported in part by a grant from the Office of Juvenile

Justice and Delinquency Prevention of the U.S. Department of Justice, but

the views in this article do not necessarily reflect policies or recommen-

dations of the funding agency. We thank Donna Bowman Michaelis of the

Virginia Department of Criminal Justice Services and Arlene Cundiff of

the Virginia Department of Education, and their colleagues, for their

support of the Virginia High School Safety Study. We also thank our

research assistants Sharmila Bandyopadhyay, Justin Collman, Megan

Eliot, Jennifer Klein, Talisha Lee, Erica Shirley, Aisha Thompson, and

Farah Williams.

Correspondence concerning this article should be addressed to Anne

Gregory, who is now at the Graduate School of Applied and Professional

Psychology, Rutgers University, 152 Frelinghuysen Road, Piscataway, NJ

08854. E-mail: annegreg@rci.rutgers.edu

2010, Vol. 102, No. 2, 483–496 0022-0663/10/$12.00 DOI: 10.1037/a0018562

483

2003) as well as greater school crime and victimization (D. C.

Gottfredson, 2001). Compared to schools with a greater proportion

of high-income students, schools with more low-income students

tend to have more victimization (Bauer, Guerino, Nolle, & Tang,

2008; Khoury-Kassabri, Benbenishty, Astor, & Zeira, 2004) and

more frequent fighting and suspensions (Stewart, 2003). In gen-

eral, family poverty is consistently linked to youth violence (e.g.,

LeBlanc, Swisher, Vitaro, & Tremblay, 2008; Devoe, Peter,

Noonan, Snyder, & Baum, 2005; U.S. Department of Health and

Human Services, 2001). The proportion of non-White students in

a school is also regarded as a risk factor for school violence (G. D.

Gottfredson, Gottfredson, Payne, & Gottfredson, 2005), but typical

measures of socioeconomic status do not adequately distinguish

race from social class (Hosp & Reschly, 2004). These demo-

graphic risk factors are important, but they are not determinative of

school safety, and research is needed to identify school practices

that maintain safety even under challenging conditions.

Developmental Theory and School Climate

Adolescent Developmental Needs

For decades, adolescence has been characterized as a time when

youth shift their focus from family to peers (Selman & Hickey

Schultz, 1990; Sullivan, 1953) and from dependence on adults to

autonomy and relative self-reliance (O’Connor, Allen, Bell, &

Hauser, 1996; Smetana & Gaines, 1999). However, to assert that

adolescents need little from adults would be mistaken. Instead, a

substantial body of research has shown that positive adolescent

development depends on adult relationships (e.g., Maccoby &

Martin, 1983; Steinberg, Lamborn, Dornbusch, & Darling, 1992),

although the nature of the relationship differs from the earlier

parent– child dyad. Since Baumrind (1968) presented her landmark

typology of parenting that contrasted authoritative and authoritar-

ian styles, researchers have identified two central dimensions of

effective parenting of adolescents: (a) structure—strictness and

close supervision as reflected in parental monitoring and limit

setting and (b) support—parental warmth, acceptance, and in-

volvement (Herman, Dornbusch, Hen-on, & Herting, 1997; Stein-

berg, Lamborn, Darling, & Mounts, 1994; Steinberg et al., 1992).

This conception of parenting provides the core theoretical basis for

the present study.

In combination, high structure (firm enforcement of rules) and

support (responsiveness to children’s developmental needs) com-

prise what Baumrind (1968) has called authoritative parenting.

Early studies supported the benefits of this parenting style for

achievement outcomes in a predominantly Caucasian adolescent

sample (Baumrind, 1968, 1991). Since then, numerous researchers

have examined whether an authoritative parenting style benefits

adolescents from different racial and ethnic groups. Evidence has

been somewhat mixed. Several studies have shown that authorita-

tive parenting is more consistently beneficial to Caucasian adoles-

cents’ achievement compared to the achievement of adolescents

from other racial and ethnic groups (Park & Bauer, 2002; Stein-

berg et al., 1994). Some argue that an authoritarian parenting style,

which emphasizes close monitoring and strict discipline, may be

more culturally congruent for Asian American adolescents or more

protective of African American adolescents living in dangerous

neighborhoods (Baumrind, 1995; Furstenberg, Eccles, Elder,

Cook, & Sameroff, 1997; Gonzales, Cauce, Friedman, & Mason,

1996; Kelley, Power, & Winbush, 1992). That said, a collection of

other studies has found support for the benefits of an authoritative

parenting style for adolescents of color. For instance, authoritative

parenting predicted higher educational aspirations in low-income

Hispanic and African American youth (Gorman-Smith, Tolan, &

Henry, 2000), lower problem behavior with adolescents from

low-income urban families (Shumow, Vandell, & Posner, 1999),

and higher academic grades (Taylor, Hinton, & Wilson, 1995) and

self-esteem (Mandara & Murray, 2002) for African American

teens. Taken together, these studies suggest that, for adolescents

from different racial and ethnic groups, authoritative parents may

meet their children’s developmental needs with structure that

includes establishing clear rules, monitoring behavior, and enforc-

ing rules consistently, yet also support their children with warmth

and encouragement.

The model of authoritative parenting provides an illuminating

theoretical perspective on discipline practices in schools. Baum-

rind (1996) wrote

Within the authoritative model, behavioral compliance and psycho-

logical autonomy are viewed not as mutually exclusive but rather as

interdependent objectives: Children are encouraged to respond habit-

ually in prosocial ways and to reason autonomously about moral

problems, and to respect adult authorities and learn how to think

independently. (p. 405)

Respect for and cooperation with authority, according to Baum-

rind (1996), should be nurtured along with autonomous reasoning

and independent thinking. This is particularly important for ado-

lescents as they seek greater control in decision making (Smetana

& Gaines, 1999) and expect fair and legitimate adult authority

(Turiel, 2005). Adolescents are especially sensitive to issues of

fairness and autonomy, so that efforts to manage and control their

behavior must be tempered with efforts to demonstrate that they

are regarded with respect.

Darling and Steinberg (1993) theorized that authoritative par-

enting develops an emotional climate between adolescents and

their parents that fosters adolescents’ openness to parental social-

ization. In other words, the positive climate of authoritative par-

enting may help students to be more responsive to parenting

behaviors. In a school setting, an authoritative approach to disci-

pline combines both firm enforcement of school rules and a

concerted effort to communicate warmth and concern for the

well-being of each student as an individual. In theory, authoritative

discipline in school could offer the right conditions for student

cooperation with school rules and safe interactions among stu-

dents. Like with authoritative parents, students may be more open

and responsive to school administrator and teacher efforts to

uphold school rules and redirect misbehavior.

Authoritative Teaching

At the classroom level, research has shown that adolescents

benefit from teachers who use an approach similar to authoritative

parental guidance (Gregory & Weinstein, 2004; Walker, 2008;

Wentzel, 2002). Recent studies found that the combination of

teacher structure and support predicted growth in achievement,

especially for students experiencing the negative effects of low

family income (Gregory & Weinstein, 2004). Moreover, Gregory

484 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

and Weinstein (2008) showed that African American students with

negative discipline trajectories had greater acceptance of the au-

thority of teachers whom they perceived as caring and as main-

taining high academic expectations. Wentzel (2002) provided ev-

idence that teacher characteristics are related to behavioral

outcomes at the middle school level. She found that student per-

ceptions that their teachers maintained high expectations, yet gave

infrequent negative feedback, were associated with less irrespon-

sible student behavior as reported by the teacher. Moreover, qual-

itative research has found that teachers who are warm demanders

(Irvine, 2002) or compassionate disciplinarians (Vasquez, 1988)

build trusting relationships with students of color and low-income

students. These studies show the benefits of an authoritative ap-

proach by individual teachers, but these researchers have not

examined the joint influence of structure and support at the school

level.

The present study extends the concepts of authoritative parent-

ing and teaching to the level of schoolwide discipline. In this

study, we test a new theory of school discipline policies and

practices called authoritative school discipline (Gregory & Cor-

nell, 2009). According to this theory, neither structure nor support

alone is sufficient to maintain a safe and orderly school climate; in

contrast, an authoritative approach with both structure and support

has an optimal impact on school safety. Pellerin (2005) applied a

similar model to high schools using 1990 and 1992 archival data

from the High School Effectiveness Study. She found that schools

using authoritative practices had less truancy and fewer dropouts

than did schools using an authoritarian approach.

Structure

We conceptualize school structure as the degree to which

schools consistently and fairly enforce rules. Undoubtedly, schools

must maintain sufficient order to ensure an environment conducive

to learning, but there is great variation in how schools approach

this task. Permissive schools that tolerate a wide range of student

behavior run the risk of suffering too much disorder, while schools

that seem too strict or unfair may elicit antagonistic responses from

adolescents who are developmentally inclined to challenge author-

ity and seek autonomy (Mayer & Leone, 1999; Smetana, 2005).

The student perspective is valuable for assessing the degree of

structure in schools. Decades of research on school climate has

highlighted the role of student perceptions of the school environ-

ment in positive youth development (Anderson, 1982; Moos &

Moos, 1978; Sprott, 2004; West, 1985). Adolescents’ perceptions

of the clarity and fairness of rules at their school are consistently

linked to better behavior (D. C. Gottfredson, Gottfredson, & Hybl,

1993; Hollingsworth, Lufler, & Clune, 1984; Welsh, 2000). Sim-

ilar results were found in a national sample of adolescents. Re-

search from the 1995 School Crime Supplement to the National

Crime Victimization Survey showed that adolescents who reported

greater understanding of school rules and consequences experi-

enced lower school crime and violence (Mayer & Leone, 1999).

The positive experience of school structure as clear and fair has

held for diverse groups of adolescents. G. D. Gottfredson and

colleagues (2005) found that student-perceived clarity and fairness

of rules predicted school-level differences in victimization and

delinquency. These school-level influences held beyond the effects

of student risk factors such as school rates of poverty.

Support

Adolescents’ perceptions of their teachers as caring and sup-

portive have been linked to higher grade point averages (Good-

enow, 1993), achievement growth (Gregory & Weinstein, 2004;

Hanson & Austin, 2003), and engagement in school (Maehr, 1991;

Midgley, Maehr, Hruda, Anderman, & Freeman, 2000). Availabil-

ity of adult assistance is a central aspect of school support. When

students feel that their teachers are caring and concerned, they are

more likely to seek help (Unnever & Cornell, 2004; Wilson &

Deanne, 2001). Help seeking fosters safer schools; victims of

bullying who seek help are less likely to experience revictimiza-

tion (Ladd & Ladd, 2001; Smith, Talamelli, & Cowie, 2004). In

addition, adult supports may be especially important for at-risk

adolescents. Croninger and Lee (2001) found that positive and

supportive relationships with teachers substantially reduced the

risk of dropping out for academically challenged students. Fur-

thermore, positive teacher–student relationships are associated

with lower use of weapons (Henrich, Brookmeyer, & Shahar,

2005) and reduced problem behavior (Jessor et al., 2003). Feeling

cared for and respected by adults in the school may elicit a greater

student willingness to cooperate with school rules and adult direc-

tion (Darling & Steinberg, 1993).

Present Study

Research on high school safety needs a guiding theory that is

developmentally grounded. Authoritative discipline theory offers a

promising framework to examine the conditions that are associated

with school safety. Like Pellerin (2005), we conceptualize schools

comprising multiple socializing agents (i.e., teachers, administra-

tors). The nature of their interactions with students is partially

shaped by the informal rules, formal policies, and cultural norms in

the school. Mutually reinforcing influences across the school ecol-

ogy compose the school discipline climate. A climate of structure

and support is developmentally appropriate for adolescents be-

cause adolescents need adult monitoring and clarity of rules and

expectations, yet they also need supportive adults who understand

their perspective. Clear and fair rules with adult support serves

adolescent autonomy needs. With support, adolescents will feel

their voices are heard and opinions seriously considered. With

structure that involves consistent and fair enforcement of school

rules, adolescents will experience legitimate and nonarbitrary use

of authority. With both structure and support in school, adolescents

may be more likely to cooperate with rules and to seek help when

needed. In these schools, both adults and adolescents may set a

tone of respect and establish norms against bullying and victim-

ization. This theoretical framework is promising, yet additional

empirical research is needed to test whether schools with author-

itative discipline are safer for adolescents.

Addressing this gap in the literature, we tested in the current study

the hypothesis that schools high on both structure and support would

be safer than other schools, as indicated by lower student victimiza-

tion and bullying. We selected two scales of structure, which we

conceptualized as students’ experiences of fair and consistently en-

forced school rules. From the students’ perspectives, school staff

did not overlook rule infractions in classrooms and hallways, and

consequences were issued in an even-handed and fair manner. We

selected two scales of support, which we conceptualized as stu-

485

AUTHORITATIVE SCHOOL DISCIPLINE

dents’ experience of the school staff as caring and helpful. Staff

were seen as open to hearing student difficulties and as effective in

providing help. We further hypothesized that structure and support

would be predictive of school safety after taking into account

demographic factors of school size, the racial/ethnic composition

of the enrolled students, and their eligibility for free- and reduced-

price meals.

School safety was assessed using three indicators: student re-

ports of being the victim of aggressive acts, student perceptions of

the extent of bullying and teasing at school, and teacher percep-

tions of the extent of bullying and teasing at school. We assessed

student victimization using an adaptation of the scale developed by

G. D. Gottfredson (1999) as part of the Effective School Battery.

This scale covers a range of victim experiences from being the

recipient of abusive remarks to being physically attacked and

injured. The victimization scale has been widely used in studies of

school climate and safety (e.g., G. D. Gottfredson et al., 2005).

Since the victimization scale measures personal experiences of

being a victim, and victim rates may be relatively low in many

schools, an additional scale was selected to measure student per-

ceptions of the extent of bullying and hostility at school that affects

all students (Bandyopadhyay, Cornell, & Konold, 2009; Cornell &

Sheras, 2003). For example, the scale asks whether students from

different neighborhoods get along with one another, and whether

they observe a lot of teasing about clothing or appearance at

school. Student observations of how much bullying and teasing

occurs at school are an important aspect of school climate. Re-

search has shown that a climate of bullying and hostility can affect

all students, even those who are bystanders but not direct victims

(Olweus & Limber, 2000). Unnever and Cornell (2003) described

a “culture of bullying” that develops in a school when students

come to perceive that bullying is a pervasive and generally ac-

cepted event. In addition to student perceptions, we assessed

teacher perceptions by administering a parallel version of the

Bullying Climate Scale (Cornell, 2006) asking teachers for their

perspective on the extent of bullying and hostility among students.

Method

Procedures and participants. In spring 2007, we collected

school climate surveys from both ninth-grade students and teach-

ers in 290 of the 314 public high schools in the state of Virginia.

We selected ninth grade because these students are completing the

first year of high school and have a high rate of discipline prob-

lems (45% of all discipline violations for Grades 9 –12 in Vir-

ginia). The school participation rate was over 92%, which was

achieved with the cooperation of the Virginia Department of

Education and the Virginia Department of Criminal Justice Ser-

vices, who endorsed the study and encouraged participation. Sur-

veys were completed online by samples of ninth-grade students

and teachers from each school. With a few exceptions (e.g., small

rural schools), each school selected around 25 ninth-grade students

from their enrollment list using a set of random numbers generated

for each school based on class size. From each school, about 10

ninth-grade teachers were selected using a similar set of random

numbers based on the estimated number of ninth-grade teachers in

each school.

Online surveys were obtained from 7,318 students and 2,922

teachers. Of these students, 49% were girls and 51% were boys.

Based on student self-report, the study sample was 63% Cauca-

sian, 22% African American, 5% Hispanic, 3% Asian American,

5% other, and less than 1% American Indian. The ninth-grade

population reported by the Virginia Department of Education

(2007) was 56% Caucasian; 30% African American; 8% Hispanic;

4% Asian American; and 2% American Indian, Hawaiian, and

unspecified. The state had no “other” category. Using the compa-

rable racial categories, we found that the study sample had slightly

more Caucasian students and fewer African American, Hispanic,

and Asian students than the state’s ninth-grade population, ␹

(4,

N⫽6,962) ⫽276.39, p⬍.001.

Of the 2,922 teachers who completed the survey, 64% were

female and 37% were male. Eighty-three percent of the teachers

were Caucasian, 12% were African American, 2% were Hispanic,

1% were Asian American, 1% were other, and less than 1% were

American Indian. State demographics for high school teachers

were not available. Teachers reported that they had 1–5 years

(36%), 6 –10 years (21%), 11–15 years (13%), or more than 15

years (30%) of teaching experience.

Measures

School-level risk factors. Previous research has shown that

the overall composition and size of the student body can increase

the likelihood of victimization (D. C. Gottfredson, 2001; U.S.

Department of Health and Human Services, 2001). Therefore, all

analyses took into account the enrollment size of the school

(Grades 9 –12), the proportion of students of color, and the pro-

portion of students participating in the free- and reduced-price

meal program.

Individual-level risk factors. Boys are more likely than girls

to report being the victim of violence in schools (NCES, 2007),

and low-income Hispanic and African American students are more

likely to be exposed to violence than more affluent, Caucasian

students (Ozer & Weinstein, 2004). Given these trends, we con-

trolled for student gender, which was coded as 1 for boys and 0 for

girls. To analyze the effects of student race, we dummy coded the

race and ethnicity variables, using the largest student group as the

reference group (Hardy, 1993). The largest group in our sample

comprised Caucasian adolescents. Dummy variables were con-

structed in the following way: Caucasian (0) versus African Amer-

ican (1), Caucasian (0) versus Hispanic (1), Caucasian (0) versus

Asian (1), and Caucasian (0) versus Other (1). We developed

similar comparisons for teacher race and gender, and also took into

account years of teaching experience.

School safety. As part of the online survey, students and

teachers completed a Bullying scale taken from the School Climate

Bullying Survey (Cornell & Sheras, 2003; McConville & Cornell,

2003). This scale asks students and teachers to rate (strongly

disagree,disagree,agree,strongly agree) the extent of teasing and

bullying at school (e.g., “Students here often get teased about their

clothing or physical appearance).” The Bullying scale includes the

4-item Prevalence of Teasing and Bullying scale reported by

Bandyopadhyay et al. (2009), but also includes three items inquir-

ing whether students are made to feel welcome and accepted by

other students, and whether students from different neighborhoods

get along. The 7-item scales had Cronbach’s alphas of .77 (student

scale) and .87 (teacher scale). The study by Bandyopadhyay et al.

using the shorter version of the Bullying scale found that the

486 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

aggregated amount of bullying reported by students in a school

was correlated with several indicators of schoolwide disorder,

including the number of short-term suspensions at school and

teacher reports of gang-related violence at school.

Students also completed a Victimization index based on G. D.

Gottfredson’s (1999) nine items. We excluded two relatively triv-

ial forms of victimization (i.e., theft and damage of property worth

less than $10) and relied on the remaining seven forms of student

victimization, which ranged from theft of personal property worth

more than $10 to being physically attacked (G. D. Gottfredson,

1999). Students answered “yes” or “no” for each form of victim-

ization they had experienced in the past school year. Notably, this

scale can be distinguished from the Bullying scale because it asks

students to report their own victimization experiences rather than

how frequently they observed the victimization of others. The

scale had a Cronbach’s alpha of .72. Totals were calculated for

each participant based on how many forms of victimization he or

she reported.

Structure and support. Ninth-grade student surveys were

used to measure structure and support in the high schools. School

structure was measured by two scales completed by students.

Experience of School Rules is a 7-item scale used in the School

Crime Supplement to the National Crime Victimization Survey

(NCES, 2005). Students responded (strongly disagree,disagree,

agree,strongly agree) to seven items designed to measure percep-

tions of school rules as fair and uniformly enforced, such as “The

school rules are fair” and “The school rules are strictly enforced.”

The scale had a Cronbach’s alpha of .74.

The second measure of school structure was the Daily Structure

scale (Cornell, 2006), which was devised for this study to measure

student perceptions of how strictly rules were enforced during

the school day. Students were asked how likely students would be

caught or punished (not at all likely,not likely,likely,very likely)

for six common problems such as cutting class, coming late to

class, smoking, fighting, and speaking sarcastically to a teacher.

The scale had a Cronbach’s alpha of .54.

School support was measured by two scales. The Learning

Environment scale was used by Austin and Duerr (2005) to mea-

sure how much students perceive that adults in their school are

supportive and respectful of students. The scale consisted of eight

items asking students how much they agree (strongly disagree,

disagree,agree,strongly agree) that the adults in their school

“really care about all students,” “treat all students fairly” and show

respect and support for students in other ways. The scale had a

Cronbach’s alpha of .96.

The Help Seeking scale, like the Bullying scale, was taken from

the School Climate Bullying Survey (Cornell & Sheras, 2003).

This instrument and the Help Seeking scale in particular have been

used in a series of studies of school bullying (e.g., Bandyopadhyay

et al., 2009; Cole, Cornell, & Sheras, 2006; Williams & Cornell,

2006). The Help Seeking scale was designed to measure student

willingness to seek help from school staff members for bullying

and threats of violence. Student were asked to agree (strongly

disagree,somewhat disagree,somewhat agree,strongly agree)

with eight items, including statements such as “If another student

was bullying me, I would tell one of the teachers or staff at school”

and “If another student brought a gun to school, I would tell one

of the teachers or staff at school.” The scale had a Cronbach’s

alpha of .89 in the present study. A previous study found that the

scale structure was supported by a series of factor analyses across

middle school and high school samples, as well as a multigroup

confirmatory analysis showing full metric invariance across gen-

der and race groups (Bandyopadhyay et al., 2009). In addition,

schools whose students reported higher help seeking had fewer

short- and long-term suspensions, according to school records, as

well as lower levels of bullying and gang-related violence, as

reported by teachers (Bandyopadhyay et al., 2009).

Missing data. A small percentage of teachers (2%, n⫽55)

did not complete the Bullying scale. Chi-square tests showed that

total teacher participants did not significantly differ in terms of

gender, race/ethnicity, or years of teaching experience from the

subsample that excluded teachers with missing data. Due to the

missing teacher data, 10 schools were not included in analyses that

used teacher reports of bullying.

A small percentage of students (3%, n⫽198) did not complete

the Bullying and Victimization scales. Again, chi-square tests

confirmed that the total student respondents did not significantly

differ in terms of gender and race/ethnicity compared to the

subsample excluding students with missing data. Due to the miss-

ing student data, two of the schools in the original sample were

excluded from analyses using student-reported bullying and vic-

timization.

Data analytic plan. We conducted a confirmatory factor anal-

ysis on the scales theorized to comprise structure and support.

Multiple students and teachers were nested within each school,

which resulted in the nonindependence of their data. Given this

nesting and the focus on school-level differences on the outcomes,

it was important to use a data analytic technique that disaggregated

within- and between-school variance. Hierarchical linear modeling

(HLM) allows for comparison of school differences, after consid-

ering within-school student or teacher variability in the outcomes

(Raudenbush & Bryk, 2002). Shinn and Rapkin (2000) suggested

that scholars need to carefully consider at what levels in an

ecology the constructs of interest should be examined, then select

the most appropriate analytic strategy. Given that a majority of the

high school students in the current study changed classrooms

between four and 10 times a day, students’ experience of bullying

and victimization would not likely be confined to single class-

rooms. Their experience would be more related to the school as a

whole, suggesting that between-schools variability, not between-

classrooms variability, would be most relevant. Therefore, for the

analyses, a two-level model was most appropriate conceptually

(e.g., students within schools), but not a three-level model (e.g.,

students within classrooms within schools).

For each school safety outcome, the predictors—school struc-

ture and support—were examined after taking into account school

size and the proportion of low-income and ethnic minority stu-

dents. In the analyses, all dependent and independent continuous

variables were standardized (M⫽0,SD⫽1), which has the

beneficial effect of centering the variables and facilitating the

interpretability of the HLM estimates. The dependent variable,

student victimization, was found to be positively skewed. With

HLM using a log transformation of the variable and multilevel

logistic regression using a dichotomized variable, we found similar

results to the HLM using the variable as continuous. To increase

the interpretability of the results, we present HLM analyses with

the continuous victimization variable.

487

AUTHORITATIVE SCHOOL DISCIPLINE

Results

Preliminary Analyses

Descriptives. Descriptive statistics for all study variables are

presented in Table 1. Students expressed a wide range of perspec-

tives as to whether bullying was a problem at their schools (M⫽

16.71, SD ⫽3.34). Teachers held a similar range in perspective

about the problem (M⫽16.77, SD ⫽3.11). Students typically

reported one or two forms of victimization (M⫽1.34, SD ⫽1.49).

However, some students reported as many as seven forms of

victimization. Overall, most students and teachers reported favor-

able perceptions of structure and support in their school.

The enrollment of Virginia’s public high schools ranged from

33 to 2,881 students with a mean of 1,207. The percentage of

students qualifying for free- and reduced-price meals varied from

1% to 83% across schools (M⫽30%, SD ⫽16%). Similarly, the

percentage of ethnic minority students varied widely—from 0% to

99%—across schools (M⫽34%, SD ⫽26%). The correlation

between proportion of low-income students and proportion of

ethnic minority students was .31. School enrollment correlated

⫺.43 with the proportion of low-income students and .20 with the

proportion of ethnic minority students (see Table 2).

Confirmatory factor analysis. We conceptualized that

School Structure should be represented by two scales (Experience

of School Rules and Daily Structure), while School Support should

be represented by another two scales (Learning Environment and

Help Seeking). To empirically evaluate whether the data could

support our hypothesized structure of these two constructs, we

conducted a confirmatory factor analysis (CFA) on the full student

sample, using LISREL 8.8 (Jo¨ reskog & So¨ rbom, 2007). The CFA

results supported the formation of these two constructs as we

hypothesized, with the loadings of two scales (Experience of

School Rules and Daily Structure) on one factor (School Structure)

being 0.74 and 0.53, respectively, and the loadings of the other two

scales (Learning Environment and Help Seeking) on the second

factor (School Support) being 0.82 and 0.77, respectively. The

CFA model fit indices suggested good fit (e.g., root-mean-square

error of approximation ⫽.06, adjusted goodness-of-fit index ⫽

.98, comparative fit index ⫽1.00, nonnormed fit index ⫽0.98) for

this model. The two scales under each construct were standardized

and summed to form composite scores of each factor (i.e., School

Structure, School Support), and the composite scores were used in

later substantive analyses. The scale items composing the School

Structure and School Support had good internal consistency (Cron-

bach’s ␣s⫽.72, .91, respectively).

Correlations. Pearson product–moment correlations were run

with variables aggregated at the school level (see Table 2). Teacher

perceptions of bullying were significantly associated with student

reports of bullying (r⫽.38, p⬍.001) and student reports of

victimization (r⫽.17, p⬍.01). Teachers perceived more bullying

in larger schools and schools with more students of color (r⫽.16,

p⬍.01; r⫽.28, p⬍.001, respectively), compared to teachers in

smaller schools and schools with fewer students of color. Student

perceptions of bullying, however, were unrelated to school size,

and the proportion of ethnic minority students.

Schools with high structure were more likely to have high

support (r⫽.53, p⬍.001) compared to schools with low

structure. Related to the central questions in this study, structure

and support were, as expected, inversely related to student victim-

ization and student-reported bullying (rranges from ⫺.34 to ⫺.47,

p⬍.001). Student reports of structure and support were also

associated with teacher perceptions of bullying among students

(r⫽⫺.29 and r⫽⫺.31, p⬍.001, respectively).

HLM Analyses

HLM models were examined for each of the three Level 1

outcome variables: (a) student-reported bullying, (b) teacher-

reported bullying among students, and (c) student-reported victim-

ization. Specifically, three models were examined for each out-

come. The first model was the two-level HLM model with no

predictors (null model). Results from this model were used to

calculate the intraclass correlation coefficient (ICC), which is the

proportion of between-schools variance (i.e., ␴

) to the total

variance (i.e., ␴

⫹␴

). The second model included only Level 1

and Level 2 control variables, which were treated as fixed effects.

For student-reported bullying and victimization, Level 1 control

variables included student gender and race. For teacher-reported

bullying, Level 1 control variables were teacher gender, race, and

years of teaching experience. Level 2 control variables included

school size, a school’s proportion of minority students, and pro-

portion of students qualifying for free- and reduced-price meals.

The third model for each outcome variable included the control

variables mentioned above, plus the variables of our research

focus: perceived school structure and school support. Comparisons

among the three models identified the increase in proportion of

variance explained in the outcome when structure and support

were added to the model. The proportion of explained variance

provides an index of effect size (Kreft & de Leeuw, 1998). For

student outcome variables, Model 3 was as follows:

Level 1 model: Yij ⫽␤0j⫹␤1j共Genderij兲

⫹␤2ij共African Americanij兲⫹␤3j共Hispanic兲⫹␤4j共Asianij兲

⫹␤5j共Othersij兲⫹eij

Level 2 model: ␤0j⫽␥00 ⫹␥01 共% minorityj兲

⫹␥02 共% free and reduced-price mealj兲

⫹␥03 共School Sizej兲⫹␥04 共Structurej兲

⫹␥05 共Supportj兲⫹u0j

Table 1

School Safety, Structure, Support, and Student Demographics

Variable MSDMin. Max.

Bullying

Student report 16.71 3.34 7.00 28.00

Teacher report 16.77 3.11 7.00 28.00

Student reports of victimization 1.34 1.49 0.00 7.00

Structure 50.00 8.72 22.59 83.73

Support 50.00 9.37 25.99 74.24

School composition

% minority 34 26 0 99

% free- and reduced-price meal 30 16 1 83

School size 1,207 687 33 2,881

488 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

where irefers to individual level, jrefers to school level, erefers

to error or residual at the individual level, and urefers to residual

at the school level.

Bullying. Table 3 presents the summary of HLM analysis for

the outcome variable of student-reported bullying in school. The

between-schools variation in bullying accounted for 5.6% of the

total variation (ICC ⫽0.056). Model 2 shows that the Level 2

control variables (school size, proportion of minority students, and

students who qualified for free- and reduced-price meals) ac-

counted for about 9% of the between-school variance. Schools

with a lower proportion of minority students (␥

⫽⫺.05, p⬍.05)

and a higher proportion of students who qualified for free- and

reduced-price meals (␥

⫽.10, p⬍.001) had more bullying.

Model 2 also shows that Level 1 covariates accounted for none of

the within-school variation. Female students were more likely to

report bullying than were male students (␤

⫽⫺.08, p⬍.001). In

addition, students who selected “other” in the race categories were

more likely than Caucasian students to report bullying (␤

⫽.24,

p⬍.001).

With structure and support in Model 3, an additional 45% of

between-schools variance in bullying was explained after we took

into account school size, proportion of minority students, and

students who qualified for free- and reduced-price meals. When we

took into account the school-level control variables, higher levels

of school structure (␥

⫽⫺.07, p⬍.001) and school support

(␥

⫽⫺.12, p⬍.001) were statistically associated with less

bullying.

Table 4 presents the HLM analysis for teacher-reported bully-

ing. The between-schools variation accounted for 13% of the total

variation (ICC ⫽0.13). Model 5 shows that the Level 2 control

variables (school size, proportion of minority students, and stu-

dents who qualified for free- and reduced-price meals) accounted

for about 15% of the between-schools variance. A higher propor-

tion of minority students was significantly associated with teacher

perceptions of more bullying (␥

⫽.13, p⬍.001). All three

Level 1 teacher characteristics were significant predictors of bul-

lying. Female teachers reported more bullying among students

than did male teachers (␤

⫽⫺.21, p⬍.001). Caucasian teachers

Table 2

Intercorrelations Among Student and Teacher Variables

Variable 1 2 3 4 5 6 7

1. % minority —

2. % free- and reduced-price meal .31

ⴱⴱⴱ

—

3. School size .20

ⴱⴱⴱ

⫺.43

ⴱⴱⴱ

—

4. Structure ⫺.18

ⴱⴱ

.08 ⫺.19

ⴱⴱ

—

5. Support ⫺.34

ⴱⴱⴱ

⫺.10 ⫺.20

ⴱⴱ

.53

ⴱⴱⴱ

—

6. Bullying (SR) .07 .15

ⴱ

.11 ⫺.39

ⴱⴱⴱ

⫺.47

ⴱⴱⴱ

—

7. Bullying (TR) .28

ⴱⴱⴱ

.10 .16

ⴱⴱ

⫺.29

ⴱⴱⴱ

⫺.31

ⴱⴱⴱ

.38

ⴱⴱⴱ

—

8. Victimization (SR) .06 .03 .04 ⫺.34

ⴱⴱⴱ

⫺.36

ⴱⴱⴱ

.42

ⴱⴱⴱ

.17

ⴱⴱ

Note. SR ⫽student reported; TR ⫽teacher reported.

ⴱ

p⬍.05.

ⴱⴱ

p⬍.01.

ⴱⴱⴱ

p⬍.001.

Table 3

HLM Analysis With Student-Reported Bullying as Level 1 Outcome

Measure

Model 1 Model 2 Model 3

Estimate SE Estimate SE Estimate SE

Student-level predictors

Gender (1: male; 0: female) ␤

⫺.08

ⴱⴱⴱ

.02 ⫺.08

ⴱⴱⴱ

.02

African American ␤

⫺.01 .03 ⫺.02 .03

Hispanic ␤

⫺.03 .05 ⫺.02 .05

Asian ␤

.03 .07 .03 .07

Others ␤

.24

ⴱⴱⴱ

.05 ⫺.23

ⴱⴱⴱ

.05

School-level predictors

% minority ␥

⫺.05

ⴱ

.02 ⫺.09

ⴱⴱⴱ

.02

Free- and reduced-price meal ␥

.10

ⴱⴱⴱ

.02 .10

ⴱⴱⴱ

.02

School size ␥

.09

ⴱⴱⴱ

.02 .06

ⴱⴱⴱ

.02

Structure ␥

⫺.07

ⴱⴱⴱ

.02

Support ␥

⫺.12

ⴱⴱⴱ

.02

Random effects

Individual level ␴

.94

ⴱⴱⴱ

.02 .94

ⴱⴱⴱ

.02 .94

ⴱⴱⴱ

.02

School level ␴

.06

ⴱⴱⴱ

.01 .05

ⴱⴱⴱ

.01 .03

ⴱⴱⴱ

.01

Reduced variance within schools

.00 .00

Reduced variance between schools

.09 .54

Proportion of unexplained variance reduced from Model 1 (null model).

ⴱ

p⬍.05.

ⴱⴱⴱ

p⬍.001.

489

AUTHORITATIVE SCHOOL DISCIPLINE

reported more bullying among students than African American and

Asian teachers (␤

⫽⫺.23, p⬍.001; ␤

⫽.55, p⬍.001,

respectively). Less experienced teachers were more likely to report

bullying (␤

⫽⫺.14, p⬍.001).

Structure and support were included in Model 6. Compared to

Model 5, an additional 8% of between-schools variance in bullying

was explained. When we took into account school size, proportion

minority, and proportion receiving free- and reduced-price meals,

higher levels of school structure (␥

⫽⫺.08, p⬍.05) and school

support (␥

⫽⫺.07, p⬍.05) were statistically associated with

less bullying.

Victimization. Table 5 presents the summary of HLM analy-

ses for the outcome variable of student-reported victimization. The

between-schools variance was about 1.5% of the total variance

(ICC ⫽.015). Model 8 shows that the Level 2 school composition

variables did not account for any of the between-schools variance

and were not statistically associated with victimization. Of the

Level 1 covariates, gender was associated with victimization, with

Table 4

HLM Analysis With Teacher-Reported Bullying as Level 1 Outcome

Measure

Model 4 Model 5 Model 6

Estimate SE Estimate SE Estimate SE

Teacher-level predictors

Gender (1: male; 0: female) ␤

⫺.21

ⴱⴱⴱ

.04 ⫺.21

ⴱⴱⴱ

.04

African American ␤

⫺.23

ⴱⴱⴱ

.06 ⫺.24

ⴱⴱⴱ

.06

Hispanic ␤

⫺.05 .14 ⫺.06 .14

Asian ␤

⫺.55

ⴱⴱⴱ

.15 ⫺.55

ⴱⴱⴱ

.15

Others ␤

⫺.00 .12 ⫺.00 .12

Years of teaching ␤

⫺.14

ⴱⴱⴱ

.02 ⫺.14

ⴱⴱⴱ

.02

School-level predictors

% minority ␥

.13

ⴱⴱⴱ

.04 .09

ⴱ

.04

Free- and reduced-price meal ␥

.05 .04 .05 .03

School size ␥

.04 .04 .03 .04

Structure ␥

⫺.08

ⴱ

.03

Support ␥

⫺.07

ⴱ

.03

Random effects

Individual level ␴

.87

ⴱⴱⴱ

.02 .84

ⴱⴱⴱ

.02 .84

ⴱⴱⴱ

.02

School level ␴

.13

ⴱⴱⴱ

.02 .11

ⴱⴱⴱ

.02 .10

ⴱⴱⴱ

.02

Reduced variance within schools

.03 .03

Reduced variance between schools

.15 .23

Proportion of unexplained variance reduced from Model 4 (null model).

ⴱ

p⬍.05.

ⴱⴱⴱ

p⬍.001.

Table 5

HLM Analysis With Student-Reported Victimization as Level 1 Outcome

Measure

Model 7 Model 8 Model 9

Estimate SE Estimate SE Estimate SE

Student-level predictors

Gender (1: male; 0: female) ␤

.26

ⴱⴱⴱ

.02 .25

ⴱⴱⴱ

.02

African American ␤

.00 .03 ⫺.00 .03

Hispanic ␤

⫺.05 .05 ⫺.04 .05

Asian ␤

⫺.08 .07 ⫺.07 .07

Others ␤

.35

ⴱⴱⴱ

.05 .35

ⴱⴱⴱ

.05

School-level predictors

% minority ␥

⫺.00 .02 ⫺.03 .02

Free- and reduced-price meal ␥

.02 .02 .01 .02

School size ␥

.01 .01 ⫺.00 .02

Structure ␥

⫺.04

ⴱⴱ

.02

Support ␥

⫺.07

ⴱⴱⴱ

.02

Random effects

Individual level ␴

.99

ⴱⴱⴱ

.02 .96

ⴱⴱⴱ

.02 .96

ⴱⴱⴱ

.02

School level ␴

.02

ⴱⴱⴱ

.00 .02

ⴱⴱⴱ

.00 .01

ⴱ

.00

Reduced variance within schools

.03 .03

Reduced variance between schools

.00 .50

Proportion of unexplained variance reduced from Model 7 (null model).

ⴱ

p⬍.05.

ⴱⴱ

p⬍.01.

ⴱⴱⴱ

p⬍.001.

490 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

male students more likely to report victimization than female

students (␤

⫽.26, p⬍.001). In addition, students selecting

“other” in the racial categories were more likely than Caucasian

students to report higher victimization (␤

⫽.35, p⬍.001).

Model 9 included structure and support and accounted for 50% of

the between-schools variance in victimization. Taking into account

school size, proportion minority, and proportion receiving free-

and reduced-price meals, we found that higher levels of school

structure (␥

⫽⫺.04, p⬍.01) and school support (␥

⫽⫺.07,

p⬍.001) were statistically associated with less student victim-

ization.

To understand the differences related to the outcomes (bullying

and victimization) in authoritative versus nonauthoritative schools,

we classified schools by median and split them into four groups:

(a) low structure/low support, (b) low structure/high support, (c)

high structure/low support, and (d) high structure/high support. An

analysis of variance found significant group effects for student

victimization and for both student and teacher perceptions of

bullying. Follow-up group comparisons showed that schools low

on structure and support significantly differed on bullying and

victimization from schools high on both structure and support, and

the magnitude of the difference was considerable. Schools with

high structure and support were three-quarter to one standard

deviation lower on bullying and victimization compared to schools

with low structure and support.

Discussion

This study found support for authoritative discipline theory, a

new framework for conceptualizing developmentally appropriate

school discipline for adolescents. Within a large and diverse sam-

ple of public high schools, both structure and support were asso-

ciated with between-schools differences in safety. More specifi-

cally, student perceptions of structure and support, aggregated at

the school-level, were associated with less student victimization

and less bullying among students, even after we controlled for

school size and the proportions of low-income and ethnic minority

students in the student body. These findings support a new appre-

ciation for an authoritative approach to discipline at the school

level, as such approaches have been almost exclusively studied at

the classroom level.

A unique contribution of this study was the ability to examine

school variation in a nearly complete state population of public

high schools. The 92% school participation rate for this study is

noteworthy because schools with high levels of discipline prob-

lems or low investment in student support may be less likely to

participate in research. Two excellent studies (G. D. Gottfredson et

al., 2005; Hanson & Austin, 2003) acknowledged that low partic-

ipation limited their findings; for example, the G. D. Gottfredson

et al. study (2005) was limited to just 254 (30%) of the 847 invited

schools. The current study is much less vulnerable to selection

biases found in previous studies and because we were able to

examine student-perceived structure and support in almost an

entire state population of high schools. This increases confidence

in the applicability of the findings for diverse schools.

Structure

Structure comprised student perceptions of the rules as fair and

consistently enforced for common problems such as cutting class,

smoking, fighting, and speaking sarcastically to a teacher. Schools

high on structure did not overlook such common problems. The

HLM analyses showed that schools with more structure had less

bullying and student victimization. This study extends early re-

search on structure, mostly concerned with classroom settings,

which found that students had positive perceptions of their teach-

ers if they offered clear and consistently enforced rules (e.g.,

Trickett & Moos, 1974). The present study suggests that there are

meaningful differences between schools in how school rules are

perceived by students. Strategies for student behavior management

often focus on classroom-level interventions, but evidence that

disciplinary structure is a construct that can be expanded from the

level of classroom practice to a schoolwide level suggests that a

broader approach should be considered. This is especially relevant

in high school settings, where students regularly change classes,

and bullying and victimization are not necessarily confined to

specific classrooms. Several studies have found that schools char-

acterized by greater structure provide a safer learning environment.

Specifically, studies have found that adolescents’ greater under-

standing of rules was associated with lower school crime (Mayer

& Leone, 1999), and their perceptions of the rules as fair have been

linked to better student behavior (D. C. Gottfredson et al., 1993).

These findings are consistent with theoretical expectations that

adolescents have a developmental need for both predictability of

their environment and reasonable application of rules. Adolescents

are more willing to accept and trust school authority if they

perceive rule enforcement to be fair (Tyler, 2006).

Support

Characteristics of school support were found to be distinct from

the characteristics of school structure, as demonstrated in the factor

analytic results. Previous research shows that adolescents who

experience adults in the school as supportive are more likely to

have academic success (e.g., Goodenow, 1993; Gregory & Wein-

stein, 2004). The current study extends these findings to the

domain of school safety and corroborates previous research show-

ing that students with positive teacher and student relationships

have lower problem behavior (Jessor et al., 2003) and are less

likely to use weapons (Henrich et al., 2005). How and why school

support is related to higher school safety deserves further explo-

ration. Future research may consider whether students in schools

with adult support turn to adults for help before conflict escalates.

Extrapolating from Darling and Steinberg’s (1993) proposed

mechanisms of parenting styles, students in highly supportive

schools may simply be more open and responsive to directives

from school staff whom they experience as fair and respectful.

Additive Effects of Structure and Support

Discussions about school discipline policy often contrast “get

tough” practices with “give support” practices as though the two

were mutually exclusive. For example, Nickerson and Martens

(2008) found that principal attitudes toward discipline could be

categorized into a security/enforcement approach (i.e., use of se-

curity cameras, police officers) and an education/therapeutic ap-

proach (i.e., use of violence prevention programs, teacher training

in classroom management). In contrast, this study found that

structure and support were positively correlated (r⫽.53), at least

491

AUTHORITATIVE SCHOOL DISCIPLINE

from a student perspective. Schools where students perceived that

rules were strictly enforced also described more supportive rela-

tionships with the adults in their school.

It should be noted, however, that our measure of structure was

based on student perceptions of whether rules were fair and strictly

enforced. We did not assess the use of punitive disciplinary con-

sequences such as suspension and expulsion, which we regard as

conceptually distinguishable from our concept of structure. Zero-

tolerance discipline policies impose severe sanctions (often long-

term suspension or expulsion) for even minor violations of a

school rule, with little or no consideration of the circumstances of

the behavior or the student’s intentions (Heaviside et al., 1998;

Skiba & Peterson, 1999; Tebo, 2000). Zero-tolerance policies may

reflect the rigid and controlling practices that Baumrind (1968)

described as authoritarian (Arum, 2003; Nickerson & Spears,

2007). Although not studying zero-tolerance policies specifically,

Pellerin (2005) found that authoritarian schools, which emphasized

punishment, had higher rates of drop out compared to schools that

responded to student needs and demanded academic performance

and cooperation with the rules. Future research should determine

whether schools with authoritarian discipline rely on zero-

tolerance policies, which are inherently unsupportive, and elicit

negative student behavior compared to schools with authoritative

discipline.

Although structure and support were highly correlated in our

study, both made independent contributions to the prediction

of school safety conditions. Notably, structure and support were

significantly associated with all three safety indicators. When

simultaneously entered into HLM analyses, structure and support

predicted less student-reported victimization and less student- and

teacher- perceived bullying among students, compared to schools

with less structure and support. One strength of these findings is

their consistency across informants. Although student and teacher

perceptions of bullying were only modestly correlated (.38), both

were associated with student perceptions of support. Of particular

relevance is that student perceptions of structure and support

predicted teacher perceptions of bullying among students. In con-

trast, previous studies have often relied on measures obtained from

the same informants, which can inflate correlations due to response

bias and other sources of shared method variance (Nickerson &

Martens, 2008; Roberts, Wilcox, May, & Clayton, 2007).

Overall, structure and support explained between 8% and 50%

of the between-schools variance in safety outcomes. Said with

caution given the correlational nature of the study, these effects are

large enough to suggest that schools could achieve meaningful

differences in levels of bullying and student victimization by

improving their structure and support (Rosenthal, 1990). More-

over, the findings took into account the composition of the student

body related to race, income, and school size. This is important

given the need to identify how to increase school safety in larger

high schools (G. D. Gottfredson et al., 2005) and how to lower the

safety fears of African American and Hispanic students, which

tend to be higher than the fears of Caucasian students (NCES,

2007). Comparisons of HLM coefficients offer another perspective

on the potential of structure and support for addressing school

safety. In the HLM model predicting student-perceived bullying,

the statistically significant estimates for the school-level control

variables (e.g., school poverty) ranged from .05 to .10. The esti-

mates for structure (⫺.07) and support (⫺.12) were of the same

magnitude. This suggests for schools facing risk factors, which are

largely out of their control, structure and support hold promise as

potential buffers to these risks. In addition, the present study’s

findings have important implications for future intervention re-

search, which may find that disparity in experiences of victimiza-

tion could be substantially lowered in schools with increases in

structure and support.

Perhaps the most important implication of this study is that

structure and support should be considered in tandem. We conjec-

ture that students who feel supported and respected at school are

more accepting of structure (Arum, 2003); in the absence of

support, no degree of structure may be adequate (D. C. Gottfredson

et al., 1993; Hollingsworth et al., 1984). Offering support does not

preclude upholding behavioral expectations through fair and con-

sistent rule enforcement. Administrators should try to ensure that

every student feels connected to at least one teacher and feels

comfortable enough to seek help when needed. At the same time,

an administrator might consider strategies to clearly communicate

rules and demonstrate fair and consistent enforcement of those

rules.

Characteristics of Students and Teachers

The large within-school variation in school bullying and vic-

timization suggests that students within the same building have

divergent experiences of school safety. The student demographic

characteristics examined in this study explained only a small

portion of the within-school variance in school safety. Boys were

more likely to report being victimized, yet girls were more likely

to report bullying among students. Girls may be the victims of

fewer physical assaults and thefts than boys and, at the same time,

be more sensitive to the climate of bullying than boys. It is also

quite plausible that girls were more likely to report bullying than

were boys because the measure of bullying included verbal and

social forms of bullying that are more common among girls than

boys (Crick & Grotpeter, 1995).

In the analyses of student-reported safety, the “other” versus

Caucasian contrast was the only race comparison variable that was

significant. Compared to Caucasian students, those who chose

“other” and presumably did not consider themselves a member of

the Asian, African American, or Hispanic groups were more likely

to report bullying and victimization. About 60% of the students

who chose “other” wrote that they were mixed or biracial. Future

research needs to address the experience of biracial adolescents,

many of whom may feel forced to identify with one racial group

over another in order to fit in with the racial social segregation

typical of high schools (Tatum, 1997; Wardle, 1992). If they cross

boundaries between racial groups, they may be at risk for teasing

and bullying.

Future research needs to consider additional student character-

istics, such as social status, popularity, and social class (Allen,

Porter, McFarland, Marsh, & McElhaney, 2005) that may help

explain divergent experiences of safety. For instance, a recent

national report showed that low-income students report greater

victimization than high-income students (Bauer et al., 2008). Close

to 8% of students ages 12–18 with household incomes of $15,000

or less reported being a victim in school compared with about 5%

of their peers with household incomes of $50,000 or more. With-

out the socioeconomic status of individual students, we were not,

492 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

in the current study, able to determine whether low-income stu-

dents experienced higher rates of victimization, regardless of racial

group membership. Future researchers might try to tease apart the

effects of student race and socioeconomic status, which are over-

lapping in many communities. Additional knowledge of student

characteristics and school safety could inform the substantial lit-

erature on peer aggression and victimization, as well as prevention

and intervention programs to reduce bullying and other forms of

peer aggression and bullying (Jimerson & Furlong, 2006; Ladd &

Ladd, 2001).

Less experienced teachers and female teachers were more likely

to perceive bullying among students compared to more experi-

enced teachers and male teachers, respectively. In addition, Cau-

casian teachers, compared to Asian and African American teach-

ers, were more likely to perceive bullying. Multiple explanations

for these findings are possible. Less experienced teachers, Cauca-

sian teachers, and female teachers may be more sensitive to bul-

lying and teasing as a problem or may be more likely to overiden-

tify a student interaction as a manifestation of bullying than other

teachers. Another explanation is that students are less likely to

display such behavior in the presence of a more experienced

teacher, a male teacher, an Asian teacher, or an African American

teacher.

Given that teacher characteristics explained only 3% of the

within-school variation in teacher-perceived bullying, additional

teacher variables need to be examined in future research. For

instance, teachers’ perceptions of bullying may relate to the stu-

dent composition in their classrooms. Compared to teachers with

greater academic homogeneity in their classrooms, teachers in

mixed-ability classrooms have a more difficult time eliciting co-

operation and engagement from their students (Evertson, Sanford,

& Emmer, 1981). Future research might determine whether teach-

ers’ perceptions of bullying among students differ depending on

the academic track level or heterogeneity of ability in their class-

rooms. In addition, the likelihood of bullying may be higher in

some course subjects than others.

Limitations

This study examined relationships between measures of school

climate and safety; correlational studies do not establish causal

relationships and are open to multiple interpretations. It is certainly

plausible that there are bidirectional causal effects and that school

safety conditions could and likely do affect school structure and

support. Future studies may consider alternative explanations for

these findings by testing other models or may demonstrate causal

effects through experimental interventions. Nevertheless, the find-

ings in this study are consistent with our presumed causal model

by demonstrating a statistical effect of school structure and support

on school safety after we controlled for known demographic risk

factors for school disorder and used a model that considers the

nesting of student and teacher measures within schools. The anal-

yses show that, although structure and support are moderately

correlated (.53), they make independent contributions to each of

three measures of school safety.

Another limitation is that school structure and support were

based on perceptions of ninth-grade students. There are two po-

tential limitations here. First, students might not be able to accu-

rately assess the degree of structure and support in their school

and, second, the perceptions of ninth graders might not be repre-

sentative of the experiences of all students. However, the percep-

tions of students have intrinsic importance because their experi-

ences of school climate may be more influential on their school

adjustment and behavior than more objective indicators (Loukas,

Suzuki, & Horton, 2006). In a sense, student perceptions are an

objective indicator in their own right. Students react to school rules

as they perceive them, and if they regard the rules as easily flouted,

they are more likely to disobey them. Furthermore, their subjective

perceptions of teachers as warm and supportive are essential to

creating a positive school climate. Any effort by school authorities

to create a supportive environment cannot be regarded as success-

ful if the students do not perceive the adults in the school as

supportive.

There is the additional concern that ninth grader perceptions

might differ from those of other students. Only a study that

includes students at other grades can resolve this question. How-

ever, ninth grade is the first year of high school, and adjustment

problems at this level may lead to more serious difficulties and

failure to complete high school. The creation of a positive school

climate that is discernible to ninth graders would seem to be an

essential step in facilitating students’ high school adjustment and a

kind of acid test in determining whether school practices have their

intended impact.

Nevertheless, it would be useful for future studies to examine

additional perspectives on school structure and support. For ex-

ample, it may be useful to examine how school administrators

differ in their approaches to discipline and how these differences

are communicated to students (Fenning & Bohanon, 2006). Some

principals focus on how underlying student needs contribute to

their unsafe behavior. Other principals are less likely to emphasize

support and more likely to adopt exclusively punitive approaches

to unsafe behavior (Morrison & Skiba, 2001). Additional research

would need to identify whether differences in how school leaders

approach school safety are reflected in student experience of

structure and support.

The measures of school safety were also limited to student

victimization and the extent of bullying and teasing at school. Both

student and teacher perceptions of bullying/teasing were exam-

ined, but additional measures of school safety deserve consider-

ation, including teacher victimization. Administrative records of

school disciplinary infractions could be compared to student and

teacher reports of victimization. This comparison will help identify

the degree to which serious disciplinary action indicates school

safety (Heaviside et al., 1998), given that some office referrals,

suspensions, and expulsions are issued for minor misbehavior

(Morrison et al., 2006) or are applied for more subjective reasons

to one racial group compared to another (Skiba, Michael, Nardo, &

Peterson, 2002).

The between-schools variance of student victimization was only

1.5% of the total variance. It is unclear whether the low ICC for

this variable is due to the low base rate of serious victimization in

schools or some other difficulty in the measurement of victim-

ization. We examined several variations in this scale (such as

examining only the most serious forms of victimization and log

transforming the variable to obtain a more normal distribution) but

found no more illuminating results. However, the low ICC for

victimization was similar to the findings by G. D. Gottfredson and

colleagues (2005), who found the between-schools variance for

493

AUTHORITATIVE SCHOOL DISCIPLINE

victimization was only 4% of the total variance. Similarly,

LeBlanc and colleagues (2008) found that between-schools vari-

ance for student-reported violent antisocial behavior was only

3.6% of the total variance, and Koth, Bradshaw, and Leaf (2008)

found that a majority of variance in student-perceived order and

discipline was between students with less variance between class-

rooms and schools.

It appears that there are large variations in antisocial behavior

and victimization within schools, even in those schools with the

highest and lowest rates of overall safety problems. Student gender

and race explained only 3% of the within-school variation in

student-reported victimization. The study would have benefitted

from additional student characteristics in the HLM models, such as

student socioeconomic status and academic performance. Previous

research has found links between these student characteristics and

school safety. For instance, a greater percentage of low-income,

compared to high-income, students report being victimized by

crime and violence (Bauer et al., 2008). Students with low aca-

demic performance are more likely to victimize others compared

to their higher achieving peers (e.g., Choi, 2007). Future studies

including additional student characteristics may shed more light on

the variability of student victimization within schools.

Summary

In a statewide sample of ninth-grade students and teachers,

student perceptions that school rules were fair and strictly en-

forced, and that adults were supportive and willing to help stu-

dents, were associated with less student victimization and bullying.

As posited by authoritative discipline theory, both structure and

support were more common in safer schools. Just as many ado-

lescents benefit from authoritative parenting in their home, stu-

dents may benefit from a similarly authoritative environment in

their school.

References

Allen, J. P., Porter, M. R., McFarland, F. C., Marsh, P., & McElhaney,

K. B. (2005). The two faces of adolescents’ success with peers: Ado-

lescent popularity, social adaptation, and deviant behavior. Child Devel-

opment, 76, 747–760.

American Psychological Association Zero Tolerance Task Force. (2008).

Are zero tolerance policies effective in the schools? An evidentiary

review and recommendations. American Psychologist, 63, 852– 862.

Anderson, C. (1982). The search for school climate: A review of the

research. Review of Educational Research, 52, 368 – 420.

Arum, R. (2003). Judging school discipline: The crisis of moral authority.

Cambridge, MA: Harvard University Press.

Austin, G., & Duerr, M. (2005). Guidebook for the California Healthy Kids

Survey. Part III: School Climate Survey for teachers and other staff 2005–

2006 edition. Retrieved January 5, 2006, from http://www.wested.org/pub/

docs/chks_surveys_summary.html

Bandyopadhyay, S., Cornell, D., & Konold, T. (2009). Internal and exter-

nal validity of three school climate scales from the School Climate

Bullying Survey. School Psychology Review, 38, 338 –355.

Bauer, L., Guerino, P., Nolle, K. L., & Tang, S. (2008). Student victim-

ization in U.S. schools: Results from the 2005 school crime supplement

to the National Crime Victimization Survey (NCES Rep. No. 2009-306).

Washington, DC: U.S. Department of Education, National Center for

Education Statistics, Institute of Education Sciences.

Baumrind, D. (1968). Authoritarian vs. authoritative parental control. Ad-

olescence, 3, 255–272.

Baumrind, D. (1991). Parenting styles and adolescent development. In R.

Lerner, A. C. Petersen, & J. Brooks-Gunn (Eds.), The encyclopedia of

adolescent (pp. 746 –758). New York, NY: Garland.

Baumrind, D. (1995, April). Socialization effects in cultural contexts. Paper

presented at the Chicago Consortium on Child and Family at Loyola

University, Chicago, IL.

Baumrind, D. (1996). The discipline controversy revisited. Family Rela-

tions, 45, 405– 414.

Bohanon, H., Fenning, P., Carney, K. L., Minnis-Kim, M., Anderson-

Harris, S., Moroz, K. B., . . . Pigott, T. D. (2006). Schoolwide applica-

tion of positive behavior support in an urban high school: A case study.

Journal of Positive Behavior Interventions, 8, 131–145.

Choi, Y. (2007). Academic achievement and problem behaviors among

Asian Pacific Islander American adolescents. Journal of Youth and

Adolescence, 36, 403– 415.

Cole, J., Cornell, D., & Sheras, P. (2006). Identification of school bullies

by survey methods. Professional School Counseling, 9, 305–313.

Cornell, D. G. (2006). Bullying Climate Scale. Unpublished scale, Curry

School of Education, University of Virginia.

Cornell, D. G., & Sheras, P. L. (2003). The School Climate Bullying

Survey. Unpublished survey and report, Curry School of Education,

University of Virginia.

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

social–psychological adjustment. Child Development, 66, 710 –722.

Croninger, R. G., & Lee, V. E. (2001). Social capital and dropping out of

high school: Benefits to at-risk students of teachers’ support and guid-

ance. Teachers College Record, 103, 548 –581.

Darling, N., & Steinberg, L. (1993). Parenting style as context: An inte-

grative model. Psychological Bulletin, 113, 487– 496.

Devoe, J. F., Peter, K., Noonan, M., Snyder, T. D., & Baum, K. (2005).

Indicators of School Crime and Safety: 2005 (NCES Rep. No. 2006-

001/NCJ Rep. No. 210697). Washington, DC: U.S. Departments of

Education and Justice.

Evertson, C. M., Sanford, J. P., & Emmer, E. T. (1981). Effects of class

heterogeneity in junior high school. American Educational Research

Journal, 18, 219 –232.

Fenning, P. A., & Bohanon, H. (2006). Schoolwide discipline policies: An

analysis of discipline codes of conduct. In C. M. Evertson & C. S.

Weinstein (Eds.), Handbook of classroom management: Research, prac-

tice, and contemporary issues (pp. 1021–1039). Mahwah, NJ: Erlbaum.

Furstenberg, F. J., Eccles, J., Elder, G. J., Cook, T., & Sameroff, A. (1997).

Managing to make it. Chicago, IL: University of Chicago Press.

Gonzales, N., Cauce, A., Friedman, R., & Mason, C. (1996). Family, peer

and neighborhood influences on academic achievement among African-

American adolescents: 1 year prospective effects. American Journal of

Community Psychology, 24, 365–387.

Goodenow, C. (1993). Classroom belonging among early adolescent stu-

dents: Relationships to motivation and achievement. Journal of Early

Adolescence, 3, 21– 43.

Gorman-Smith, D., Tolan, P. H., & Henry, D. B. (2000). Patterns of family

functioning and adolescent outcomes among urban African American

and Mexican American families. Journal of Family Psychology, 14,

436 – 457.

Gottfredson, D. C. (2001). Schools and delinquency. Cambridge, MA:

Cambridge University Press.

Gottfredson, D. C., Gottfredson, G. D., & Hybl, L. G. (1993). Managing

adolescent behavior: A multiyear, multischool study. American Educa-

tional Research Journal, 30, 179 –215.

Gottfredson, G. D. (1999). User’s manual for the Effective School Battery.

Ellicott City, MD: Gottfredson Associates.

Gottfredson, G. D., Gottfredson, D. C., Payne, A. A., & Gottfredson, N. C.

(2005). School climate predictors of school disorder: Results from a

494 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

national study of commitment to prevention and schools. Journals of

Research in Crime and Delinquency, 42, 412– 444.

Gregory, A., & Cornell, D. (2009). “Tolerating” adolescent needs: Moving

away from zero tolerance policies in high school. Theory into Practice,

48, 106 –113.

Gregory, A., & Weinstein, R. S. (2004). Connection and regulation at home

and in school: Predicting growth in achievement for adolescents. Journal

of Adolescent Research, 19, 405– 427.

Gregory, A., & Weinstein, R. S. (2008). A window on the discipline gap:

Cooperation or defiance in the classroom. Journal of School Psychology,

46, 455– 475.

Hanson, T. L., & Austin, G. (2003). Student health risks, resilience, and

academic performance in California: Year 2 report, longitudinal anal-

yses. Los Alamitos, CA: WestEd.

Hardy, A. M. (1993). Regression with dummy variables. Newbury Park,

CA: Sage.

Hawker, D. S. J., & Boulton, M. J. (2000). Twenty years’ research on peer

victimization and psychosocial maladjustment: A meta-analytic review

of cross-sectional studies. Journal of Child Psychology and Psychiatry,

41, 441– 455.

Heaviside, S., Rowand, C., Williams, C., & Farris, E. (1998). Violence and

discipline problems in U.S. public schools: 1996 –97. (NCES Rep. No.

98 –030). Washington, DC: U.S. Department of Education, National

Center for Education Statistics.

Henrich, C. C., Brookmeyer, K. A., & Shahar, G. (2005). Weapon violence

in adolescence: Parent and school connectedness as protective factors.

Journal of Adolescent Health, 37, 306 –312.

Herman, M., Dornbusch, S., Hen-on, M., & Herting, J. (1997). The

influence of family regulation, connection and psychological autonomy

on six measures of adolescent functioning. Journal of Adolescent Re-

search, 12, 34 – 67.

Hollingsworth, E. J., Lufler, H. S., & Clune, W. H. (1984). School disci-

pline: Order and autonomy. New York, NY: Praeger.

Holt, M. K., & Espelage, D. L. (2003). A cluster analytic investigation of

victimization among high school students: Are profiles differentially

associated with psychological symptoms and school belonging? In M. J.

Elias & J. E. Zins (Eds.), Bullying, peer harassment, and victimization in

the schools:The next generation of prevention (pp. 81–98). New York,

NY: Haworth Press.

Holt, M. K., Finkelhor, D., & Kantor, G. K. (2007). Multiple victimization

experiences of urban elementary school students: Associations with

psychosocial functioning and academic performance. Child Abuse &

Neglect, 31, 503–515.

Hosp, J. L., & Reschly, D. J. (2004). Disproportionate representation of

minority students in special education: Academic, demographic, and

economic predictors. Exceptional Children, 70, 185–199.

Irvine, J. J. (2002). In search of wholeness: African American teachers and

their culturally competent classroom practices. New York, NY: Pal-

grave.

Jessor, R., Turbin, M. S., Costa, F. M., Dong, Q., Zhang, H., & Wang, C.

(2003). Adolescent problem behavior in China and the United States: A

cross-national study of psychosocial protective factors. Journal of Re-

search on Adolescence, 13, 329 –360.

Jimerson, S., & Furlong, M. (2006). The handbook of school violence and

school safety: From research to practice. Mahwah, NJ: Erlbaum.

Jo¨reskog, K., & So¨rbom, D. (2007). LISREL 8.8 for Windows [Computer

software and manual].Chicago, IL: Scientific Software International.

Kelley, M. L., Power, T. G., & Winbush, D. D. (1992). Determinants of

disciplinary practices in low-income Black mothers. Child Development,

63, 573–582.

Khoury-Kassabri, M., Benbenishty, R., Astor, R. A., & Zeira, A. (2004).

The contributions of community, family, and school variables to student

victimization. American Journal of Community Psychology, 34, 187–

204.

Koth, C. W., Bradshaw, C. P., & Leaf, P. J. (2008). A multilevel study of

predictors of student perceptions of school climate: The effect of

classroom-level factors. Journal of Educational Psychology, 100, 96 –

104.

Kreft, I., & de Leeuw, J. (1998). Introducing multilevel modeling. Newbury

Park, CA: Sage.

Ladd, B., & Ladd, G. W. (2001). Variations in peer victimization: Rela-

tions to children’s maladjustment. In J. Juvonen & S. Graham (Eds.),

Peer harassment in school: The plight of the vulnerable and victimized

(pp. 25– 48). New York, NY: Guilford Press.

LeBlanc, L., Swisher, R., Vitaro, F., & Tremblay, R. E. (2008). High

school social climate and antisocial behavior: A 10 year longitudinal and

multilevel study. Journal of Research on Adolescence, 18, 395– 419.

Loukas, A., Suzuki, R., & Horton, K. D. (2006). Examining school con-

nectedness as a mediator of school climate effects. Journal of Research

on Adolescence, 16, 491–502.

Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the

family: Parent– child interaction. In P. H. Mussen & M. E. Hetherington

(Eds.), Handbook of child psychology: Vol. 4. Socialization, personality

and social development (pp. 1–101).New York, NY: Wiley.

Maehr, M. L. (1991). The “psychological environment” of the school: A

focus for school leadership. Advances in Educational Administration, 2,

51– 81.

Mandara, J., & Murray, C. B. (2002). Development of an empirical

typology of African American family functioning. Journal of Family

Psychology, 16, 318 –337.

Mayer, M. J., & Leone, P. E. (1999). A structural analysis of school

violence and disruption: Implications for creating safer schools. Educa-

tion and Treatment of Children, 22, 333–356.

McConville, D., & Cornell, D. (2003). Attitudes toward aggression and

aggressive behavior among middle school students. Journal of Emo-

tional and Behavioral Disorders, 11, 179 –187.

Midgley, C., Maehr, M., Hruda, L., Anderman, E., & Freeman, K. (2000).

Manual for patterns of adaptive learning. Ann Arbor, MI: University of

Michigan.

Moos, R. H., & Moos, B. S. (1978). Classroom social climate and student

absences and grades. Journal of Educational Psychology, 70, 263–269.

Morrison, G. M., Redding, M., Fisher, E., & Peterson, R. (2006). Assessing

school discipline. In S. Jimerson & M. Furlong (Eds.), The handbook of

school violence and school safety: From research to practice (pp.

211–220).Mahwah, NJ: Erlbaum.

Morrison, G. M., & Skiba, R. (2001). Predicting violence from school

misbehavior: Promises and perils. Psychology in the Schools, 38, 173–

184.

National Center for Education Statistics. (2005). School crime supplement

to the National Crime Victimization Survey 2005. Retrieved October

2006 from http://nces.ed.gov/Programs/Crime/surveys.asp

National Center for Education Statistics. (2007). Indicators of school crime

and safety: 2007. Retrieved March 6, 2008, from http://nces.ed.gov/

programs/crimeindicators/crimeindicators2007/ind_03.asp

Nickerson, A. B., & Martens, M. P. (2008). School violence: Associations

with control, security/enforcement, educational/therapeutic approaches,

and demographic factors. School Psychology Review, 37, 228 –243.

Nickerson, A. B., & Spears, W. H. (2007). Influences on authoritarian and

educational/therapeutic approaches to school violence prevention. Jour-

nal of School Violence, 6, 3–31.

O’Connor, T. G., Allen, J. P., Bell, K. L., & Hauser, S. T. (1996).

Adolescent–parent relationships and leaving home in young adulthood.

In J. A. Graber & J. S. Dubas (Eds.), Leaving home: Understanding the

transition to adulthood (pp. 39 –52). San Francisco, CA: Jossey-Bass.

Olweus, D., & Limber, S. (2000). Blueprints for violence prevention, book

nine: Bullying prevention program. Golden, CO: Venture Publishing.

Ozer, E. J., & Weinstein, R. S. (2004). Urban adolescents’ exposure to

community violence: The role of support, school safety, and social

495

AUTHORITATIVE SCHOOL DISCIPLINE

constraints in a school-based sample of boys and girls. Journal of

Clinical Child and Adolescent Psychology, 33, 463– 476.

Park, H., & Bauer, S. (2002). Parenting practices, ethnicity, socioeconomic

status and academic achievement in adolescents. School Psychology

International, 23, 386 –397.

Pellerin, L. A. (2005). Applying Baumrind’s parenting typology to high

schools: Toward a middle-range theory of authoritative socialization.

Social Science Research, 34, 283–303.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models:

Applications and data analysis methods (2nd ed.). Newbury Park, CA:

Sage.

Roberts, S. D., Wilcox, P., May, D. C., & Clayton, R. R. (2007). My school

or our school: The effects of individual versus shared school experiences

on teacher perceptions of safety. Journal of School Violence, 6, 33–55.

Rosenthal, R. (1990). How are we doing in soft psychology? American

Psychologist, 45, 775–777.

Selman, R. L., & Hickey Schultz, L. (1990). Making a friend in youth:

Developmental theory and pair therapy. Chicago, IL: University of

Chicago Press.

Shinn, M., & Rapkin, B. D. (2000). Cross-level analysis without cross-ups.

In J. Rappaport & E. Seidman (Eds.), Handbook of community psychol-

ogy (pp. 669 –695). New York, NY: Plenum Press.

Shumow, L., Vandell, D. L., & Posner, J. (1999). Harsh, firm, and per-

missive parenting in low-income families: Relations to children’s aca-

demic achievement and behavioral adjustment. Journal of Family Issues,

19, 483–507.

Skiba, R., Michael, R. S., Nardo, A. C., & Peterson, R. (2002). The color

of discipline: Sources of racial and gender disproportionality in school

punishment. Urban Review, 34, 317–342.

Skiba, R., & Peterson, R. (1999, March). The dark side of zero tolerance:

Can punishment lead to safe schools? Phi Delta Kappan, 372–379.

Smetana, J. G. (2005). Adolescent–parent conflict: Resistance and subver-

sion as developmental process. In L. Nucci (Ed.), Conflict, contradic-

tion, and contrarian elements in moral development and education (pp.

69 –91). Mahwah, NJ: Erlbaum.

Smetana, J., & Gaines, C. (1999). Adolescent–parent conflict in middle-

class African American families. Child Development, 70, 1447–1463.

Smith, P. K., Talamelli, L., & Cowie, H. (2004). Profiles of non-victims,

escaped victims, continuing victims and new victims of school bullying.

British Journal of Educational Psychology, 74, 565–581.

Sprott, J. B. (2004). The development of early delinquency: Can classroom

and school climates make a difference? Canadian Journal of Criminol-

ogy and Criminal Justice, 46, 553–572.

Steinberg, L., Lamborn, S. D., Darling, N., & Mounts, N. S. (1994).

Over-time changes in adjustment and competence among adolescents

from authoritative, authoritarian, indulgent, and neglectful families.

Child Development, 65, 754 –770.

Steinberg, L., Lamborn, S. D., Dornbusch, S. M., & Darling, N. (1992).

Impact of parenting practices on adolescent achievement: Authoritative

parenting, school involvement, and encouragement to succeed. Child

Development, 63, 1266 –1281.

Stewart, E. A. (2003). School social bonds, school climate, and school

misbehavior: A multilevel analysis. Justice Quarterly, 20, 575– 604.

Stronach, I., & Piper, H. (2008). Can liberal education make a comeback?

The case of “relational touch” at Summerhill School. American Educa-

tional Research Journal, 45, 6 –37.

Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York,

NY: Norton.

Tatum, B. D. (1997). “Why are all the Black kids sitting together in the

cafeteria?” And other conversations about race. New York, NY: Basic

Books.

Taylor, L. C., Hinton, I. D., & Wilson, M. N. (1995). Parental influences

on academic performance in African-American students. Journal of

Child and Family Studies, 4, 293–302.

Tebo, M. G. (2000, April). Zero tolerance, zero sense. American Bar

Association Journal, 86, 40 – 46.

Trickett, E. J., & Moos, R. H. (1974). Personal correlates of contrasting

environments: Student satisfactions in high school classrooms. Ameri-

can Journal of Community Psychology, 2, 1–12.

Turiel, E. (2005). Resistance and subversion in everyday life. In L. Nucci

(Ed.), Conflict, contradiction, and contrarian elements in moral devel-

opment and education (pp. 3–20). Mahwah, NJ: Erlbaum.

Tyler, T. (2006). Psychological perspectives on legitimacy and legitima-

tion. Annual Review of Psychology, 57, 375– 400.

U.S. Department of Education, National Center for Education Statistics.

(2000). Indicators of school crime and safety, 2000. Retrieved from

http://nces.ed.gov/pubs2001/crime2000/

U.S. Department of Health and Human Services. (2001). Youth violence: A

report of the Surgeon General. Rockville, MD: Author.

Unnever, J., & Cornell, D. (2003). The culture of bullying in middle

school. Journal of School Violence, 2, 5–27.

Unnever, J., & Cornell, D. (2004). Middle school victims of bullying: Who

reports being bullied? Aggressive Behavior, 30, 373–388.

Vasquez, J. A. (1988). Context of learning for minority students. Educa-

tional Forum, 56, 6 –11.

Virginia Department of Education. (2007). Enrollment & demographics:

Fall membership. Retrieved from http://www.doe.virginia.gov/

statistics_reports/enrollment/fall_membership/index.shtml

Walker, J. M. (2008). Looking at teacher practices through the lens of

parenting style. Journal of Experimental Education, 76, 218 –240.

Wardle, F. (1992). Supporting biracial children in the school setting.

Education & Treatment of Children, 15, 163–172.

Welsh, W. (2000). The effects of school climate on school disorder. Annals

of the American Academy of Political and Social Science, 567, 88 –107.

Wentzel, K. R. (2002). Are effective teachers like good parents? Teaching

styles and student adjustment in early adolescence. Child Development,

73, 287–301.

West, C. A. (1985). Effects of school climate and school social structure on

student academic achievement in selected urban elementary schools.

Journal of Negro Education, 54, 451– 461.

Williams, F., & Cornell, D. (2006). Student willingness to seek help for

threats of violence. Journal of School Violence, 5, 35– 49.

Wilson, C. J., & Deanne, F. P. (2001). Adolescent opinions about help-

seeking barriers and increasing appropriate help engagement. Journal of

Educational and Psychological Consultation, 12, 345–364.

Received March 2, 2009

Revision received November 23, 2009

Accepted November 29, 2009 䡲

496 GREGORY, CORNELL, FAN, SHERAS, SHIH, AND HUANG

A preview of this full-text is provided by American Psychological Association.

Learn more

Content available from Journal of Educational Psychology

This content is subject to copyright. Terms and conditions apply.

Geographical Variation in the Associations Between School Characteristics and Homophobic Bullying: a Contextual Analysis

Article

Full-text available

Sep 2024
PREV SCI

Homophobic bullying constitutes a serious threat to adolescent well-being and could be understood as an ecological phenomenon, influenced by diverse school, regional, and community contexts. This study examined geographic variations in the relationship between school characteristics and homophobic bullying. Data from 2244 California schools, including student surveys and administrative records, were combined and analyzed using geographically weighted regression (GWR). Results showed that the associations between school characteristics and general victimization were consistent across geographic areas. However, when it came to homophobic bullying, the relationships with school characteristics varied significantly based on location. Notably, regions with high intolerance, urbanity, large school size, and small student–teacher ratios appeared to offer protection against bullying. Additionally, student socioeconomic status influenced bullying in disadvantaged rural schools with limited Gay-Straight Alliance (GSA) presence. Ethnic diversity also played a role, with low diversity or dominance of two ethnic groups linked to higher bullying rates.

Association between bullying victimization and mental health problems among Chinese left-behind children: a cross-sectional study from the adolescence mental health promotion cohort

Article

Full-text available

Nov 2024

Background Left-behind children (LBC) refer to those who have been separated from at least one parent for six months or more due to parental migration for work. This phenomenon poses a significant threat to the mental health of over 61 million LBC in China. This study aims to compare the prevalence of mental health symptoms between LBC and non-left-behind children (non-LBC) and to explore the predictive effect of bullying victimization on adolescent mental health problems. Methods In 2019, we conducted a cross-sectional analysis involving 28,036 children and adolescents in Mianyang City, Sichuan Province, China, with ages ranging from 8 to 19 years. Mental health symptoms were assessed using the 7-item Generalized Anxiety Disorder (GAD-7) scale and the 9-item Patient Health Questionnaire (PHQ-9). The Delaware Bullying Victimization Scale-Student (DBVS-S) was employed to gather data on experiences of bullying victimization. Information on self-injury was collected by inquiring whether participants had engaged in self-injurious behavior and the reasons for such behavior. Multivariable logistic regression was utilized to analyze the risk and protective factors associated with mental health symptoms, with a particular focus on different types of bullying victimization. Results Compared to non-left-behind children (non-LBC), left-behind children (LBC) exhibited a higher prevalence of mental health issues: anxiety symptoms (24.0% vs. 18.0%, p<0.001), depressive symptoms (27.9% vs. 19.4%, p<0.001), and self-injurious behavior (17.7% vs. 12.2%, p<0.001). Among LBC, physical bullying was identified as the most significant predictor of anxiety symptoms (OR = 1.62). Additionally, LBC who experienced verbal bullying had a higher risk of depressive symptoms (OR = 2.23) and self-injurious behaviors (OR = 1.54). Enhanced family functioning, positive teacher-student relationships, and strong peer relationships were found to offer protective effects against mental health problems. Conclusion Our results suggested that LBC experienced a higher incidence of mental health symptoms, particularly among those who had been victims of bullying. This underscores the urgent need for supportive strategies focused on the school environment and interpersonal relationships to mitigate negative mental health outcomes for LBC.

Teachers’ Roles in Coping with School Violence from the Perspectives of Prospective Teachers: A Q Methodological Approach

Article

Full-text available

Nov 2024

This study explores prospective teachers’ perceptions of school violence and their role in addressing it. Using a mixed method called Q methodology, we quantitatively analyzed the subjective views of 37 prospective teachers. Based on 33 statements, the study categorizes teachers’ roles in managing school violence and analyzes the characteristics of each role type. The research results provide basic data for prospective teachers to develop their capabilities as experts in dealing with school violence. The study identifies the following three types of teachers: prevention-oriented (Type 1), reality-avoiding (Type 2), and legal-regulation-oriented (Type 3). Type 1 emphasizes that teachers can significantly prevent school violence and believes that trusting relationships between teachers and students are key to reducing school violence. Type 2 teachers tend to ignore school violence out of fear of harm, even though they acknowledge its seriousness. Type 3 emphasizes the strict enforcement of legal responsibilities and rules, believing that clear regulations and punishments are essential for reducing school violence. This study highlights the need for customized educational programs that reflect the characteristics of different teacher types in handling school violence. It suggests incorporating practical strategies for preventing and responding to school violence in teacher training, expanding mental-health education, and promoting cooperative conflict-resolution methods between students and teachers.

Empowering Teacher Intention to Intervene: The Role of Self-Efficacy in Addressing Homophobic Bullying

Article

Full-text available

Nov 2024

Teacher interventions play a crucial role in fostering a more inclusive school climate amidst homophobic bullying incidents. However, the strategies employed by teachers and the influencing factors are understudied. This study explored individual and contextual factors associated with teachers' intentions to intervene in situations of homophobic bullying. It emphasized the role of self-efficacy and how its impact may be influenced by school efforts to combat homophobic bullying. A sample of 465 teachers (76.34% women, Age: M = 49.58, SD = 10.25) from nine primary and secondary schools in central Italy completed a scale assessing different types of behaviors they would employ in response to homophobic bullying episodes, including non-intervention, disciplinary sanctions, peer mediation, victim support, and classroom discussion. The study also examined teachers' perceived school effort to address homophobic bullying and self-efficacy in handling educational tasks and homophobic bullying situations. Age, gender, grade, training on bullying, perceived prevalence of homophobic bullying incidents, and witnessing such incidents were considered as covariates. Linear regression showed that self-efficacy (general and homophobic bullying-related) was positively associated with intentions to use all intervention types and negatively non-intervention. Non-intervention was more common among teachers who reported more homophobic bullying incidents among students, while witnessing incidents was associated with less victim support and classroom discussion. The interaction of general self-efficacy and school effort was significant for disciplinary sanctions and victim support, indicating that in schools with low effort to address bullying, low teacher efficacy was associated with fewer intentions to intervene. The findings emphasize the importance of a twofold approach: fostering a strong school effort to address homophobic bullying and offering comprehensive teacher training to enhance self-efficacy in addressing these incidents.

Adolescent Exposure to Restorative Practices and Their Perceptions of Support, Structure, and Bullying in the School Climate

Article

Full-text available

Oct 2024

This study investigated whether a restorative practice initiative improved school climate. The study presents findings from a cluster randomized, controlled trial conducted with 2,248 students from grades 5 to 12 (38% Black, 32% Hispanic) attending 18 schools in the U.S. Northeast. After 1 year of implementation, the experimental analyses of the whole-school initiative did not result in more positive perceptions of school climate for students in the intervention schools compared with students in the control schools. Correlational analyses were also conducted with student surveys in both intervention and control schools. Based on student report, restorative practice exposure in all 18 schools was associated with positive perceptions of the school climate. This suggests that initiatives may need to focus on fidelity of implementation and consistent implementation across more years for substantive school climate gains.

DETERMINANTS OF BULLYING AMONG PUBLIC SECONDARY SCHOOL STUDENTS IN MUSHIN, LAGOS STATE, NIGERIA

Article

Full-text available

Jul 2024

p>The rate at which bullying is been perpetrated in schools on a daily basis is alarming and worrisome. It is a big threat that should not be handled with levity. Bullying is not only dangerous to the victims but also to the perpetrators, parents, schools and the society. The consequences could have a life-long effect on the individual that is bullied. In view of this, this study, therefore, examined the school climate, peer influence, social competence and family cohesion as facilitators of bullying among public secondary school students in Mushin, Lagos State. Two hundred and two (202) public secondary school students from 6 (six) public schools in Mushin, Lagos State were randomly selected for the study. Standardized instruments were used to measure the variables. Three research questions were raised and answered. Data collected were analyzed using the Multiple Regression Analysis and Pearson Product Moment Correlation at 0.05 level of significance. Findings of the study showed that bulling among the participants correlated positively with the four predicting variables and a coefficient of multiple correlations (R) of 0.812 and adjusted R square of 0.646 was noted. Family cohesion is the most potent contributor to the prediction (β = 0.308, t = 6.938, p < 0.05) followed by peer influence (β = 0.292, t = 6.017, p < 0.05), school climate (β = 0.126, t = 2.466, p < 0.05) and finally, social competence (β = 0.015, t = 0.343, p < 0.05). Based on the findings, it was recommended that counselling psychologists and allied professionals should endeavour to sensitize the government, school owners and public on the benefits of positive school climate and family cohesion in reducing bullying in the school system. Also, bullying may be reduced through anti-bullying programs in schools that serve to raise awareness of the problem. Lastly, cordial relationship should be established at homes among the siblings and between the parents and their children as this will go a long way in raising children with high social competence skills as well as promoting healthy relationship that fosters friendliness in school setting rather than bullying among the students. Article visualizations: </p

Mengapa Siswa SMP di Indonesia melakukan Perundungan?: Menelusuri Faktor-Faktor Penyebab Kekerasan di Sekolah

Article

Full-text available

Aug 2024

Wafi Mohammad Sholihul

Bullying is an aggressive behavior that has become a significant issue among junior high school students in Indonesia. This study employed a literature review method to identify the underlying reasons behind this phenomenon. To provide a comprehensive explanation, this study examined three primary perspectives: peers, school environment, and parental influence, to answer the research question, "Why do junior high school students in Indonesia bully their peers?" The findings indicate that bullying among peers is often influenced by social constructions, such as group norms or peer pressure. Within the school environment, factors contributing to bullying include weak enforcement of regulations and non-compliance with existing rules. Meanwhile, inappropriate parenting styles and poor quality relationships between parents and children also contribute to bullying behavior. This research is expected to serve as a reference for bullying prevention efforts among students. However, further research is still needed to identify additional factors that may contribute to the bullying phenomenon in secondary schools. Perundungan merupakan perilaku agresif yang telah menjadi masalah signifikan di kalangan siswa sekolah menengah pertama (SMP) di Indonesia. Penelitian ini menggunakan metode kajian pustaka untuk mengidentifikasi alasan-alasan di balik fenomena tersebut. Dalam upaya memberikan penjelasan yang komprehensif, penelitian ini mengkaji tiga perspektif utama: teman sebaya, lingkungan sekolah, dan pengaruh orang tua, untuk menjawab pertanyaan penelitian, "Mengapa siswa SMP di Indonesia melakukan perundungan terhadap teman sebaya mereka?" Temuan penelitian menunjukkan bahwa perundungan di antara teman sebaya sering kali dipengaruhi oleh konstruksi sosial, seperti norma kelompok atau tekanan teman sebaya. Di lingkungan sekolah, faktor yang berkontribusi terhadap perundungan meliputi lemahnya penegakan peraturan dan ketidakpatuhan terhadap aturan yang ada. Sementara itu, gaya pengasuhan yang kurang tepat serta kualitas hubungan yang buruk antara orang tua dan anak turut menjadi faktor yang berpengaruh terhadap perilaku perundungan. Penelitian ini diharapkan dapat menjadi referensi dalam upaya pencegahan perundungan di kalangan siswa. Namun, penelitian lebih lanjut tetap diperlukan untuk mengidentifikasi faktor-faktor tambahan yang mungkin berkontribusi terhadap fenomena perundungan di sekolah menengah.

Longitudinal Association Between School Climate and Psychological Flexibility and Mental Health: A Random Intercept Cross Lagged Panel Model

Article

Full-text available

Aug 2024
J PSYCHOPATHOL BEHAV

Previous research found the effect of school climate on students’ psychological distress. No longitudinal studies examined the effect of school climate on college students’ psychological flexibility and the possible mediating role of psychological flexibility between school climate and psychological distress. These were the primary purpose of the present study. Medical college students were recruited starting in September 2020; assessments were conducted at five points, with intervals of two months between each assessment. Self-reported questionnaires were completed by 210 college students (33.3% male and 66.7% female), with an average age of 20.62. The Perceived School Climate scale, Acceptance and Action Questionnaire-II, and Depression Anxiety Stress Scales (DASS-21) were used to assess participants’ perceived school climate, psychological flexibility, and mental health (i.e., depression, anxiety, and stress) respectively. The temporal directional relationship between study variables was analyzed using the random intercept cross-lagged panel model. There was a unidirectional positive effect from perceived school climate to psychological flexibility, and a unidirectional negative effect from perceived school climate to anxiety and depression. There was no directional effect between psychological flexibility and mental health at the within-person level, but at the between-person level, they were strongly associated. Furthermore, psychological flexibility did not mediate the relationship between perceived school climate and mental health. This study supported the bio-ecological theory and highlighted the importance of fostering a positive school climate to improve college students’ psychological flexibility and decrease their depression and anxiety.

The Problematic Internet Use and Its Impact on Young People’s Online Moral Disengagement

Chapter

Jul 2024

Areej Elsayary

This chapter examines the problematic use of the internet and its impact on young people’s online moral disengagement. With the widespread availability and accessibility of online platforms, young individuals are increasingly exposed to various online activities that can potentially undermine their moral judgment and ethical behavior. This research investigates the relationship between excessive internet usage, including addictive behaviors, and the tendency of young people to disengage morally when interacting online. This study sheds light on how the Internet influences moral disengagement among young individuals through a comprehensive literature review and empirical analysis. The findings contribute to our understanding of the negative consequences associated with the problematic use of the internet and provide valuable insights for parents, educators, and policymakers to mitigate the impact on young people’s moral development.

Association between childhood emotional abuse and non-suicidal self-injury among Guangxi adolescents in China: A moderated mediation model

Article

Jul 2024
J AFFECT DISORDERS

Identification of School Bullies by Survey Methods

Article

Full-text available

Jan 2006

How can middle school counselors identify bullies? This study compared two methods of identifying bullies in a sample of 386 middle school students. A peer nomination survey identified many more bullies than did student self-report. Moreover, self-reported and peer-nominated bullies differed in their types of bullying behaviors, level of general self-concept, attitudes toward aggression, and disciplinary infractions. Overall, this study raises concern about reliance on student self-report and supports the use of peer nomination as a means of identifying school bullies. These findings have implications for school counselors in undertaking efforts to reduce school bullying.

Connection and regulation at home and in school: Predicting growth in achievement for adolescents

Article

Full-text available

Jul 2004

Qualities of adolescent-adult relationships across home and school environments are examined as predictors of academic growth in mathematics. An ethnically diverse sample of adolescents was drawn from the National Educational Longitudinal Study, 1988. In separate analyses, adolescents' perceptions of (a) connection with parents and teachers and (b) regulation from parents and teachers uniquely predicted academic growth in math from 8th to 12th grade. Thus, assets across home and school were additive. No evidence supported a compensatory process in which less connection or regulation tit home was compensated by the presence of these experiences in school. Within school, teacher connection was the strongest predictor for all adolescents, but a combination of connection and regulation, making up an authoritative teaching style, predicted even greater academic growth in math for adolescents from low socioeconomic backgrounds.

Regression with Dummy Variables

Article

Aug 1994
TECHNOMETRICS

Managing Adolescent Behavior A Multiyear, Multischool Study

Article

Sep 1994

Zero tolerance, zero sense

Article

Apr 2000

MG Tebo

The Dark Side of Zero Tolerance: Can Punishment Lead to Safe Schools?

Article

Jan 1999

Seriousness of purpose in seeking to avert the tragedy of school violence does not necessarily demand rigid adherence to harsh and extreme measures. There are alternatives to politically facile get-tough strategies, the authors point out.

Schools and Delinquency

Article

Jan 2011

Gary D. Gottfredson

This article examines the nature of delinquent and related problem behavior in schools. It suggests that public perceptions that the quality of many urban schools is low has the effect of exacerbating the concentration of populations of young people at elevated risk of both delinquent behavior and poor educational outcomes in some communities. It describes delinquency and related problem behaviors in schools and suggests that delinquents and dropouts are engaged in a variety of problem behaviors, and they are low achieving, poorly motivated, and uncommitted to school. It also discusses whether individual characteristics predispose young people to problem behavior and poor school achievement. Furthermore, the chapter reveals the implications of school demography for delinquency and educational outcomes. Finally, it presents an argument concerning whether or not schools should be concerned with preventing delinquency.

Variations in peer victimization: Relations to children's maladjustment

Article

Jan 2001

Cross-Level Research without Cross-Ups in Community Psychology

Chapter

Jan 2000

A central tenet of community psychology is that human behavior must be understood in context. Community and other ecologically oriented psychologists have proposed typologies of contexts for behavior (Altman &Rogoff, 1987 ;Barker, 1968; Bronfenbrenner, 1979, 1986; Murrell, 1973; Rappaport, 1977; Seidman, 1988), with the twin goals of understanding the interplay of people and contexts, and of exhorting psychologists to include ever-broader contexts of behavior in their theory, research, and intervention practices. Typically, the typologies are organized from narrower to broader contexts or from lower to higher levels of analysis

How are we doing?

Article

May 2002
AM J NURS

Marie G. LoGuercio

Authoritative School Discipline: High School Practices Associated With Lower Bullying and Victimization

Abstract

Recommended publications

Adolescent Female Gender Identity and Attraction to Male Bullies and Victims

Sexting in adolescence: frequency and association with victimization of cyberbullying and dating vio...

Associations of Timing of Sexual Orientation Developmental Milestones and Other Sexual Minority Stre...

National Prevalence Rates of Bully Victimization Among Students With Disabilities in the United Stat...