See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/334251411

Comparing the Impact of Bullying Victimization on Drug Use and Weapon

Carrying among Male and Female Middle and High School Students: A Partial
Test of General Strain Theory

Article  in  Deviant Behavior · July 2019

DOI: 10.1080/01639625.2019.1637405

CITATIONS READS

5 366

3 authors, including:

Caitlin M. Brady Thomas Baker

Georgia Southern University University of Central Florida
13 PUBLICATIONS   73 CITATIONS    36 PUBLICATIONS   754 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Youth Crisis Response Program in Seminole County, Florida View project

All content following this page was uploaded by Caitlin M. Brady on 22 September 2019.

The user has requested enhancement of the downloaded file.

 Deviant Behavior

ISSN: 0163-9625 (Print) 1521-0456 (Online) Journal homepage: https://www.tandfonline.com/loi/udbh20

Comparing the Impact of Bullying Victimization on

Drug Use and Weapon Carrying among Male and
Female Middle and High School Students: A Partial
Test of General Strain Theory

Caitlin M. Brady, Thomas Baker & William V. Pelfrey Jr

To cite this article: Caitlin M. Brady, Thomas Baker & William V. Pelfrey Jr (2019): Comparing
the Impact of Bullying Victimization on Drug Use and Weapon Carrying among Male and Female
Middle and High School Students: A Partial Test of General Strain Theory, Deviant Behavior, DOI:
10.1080/01639625.2019.1637405

To link to this article: https://doi.org/10.1080/01639625.2019.1637405

Published online: 04 Jul 2019.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=udbh20
DEVIANT BEHAVIOR
https://doi.org/10.1080/01639625.2019.1637405

Comparing the Impact of Bullying Victimization on Drug Use and

Weapon Carrying among Male and Female Middle and High
School Students: A Partial Test of General Strain Theory
Caitlin M. Bradya, Thomas Bakera, and William V. Pelfrey Jrb
a
University of Central Florida, Orlando, USA; bVirginia Commonwealth University, Richmond, USA

ABSTRACT ARTICLE HISTORY

Prior research testing general strain theory (GST) suggests potential differ- Received 29 May 2019
ences in how males and females respond to strain. Using a sample of 3,139 Accepted 21 June 2019
middle and high school students (1,515 males and 1,624 females), this study
evaluates male and female delinquent coping behaviors to traditional
bullying and cyberbullying victimization. Three delinquent coping beha-
viors are investigated: soft drug use, hard drug use, and weapon carrying.
This study also examines the impact of bullying on anticipated strain and
tests whether anticipated strain mediates the relationship between bullying
and delinquency. Results from group-based structural equation models
indicate sex differences in the effect of bullying on delinquent coping
behavior. Specifically, bullying victimization has a significantly larger asso-
ciation with anticipated strain for males and cyberbullying victimization has
a significantly larger association with drug use and weapon carrying for
males. Anticipated strain partially mediates the relationship between bully-
ing and delinquency. Policy implications and suggestions for future
research are discussed.

In a follow up to his original positing of General Strain Theory (GST), Agnew (2001) directly
identified bullying, or “peer abuse,” as a key strain related to delinquent coping. Unsurprisingly, this
led to a great deal of bullying research framed around the propositions of GST (Agnew 1992)and
Agnew’s (2001, 2002)extensions thereof. Findings show considerable support for the association
between bullying victimization and delinquency (Cullen et al. 2008; Hay, Meldrum, and Mann 2010;
Patchin and Hinduja 2011). For example, in a recent test of GST and bullying Baker and Pelfrey
(2016) found that both traditional bullying and cyberbullying are associated with anticipated strain –
a concept exemplified by emotions like fear of future victimization (Agnew 2002). In addition, both
types of bullying were associated, directly and indirectly through anticipated strain, with soft and
hard drug use. They also linked cyberbullying with weapon carrying (Baker and Pelfrey 2016).
Few studies, though, have tested the propositions of GST and bullying victimization within the
context of differences between females and males. Broidy and Agnew (1997) were among the first to
identify ways in which strain and delinquency may be moderated by sex. Studies since have
supported their hypotheses (e.g. Higgins, Piquero, and Piquero 2011; Kaufman 2009; Piquero and
Sealock 2004) finding sex differences in strain and deviancy. Such research suggests variations in
coping with bullying victimization may be due in part to how males and females cope with strain in
general (Broidy and Agnew 1997). In response to bullying victimization more specifically, it appears
males and females may employ different mechanisms for coping (Stubbs-Richardson et al. 2018).
Stubbs-Richardson et al. (2018) found that males were more likely to cope with bullying in an

CONTACT Caitlin M. Brady caitlin.brady@ucf.eduDepartment of Criminal Justice,University of Central Florida,12805 Pegasus

Drive, Bldg. 80, Suite 311, Orlando,FL 32816
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/udbh.
© 2019 Taylor & Francis Group, LLC
2 C. M. BRADY ET AL.

antisocial manner compared to females. Although, another recent study suggests that the impact of
experienced and vicarious strains on violent offending were not moderated by sex (Lee and Kim
2018). As such, it is unclear how females and males differentially respond to bullying. In addition,
studies to date have yet to explore sex differences in the impact of bullying on the anticipated strain
or whether anticipated strain affects the relationship between bullying and delinquent coping
differently for females compared to males.
Using GST as a theoretical framework, the current study builds on prior research by Broidy and
Agnew (1997), Agnew (2002), and Baker and Pelfrey (2016) to study sex differences in the associa-
tion of bullying and delinquent coping among middle and high school students. Specifically, we
examine the possibility that the association between bullying victimization (both traditional and
cyberbullying) and delinquent coping in the form of drug use and weapon carrying differs for males
and females. In addition, we examine the association between bullying victimization and anticipated
strain, as well as the potential that anticipated strain mediates the relationship between bullying and
delinquent coping. Prior to discussing our findings, we briefly review the literature on GST and
bullying.

Literature review
General strain theory
Agnew (1992) posited three sources of strain: the failure to achieve positively valued goals, the real or
anticipated removal of positively valued stimuli, and the presentation of negative stimuli. A large
body of prior literature consistently supports the key propositions of GST, showing that individuals
exposed to different types of strain are more likely to engage in delinquent behaviors (Agnew 2002;
Agnew et al. 2002; Aseltine, Gore, and Gordon 2000; Mazerolle and Maahs 2000; Mazerolle and
Piquero 1997; Moon, Morash, and McCluskey 2012; Peck et al. 2018; Piquero and Sealock 2004).
Importantly for the study of bullying victimization, Agnew (2001) identified bullying or what he
termed “peer abuse” as a key negative stimuli leading to delinquent coping.
Specifically, Agnew (2001) suggested that bullying (or “peer abuse”) should be important because
it satisfies four conditions that indicate how influential a strainful experience will be: (1) it is
perceived as unjust, (2) it is perceived as important, (3) it is outside the frame of conventional social
controls, (4) it exposes the individual to others who display aggressive behavior. In a study of
adolescents, Agnew (2002) found that youth who were picked on by neighborhood peers were more
involved with delinquent behavior; however, this was true only for those with personality character-
istics favorable to delinquency. The applicability of elements of GST to bullying victimization has
also been suggested by the findings from several other studies (Agnew et al. 2002; Baker and Pelfrey
2016; Cullen et al. 2008; Hay and Meldrum 2010; Hay, Meldrum, and Mann 2010; Hinduja and
Patchin 2007; Moon, Morash, and McCluskey 2012; Patchin and Hinduja 2011).
Bullying victimization and GST research have also begun focusing on the issue of cyberbullying
(Hinduja and Patchin 2008), which involves using the Internet or cell phones as the mechanism
through which the bullying occurs. This includes abusive e-mails or text messages and insulting
messages or pictures on online message boards (Hay, Meldrum, and Mann 2010). The effects of
cyberbullying, especially among school-aged children, are important to explore as research indicates
that of individuals who have not yet graduated high school: 92% own cellphones, 54% own
smartphones, and 39% own other mobile devices (Pew Research Center 2017). Also, unlike tradi-
tional bullying, cyberbullying may be especially difficult to escape. The electronic nature may allow it
to occur without attracting the attention of adults or other authority figures such as teachers.
Furthermore, because many adolescents constantly carry their cell phone, they can be exposed to
cyberbullying more often and even when physically removed from bullies. Thus, individuals who are
cyberbullied may fear for their safety offline due to harm and intimidation online.
 DEVIANT BEHAVIOR 3

Anticipated strain
When Agnew (2001, 2002) elaborated on the GST framework by identifying strains that would be
most influential in leading to delinquent behavior, he also suggested that there are three different
types of strain: (1) experienced, (2) vicarious, and (3) anticipated. Experienced strains are those that
directly happen to a person and lead to a specific coping strategy. Vicarious strains are strainful
events that happen to others such as a family member or friends (Agnew 2002). Anticipated strains
refer to the individual’s expectations that current strains will continue into the future or that new
strains will be experienced (Agnew 2002).
Recent studies have examined anticipated strain by operationalizing it in terms of fear or
perceived safety (Baker and Pelfrey 2016; Baron 2009; Keith 2018; McGrath, Marcum, and Copes
2012). As noted above, Baker and Pelfrey (2016) examined middle and high school students and
found that experienced strain is associated with anticipated strain, indicating that experienced strain
may exacerbate fear of future victimization increasing the total effect of strain on delinquent coping.
Among Canadian street youths, Baron (2009) found no direct impact of anticipated strain on violent
delinquency but found that the relationship may be moderated by other factors. McGrath, Marcum,
and Copes (2012) also explored anticipated risk and fear of victimization while in prison among
male parolees. They found no significant effect of anticipated risk and fear on self-reported violent
behavior or drug use (McGrath, Marcum, and Copes 2012). Most recently, Keith (2018) found that
traditional bullying and cyberbullying were associated with greater fear. Subsequently, when fear was
introduced into the model, the direct relationships between traditional bullying and cyberbullying on
avoidance behaviors were reduced (Keith 2018). Although Keith (2018) concluded there was no
mediation effect of fear, no actual mediation tests were reported, and the reduced effect sizes suggest
the possibility for partial mediation. So, while evidence exists that experienced strain may impact
anticipated strain, acting as a potential mediator between experienced strain and coping as Agnew
(2002) posited, only Baker and Pelfrey (2016) and Keith (2018) examined this possibility and only
Baker and Pelfrey (2016) directly tested for mediation.

Sex differences in response to strain

Broidy and Agnew (1997) expanded upon the general strain theory framework by introducing the
idea that males and females may differ in their levels of exposure and responses to strain. They
posited that when experiencing strain males and females may differ in how they cope emotionally
and behaviorally. They offered several explanations for these differences though the most important
for the present study can be summarized as follows: males may have different behavioral reactions to
strain compared to females, while females may have different emotional reactions to strain compared
to males (Broidy and Agnew 1997). Despite this work, limited prior research has explored sex
differences when examining the effect of bullying victimization on delinquent coping. However, in
one of the few studies directly evaluating these differences, Cullen et al. (2008) found that among
middle school students, bullied males were more likely to engage in substance use than bullied
females. Additionally, though not a direct study of bullying victimization, Jennings et al. (2009)
found differences in males and females property offending as a result of peer abuse. However, they
found that males and females did not significantly differ in aggressive coping in response to peer
abuse. Research exploring sex differences in coping with both traditional bullying victimization and
cyberbullying victimization is limited, and none have explored the effects of anticipated strain.

Current study
As noted above, several developments in the theoretical framework have emerged since Agnew’s
(1992) original conceptualization (2002; Agnew 2001). Particularly, the ideas that peer abuse such as
traditional bullying and cyberbullying may lead to delinquent coping, and that anticipated strain
4 C. M. BRADY ET AL.

may play a mediating role in coping with experienced strains. Prior literature examining the effect of
bullying victimization on delinquency has generally focused on either traditional bullying or
cyberbullying, and those that have included both have not assessed them in the same statistical
models (e.g. Hay, Meldrum, and Mann 2010). Furthermore, studies examining the impact of
anticipated strain within the bullying literature are limited and do not include sex differences to
explore different coping reactions by males and females. To address these gaps in the literature, the
current study seeks to examine sex differences in delinquent coping for traditional bullying and
cyberbullying victimization while incorporating anticipated strain. Based on prior findings and the
general strain literature we test six hypotheses.

Hypothesis 1: The experienced strains of traditional bullying and cyberbullying will be signifi-
cantly and positively associated with the delinquent coping behaviors of soft drug use, hard
drug use, and carrying a weapon at school.

Hypothesis 2: The experienced strains of being traditionally bullied and cyberbullied will be
significantly and positively associated with anticipated strain (fear of victimization).

Hypothesis 3: Anticipated strain will be significantly and positively associated with the delin-
quent coping behaviors of soft drug use, hard drug use, and carrying a weapon at school.

Hypothesis 4: The effect of the experienced strains on soft drug use, hard drug use, and weapon
carrying will be greater for males compared to females.

Hypothesis 5: The effect of the experienced strains on anticipated strain will be greater for
females compared to males.

Hypothesis 6: The experienced strains of being traditionally bullied and cyberbullied will have
a significant and positive indirect association with delinquent coping mediated by anticipated
strain.

Methods
Sample
A short form version of the Youth Risk Behavior Survey (YRBS) measuring individual demo-
graphic characteristics, perceptions of safety, bullying victimization, and drug use were adminis-
tered to a sample of 6th through 12th grade students in a large, urban school district in
a Midwest city of approximately 600,000 residents. The data are cross-sectional and were
collected during the spring of 2009. In total, 23 schools in the district were included in the
study. Six of the schools were high schools (four of them exceeding 1000 students). The
remaining 17 were middle schools. The total number of completed and partially completed
surveys was 3,403. In total, 59% of the sample consisted of middle school students. The final
sample size after listwise deletion was 3,139: 1,515 (48%) males and 1,624 (52%) females. The
2010 United States Census reported demographic data for the entire city from which these
schools were drawn. Approximately 45% of the population described themselves as White,
approximately 40% as Black or African American, 17% as Latino/Hispanic, and about 3.5% as
two or more races. Approximately 52% of the city population is female. The sample population
is representative in terms of sex and over-represents the Black population. This is a function of
the selection by the school board of the initial 23 schools, which were drawn from the urban
areas of the city. Schools participating in the survey were part of a separate federally funded
 DEVIANT BEHAVIOR 5

project that invited schools in the region to participate in various school-related initiatives (e.g.,
parental skill development, restorative justice, classroom management, etc.), school administra-
tors self-nominated their school, then district project administrators of the federal program
selected schools. Researchers had no say in which schools were selected.1

Measures
Dependent variables
The dependent variables in this study are Soft drug use, Hard drug use, and Weapon carrying. We
replicate the items used in prior research to measure these constructs (Baker and Pelfrey 2016). The
latent construct of Soft drug use is based on three observed measures examining the participants’ use
of cigarettes, alcohol, and marijuana. For cigarette and alcohol use, participants were asked separate
similar questions: “During the past 30 days, on how many days did you smoke cigarettes?” and
“During the past 30 days, on how many days did you have at least one drink of alcohol?” 0 (0 day), 1
(1 or 2 days), 2 (3 to 5 days), 3 (6 to 9 days), 4 (10 to 19 days), 5 (20 to 29 days), and 6 (all 30 days).
For marijuana use, participants were asked “During the past 30 days, how many times did you use
marijuana?” 0 (0 time), 1 (1 or 2 times), 2 (3 to 9 times), 3 (10 to 19 times), 4 (20 to 39 times), and 5
(40 or more times). While these three items are used to construct a latent construct in the full
structural equation model, described in further detail below, it is worth noting that the items contain
acceptable internal reliability (α= 0.73).
Hard drug use is based on two items: “During your life, how many times have you taken
a prescription drug, such as OxyContin, Percocet, Vicodin, Ritalin, or Xanax without a doctor’s
prescription?” and “During the past 12 months, how many times did you use cocaine, crack,
methamphetamines, ecstasy or heroin?” 0 (0 time), 1 (1 or 2 times), 2 (3 to 9 times), 3 (10 to 19
times), 4 (20 to 39 times), 5 (40 or more times). The two items were combined into a mean-
standardized index (α = 0.60). Weapon carrying is measured using a single item, “During the past
30 days, on how many days did you carry a weapon such as a gun, knife, or club on school
property?” 0 (0 day), 1 (1 day), 2 (2 or 3 days), 3 (4 or 5 days), and 4 (6 or more days).

Experienced strain
Two measures of experienced strain are examined: Traditional bullying victimization and
Cyberbullying victimization. To measure Traditional bullying victimization, participants were
asked, “During the past 12 months, have you been harassed, picked on, or bullied so much that
you felt unsafe at school?” 0 (no) and 1 (yes). Cyberbullying victimization is operationalized as “How
often have you been the victim of cyberbullying in a chat room or social networking website such as
MySpace.com or Facebook.com?” 0 (never), 1 (a few times per year), 2 (a few times per month), 3
(a few times per week, each week), 4 (every day). Although these measures do not capture all aspects
of bullying victimization in terms of power imbalance, repetition, and intention to harm, they do
match the operationalization of nearly one-third of all studies recently analyzed in a systematic
review of bullying measures (Vivolo-Kantor et al. 2014) and replicated the YRBS and previous-
published operationalizations of bullying (Baker and Pelfrey 2016; Messias, Kindrick, and Castro
2014). They also closely resemble operationalizations of bullying from the National Longitudinal
Survey of Youth 1997 (Quinn and Stewart 2018).

1
While the selection process of the 23 schools was non-probabilistic the scope of this project exceeds numerous prior studies on
bullying that examined fewer schools which were also selected non-probabilistically (see for example, Cullen et al. 2008 – six
middle schools; Hay and Meldrum 2010 – one middle and one high school; Hay, Meldrum, and Mann 2010 – one middle and
one high school; Tharp-Taylor, Haviland, and D’Amico 2009 – two middle schools).
6 C. M. BRADY ET AL.

Anticipated strain
The measure of Anticipated strain pertained to skipping school because respondents felt it was
unsafe. It was measured by asking participants “During the past 30 days, on how many days did you
not go to school because you felt you would be unsafe at school or on your way to or from school?” 0
(0 day), 1 (1 day), 2 (2 or 3 days), 3 (4 or 5 days), 4 (6 or more days). This measure captures the
emotional fear of victimization felt by respondents and mirrors the operationalization of anticipated
strain by Baker and Pelfrey (2016).

Control variables
Two demographic characteristics were controlled for in every model. Black is a dichotomous
measure of self-identified race measured 0 (non-Black) and 1(Black). Grade level is the academic
grade level of each participant ranging from grades 6 through 12. Age is excluded because of its high
collinearity with grade level (r = 0.95).

Analytic strategy
To test the six proposed hypotheses, we conducted a series of structural equation models (SEM)
including a measurement model and a series of structural models all testing for male and female
between-group differences. An advantage to using structural equation modeling is that SEM
estimates the measurement error of a given latent construct and tests the entire model in the
same analysis, including mediating relationships. Typically, researchers have fit a series of
regressions to estimate mediating relationships; however, recently, researchers suggest that
SEM may be a more useful technique which can simultaneously and more efficiently estimate
mediation (Iacobucci 2010). Generally, researchers who use structural equation modeling agree
that model fit should be assessed by multiple fit indices (e.g. Hu and Bentler 1999; Iacobucci
2010; Steiger 2000). Model fit statistics are reported for each model and include: root mean
square error of approximation (RMSEA), standardized root mean square residual (SRMR),
comparative fit index (CFI), Tucker-Lewis index (TLI), chi-square test (χ2) with p value, and
coefficient of determination (CD).
As soft drug use is measured as a latent construct, a measurement model will be tested.
Invariance in the measurement model will first be established before testing for male/female
between-group differences. Next, to test for sex differences in the measurement of soft drug use,
a series of nested measurement models will be estimated simultaneously for males and females.
A group-based measurement model for soft drug use will be specified. If the model fits the data
well for both groups, then equivalent form can be established. Then, to test for invariant
loadings all parameters are freely estimated (unconstrained) across sex. Next, a model will be
estimated in which all factor loadings are constrained to be equal for females and for males. Chi-
Square (χ2) difference tests will be used to compare the fit of the unconstrained measurement
model with that of the constrained model. Loadings that demonstrate better model fit when
allowed to freely estimate indicate a significant difference across groups. If equivalent form and
loadings can be demonstrated across different groups, it indicates that the factors predicted by
the latent construct do not differ across groups and indicators of soft drug use are the same for
males as they are for females. If this is not the case, then we must understand and account for
these differences in measurement prior to conducting full structural equation models.
We also test for form and loading invariance between males and females across all structural
models analyzed. Invariance in form would again indicate that the specified models would not fit
significantly better for one group over the other. Testing for invariance in loadings provide
information on how certain variables (those which are not equivalent) differentially impact
outcomes for males and females. For example, if fit statistics indicated that a model allowing
the effect of cyberbullying on soft drug use to freely estimate across males and females fit better
than one in which it was constrained, then this would indicate a significant difference in the
 DEVIANT BEHAVIOR 7

impact of cyberbullying on soft drug use of males and females. The difference in size and
direction of the coefficient for males compared to females would demonstrate how the effect
differs between males and females in the sample.

Results
Soft drug use
As noted above, we begin our analysis of bullying victimization on soft drug use by estimating
a measurement model for the latent construct Soft drug use. Tests were conducted to ensure
measurement invariance across males and females in the latent construct. Results indicate an
equivalent form. All loadings are substantial and statistically significant. Because the model is
saturated (there are zero degrees of freedom), model fit statistics could not be estimated. To be
clear, the findings for measuring invariant form indicate that soft drug use has the same meaning for
males as it does for females. Next, we tested for invariant loadings. Results indicated none of the
observed variables is more central to the meaning of soft drug use for males over females and vice
versa. Specifically, the difference in loadings between males and females for each observed variable in
the measurement model was not significant (alcohol use: χ2 [1] = .007, p = .935; marijuana use: χ2 [1]
= 1.153, p = .283; cigarette use: χ2 [1] = 1.325, p = .250). We proceed with the full structural model
using the equal loadings measurement model for Soft drug use.
Our analysis continues by testing for invariant form and loadings in the full structural model
predicting soft drug use. The model statistics all provide evidence of good model fit,
χ2 (30) = 156.26, p = .00; RMSEA = .052; SRMR = .029; CFI = .945, TLI = .919; and
CD = .181. Given these model fit statistics, we proceed with examining the results of the SEM
(presented in Table 2 and visually in Figure 1). The findings indicate equivalent form: grade,
traditional bully victimization, cyberbully victimization, and anticipated strain are positively and
significantly associated with Soft drug use. Traditional bully victimization and cyberbully victimiza-
tion are positively and significantly associated with anticipated strain. Black was not significantly
associated with soft drug use for either males or females. These findings support our first, second,
and third hypotheses as they relate to soft drug use.
Tests for equal loadings indicate that several loadings significantly differ for males and females.
While the effect of cyberbully victimization on soft drug use was significant for both males (β = .35, p <
.001) and females (β = .11, p < .001), significant group differences exist. This indicates that the
association between cyberbully victimization and engaging in soft drug use is significantly greater
for males than for females. This finding provides partial support for our fourth hypothesis. Sex
differences also existed for the relationship between cyberbullying and anticipated strain.
Unexpectedly, the association was greater for males (β = .20, p < .001) compared to females (β = .06,

Table 1. Descriptive statistics.

Full Sample Males Females
(n = 3,139) (n = 1,515) (n = 1,624)
M SD M SD M SD
Black 0.64 0.48 0.63 0.48 0.66 0.48
Grade Level 3.21 1.70 3.13 1.69 3.28 1.71
Trad. Bully Victim 0.15 0.35 0.13 0.34 0.16 0.37
Cyberbully Victim 0.19 0.66 0.18 0.68 0.19 0.65
Anticipated Strain 0.45 0.99 0.42 0.96 0.47 1.01
Hard Drug Usea 0.00 0.85 0.06 1.02 −0.06 0.64
Soft Drug Useb 0.27 0.56 0.30 0.63 0.23 0.49
Weapon Carrying 0.14 0.62 0.17 0.70 0.10 0.54
a b
NOTES: = mean standardized index; = latent construct
8 C. M. BRADY ET AL.

Table 2. Summary table for multi-group results for soft drug use.
Unconstrained Solution Constrained Solution
Males Females Males Females
(n = 1,515) (n = 1,624) (n = 1,515) (n = 1,624)
β S.E. β S.E. β S.E. β S.E.
Anticipated Strain
Black −0.01 0.02 0.04 0.02 0.01 0.02 0.01 0.02
Grade −0.04 0.02 −0.00 0.03 −0.02 0.02 −0.02 0.02
Trad. Bully Victima 0.21*** 0.02 0.14*** 0.02 0.21*** 0.02 0.14*** 0.02
Cyberbully Victima 0.20*** 0.02 0.07** 0.02 0.20*** 0.02 0.06* 0.02
Soft Drug Use
Black 0.02 0.03 −0.01 0.03 −0.00 0.02 −0.00 0.02
Grade 0.15*** 0.02 0.20*** 0.03 0.15*** 0.02 0.20*** 0.02
Anticipated Straina 0.13*** 0.03 0.07* 0.03 0.14*** 0.03 0.06* 0.03
Trad. Bully Victim 0.03 0.03 0.07* 0.03 0.04* 0.02 0.06* 0.02
Cyberbully Victima 0.35*** 0.03 0.11*** 0.03 0.35*** 0.03 0.11*** 0.03
R2 Anticipated Strain 0.09 0.03 0.09 0.02
R2 Soft Drug Use 0.18 0.06 0.19 0.05
χ2 df = 25, 151.30, p = 0.00 df = 30, 156.26, p = 0.00
CFI 0.95 0.95
RMSEA 0.06 0.05
NOTES: aVariable was left unconstrained in constrained model because it significantly differed between males and females.
*p < .05, **p < .01, ***p < .001

Figure 1. Bullying victimization on soft drug use. Standardized coefficients shown. Constrained model estimated. (*p < .05; **p <
.01; ***p < .001).

p < .05). The effect of traditional bully victimization on anticipated strain was significant for both males
(β = .21, p < .001) and females (β = .14, p < .001), and again, the association was greater for males. The
association for both experienced strains on anticipated strain was greater for males compared to
females, this contradicts our fifth hypothesis. Group differences also occurred for anticipated strain,
where the association between anticipated strain and soft drug was greater for males (β = .14, p < .001)
than females (β = .06, p < .05).
 DEVIANT BEHAVIOR 9

In our tests for the mediating effect of anticipated strain, we found that traditional bullying had
a significant positive indirect effect on soft drug use through anticipated strain for both males (β =
.06, p < .001) and females (β = .01, p < .05). This relationship indicates that experiencing traditional
bullying leads to anticipated strain, which in turn leads to delinquent coping for males and females.
Cyberbullying had a significant positive indirect effect on soft drug use through anticipated strain for
males (β = .03, p < .001), but not for females. This relationship indicates that cyberbullying leads to
anticipated strain which in turn leads to delinquent coping, via soft drug use, for males. These
findings provide partial support for Hypothesis 6 as it relates to soft drug use, though we did not
hypothesize sex differences for the indirect effects.

Hard drug use

Our analysis of bullying victimization and hard drug use begins by ensuring that the model properly
fits the data. The model fit statistics all provides evidence of good model fit, χ2(5) = 8.14, p = .15;
RMSEA = .020; SRMR = .011; CFI = .994; TLI = .980; and CD = .182. Given these statistics of model
fit, we proceed with examining the results of the SEM (presented in Table 3 and visually in Figure 2).
For males, the results indicate that traditional bully victimization, cyberbully victimization, and
anticipated strain are positively and significantly associated with Hard drug use. For females,
traditional bullying and cyberbully victimization were positively and significantly associated with
Hard drug use. For both males and females, traditional bully victimization and cyberbully victimiza-
tion are positively and significantly associated with Anticipated Strain. These results, again, provide
support or partial support for our first three hypotheses in this case as they relate to hard drug use.
None of the demographic variables were significantly associated with hard drug use for either males
or females. We again find evidence for equivalent form in the model for males and females.
However, we do not find equal loadings. Several of the coefficients significantly differ for males
and females.
While the effect of cyberbully victimization on hard drug use was significant for both males (β =
.42, p < .001) and females (β = .06, p < .05), significant group differences exist. This indicates that the
association between cyberbully victimization and engaging in hard drug use is significantly greater

Table 3. Summary table for multi-group results for hard drug use.
Unconstrained Solution Constrained Solution
Males Females Males Females
(n = 1,515) (n = 1,624) (n = 1,515) (n = 1,624)
β S.E. β S.E. β S.E. β S.E.
Anticipated Strain
Black −0.01 0.02 0.04 0.02 0.01 0.02 0.01 0.02
Grade −0.04 0.02 −0.00 0.03 −0.02 0.02 −0.02 0.02
Trad. Bully Victima 0.21*** 0.02 0.14*** 0.02 0.21*** 0.02 0.14*** 0.02
Cyberbully Victima 0.20*** 0.02 0.07** 0.02 0.20*** 0.02 0.06* 0.02
Hard Drug Use
Black 0.00 0.02 −0.04 0.03 −0.02 0.01 −0.03 0.02
Grade 0.05* 0.02 0.02 0.03 0.02 0.01 0.03 0.02
Anticipated Straina 0.09*** 0.02 0.05 0.02 0.09*** 0.02 0.04 0.02
Trad. Bully Victim 0.07** 0.02 0.06* 0.03 0.04** 0.01 0.07** 0.02
Cyberbully Victima 0.42*** 0.02 0.06* 0.02 0.42*** 0.02 0.06* 0.02
R2 Anticipated Strain 0.09 0.03 0.09 0.02
R2 Hard Drug Use 0.21 0.01 0.21 0.01
χ2 df = 0, 0.00b df = 5, 8.14, p= .15
CFI 1.00 0.99
RMSEA 0.00 0.02
NOTES: aVariable was left unconstrained in constrained model because it significantly differed between males and females.
b
This model is saturated with no available degrees of freedom to estimate model fit. *p < .05, **p < .01, ***p < .001
10 C. M. BRADY ET AL.

Figure 2. Bullying victimization on hard drug use. Standardized coefficients shown. Constrained model estimated. (*p < .05; **p <
.01; ***p < .001).

for males than for females. This finding provides partial support for our fourth hypothesis.
Anticipated strain was significantly associated with hard drug use for males (β = .09, p < .001),
but there was no direct effect of anticipated strain on hard drug use for females. Sex differences also
existed for the relationship between cyberbullying and anticipated strain. Females (β = .06 p < .05)
who were victims of cyberbullying were significantly less likely to anticipate strain than males (β =
.20, p < .001). Sex differences also existed for the relationship between traditional bullying and
anticipated strain. Unexpectedly, the association was greater for males (β = .21, p < .001) compared
to females (β = .14, p < .001), again contradicting our fifth hypothesis.
The mediating relationships through anticipated strain were again evaluated. Cyberbullying (β =
.03, p < .001) and traditional bullying (β = .06, p < .001) have significant positive indirect effects on
hard drug use through anticipated strain for males, but have no indirect effects on hard drug use for
females. This indicates that bullying leads to anticipated strain which in turn leads to delinquent
coping, in the form of hard drug use, for males again partially supporting our sixth hypothesis.

Weapon carrying at school

As with hard drug use, we begin our analysis of the structural model examining the impact of
bullying on weapon carrying at schools by ensuring that the model properly fits the data. The model
statistics all provide evidence of good model fit, χ2(6) = 6.52, p = .37;
RMSEA = .007; SRMR = .010; CFI = .998; TLI = .995; and CD = .097. Given these statistics of
model fit, we proceed with examining the results of the SEM (presented in Table 4 and visually in
Figure 3). The findings indicate that grade, traditional bully victimization, cyberbully victimization,
and anticipated strain are positively and significantly associated with Weapon carrying. For both
males and females, traditional bully victimization and cyberbully victimization are positively and
 DEVIANT BEHAVIOR 11

Table 4. Summary table for multi-group results for weapon carrying.

Unconstrained Solution Constrained Solution
Males Females Males Females
(n = 1,515) (n = 1,624) (n = 1,515) (n = 1,624)
β S.E. β S.E. β S.E. β S.E.
Anticipated Strain
Black −0.01 0.02 0.04 0.02 0.01 0.02 0.01 0.02
Grade −0.04 0.02 −0.00 0.03 −0.02 0.02 −0.02 0.02
Trad. Bully Victima 0.21*** 0.02 0.14*** 0.02 0.21*** 0.02 0.14*** 0.02
Cyberbully Victima 0.20*** 0.02 0.07** 0.02 0.20*** 0.02 0.06* 0.02
Weapon Carrying
Black 0.03 0.03 0.02 0.02 0.02 0.02 0.03 0.02
Grade 0.05 0.03 0.03 0.03 0.03* 0.02 0.04* 0.02
Anticipated Strain 0.09*** 0.03 0.06* 0.02 0.06*** 0.02 0.08*** 0.02
Trad. Bully Victim 0.06* 0.03 0.08** 0.02 0.06*** 0.01 0.08*** 0.02
Cyberbully Victima 0.21*** 0.02 0.09*** 0.02 0.21*** 0.02 0.08** 0.02
R2 Anticipated Strain 0.09 0.03 0.09 0.02
R2 Weapon Carrying 0.07 0.02 0.06 0.03
χ2 df = 0, 0.00 b
df = 6, 6.52, p = .37
CFI 1.00 1.00
RMSEA 0.00 0.01
NOTES: aVariable was left unconstrained in constrained model because it significantly differed between males and females.
b
This model is saturated with no available degrees of freedom to estimate model fit. *p < .05, **p < .01, ***p < .001

significantly associated with Anticipated strain. Once again, we find support for Hypotheses 1, 2, and
3. Black was not significantly associated with weapon carrying for either males or females. While we
again find evidence of invariant form in the structural models for males and females, we also find
significant differences in several of the coefficients.
Although cyberbullying victimization was significantly associated with weapon carrying for both
males (β = .21, p < .001) and females (β = .08, p < .01), differences in effect size are significant, males are
more likely to carry a weapon if a victim of cyberbullying again partially supporting Hypothesis 4. Sex
differences existed for the relationship between cyberbullying and anticipated strain. Once again,
males (β = .20, p < .001) who were victims of cyberbullying were more likely to anticipate strain than
females (β = .06, p < .05) failing to support our fifth hypothesis. Sex differences also existed for the
relationship between traditional bullying and anticipated strain. The association was greater for males
(β = .21, p < .001) compared to females (β = .14, p < .001) contradicting our fifth hypothesis.
Tests for indirect effects show cyberbullying victimization has a significant positive indirect effect
on carrying a weapon through anticipated strain for both males (β = .01, p < .001) and females (β =
.00, p < .05). Also, traditional bullying victimization has a significant positive indirect effect on
weapon carrying through anticipated strain for both males (β = .03, p < .001) and females (β = .02,
p < .01). This relationship indicates that experiencing cyberbullying and traditional bullying leads to
anticipated strain. Anticipated strain, in turn, leads to delinquent coping, via weapon carrying, for
both males and females.

Discussion and conclusion

The primary focus of this study was to examine the impact of bullying on delinquent coping among
males and females. We tested six hypotheses and analyses revealed several interesting results that
inform future research and policy. Across the three different types of delinquent coping behaviors
examined–soft drug use, hard drug use, and weapon carrying–each model was better at predicting
male delinquent coping than female delinquent coping. Our findings reinforce those of previous
studies in demonstrating that males and females respond and cope differently after experiencing
strain (Broidy and Agnew 1997). However, contrary to our proposed hypothesis regarding differ-
ences in emotional responses among females, the impact of bullying on anticipated strain was
stronger for males compared to females.
12 C. M. BRADY ET AL.

Figure 3. bullying victimization on weapon carrying. Standardized coefficients shown. Constrained model estimated. (*p < .05; **p
< .01; ***p < .001).

The findings from the current study also advance the literature on general strain theory, bullying
victimization, and differential delinquent coping in several ways. The finding of males experiencing
a larger effect of anticipated strain on delinquent coping suggests the possibility that males may have
greater anxiety about potentially strainful events which increases their likelihood of using drugs to
cope and/or arming themselves with weapons. The effect of both anticipated and experienced strain
were directly evident among the sample of adolescents, however, sex differences did exist. The
experienced strain was significantly associated with anticipated strain, indicating that victimization
may intensify concern over future events. Another interesting pattern emerged from the data.
Although not hypothesized, we found that the effect of traditional bullying on anticipated strain
was significantly greater than the effect of traditional bullying on delinquency for both males and
females. However, this pattern reverses for cyberbullying. In all instances, the effect of cyberbullying
on delinquency is significantly greater than or statistically equivalent to the effect of cyberbullying on
anticipated strain.
Theoretically, this implies that different strains may result in different types of coping
behavior even if those strains seem conceptually similar. Both traditional bullying and cyber-
bullying represent peer abuse and are often treated as similar behaviors. Our results demon-
strate that traditional bullying and cyberbullying may be more distinct than simply the method
of victimization. Of course, it is possible that this result is an artifact of our operationalization
of anticipated strain. Because anticipated strain analyzed in the current study involved fear
related to attending school – the location in which the traditional bullying was most likely to
occur, fear may have been more likely to result from traditional bullying in the current analysis.
However, the effect sizes of traditional bullying and cyberbullying on anticipated strain are
similar (.21 compared to .20 for males and .14 compared to .06 for females, respectively, across
all models). So, it is also possible that the direct, interpersonal, often physical nature of in-
person bullying may produce different coping behaviors than the less direct, even at times
 DEVIANT BEHAVIOR 13

anonymous nature of cyberbullying. Again, when we compare the effect sizes of traditional
bullying and cyberbullying on delinquency, cyberbullying generally has a larger or equivalent
effect compared to traditional bullying. In sum, traditional bullying and cyberbullying appear to
have similar effects on more emotional reactions, but cyberbullying seems to have a greater
effect on behavioral coping. Still, both bullying behaviors have a greater effect on the coping
behaviors of males compared to females.
Future research should consider different types of coping responses when testing GST, as
males and females appear to have different responses to strains. Many studies utilizing the GST
framework, do not include both fear and behavioral coping to strain focusing instead solely on
offending behavior (Aseltine, Gore, and Gordon 2000; Ousey, Wilcox, and Schreck 2015;
Paternoster and Mazerolle 1994). Additionally, future research should include negative affective
states in their analytic models. In addition to identifying certain types of strain that can lead to
deviant coping, Agnew (1992) posited that strain produces deviance indirectly through negative
affective states. Originally, Agnew (1992) discussed anger as a key negative affective state, but
GST has been expanded to include other negative emotions such as frustration (Agnew 2001,
2006). Including all these components would result in a more robust specification and test
of GST.
Like most prior research, this study is not without limitations. The current study uses cross-
sectional data preventing direct tests of causality. The sample, while large, was drawn from
a single city in the Midwest and participants were not selected at random. As such, the
generalization of these findings should be done cautiously. In addition, because respondents
ranged from 6th graders to 12th graders, questions were chosen that could be easily understood
by the very youngest and inexperienced readers in the sample. Future studies should consider
other more sophisticated measures which capture all the various components of bullying, fear,
drug use, and aggression.
The current study, while adding to the scholarly literature, also informs policy. First, bullying
impacts the criminal justice system in various ways but perhaps most appreciably through the
School to Prison Pipeline (Nellis 2015). When problems happen at schools, law enforcement
personnel such as school resource officers are often called to intervene (Nellis 2015). Arrests
become likely, particularly when a weapon or drugs are involved, and the future educational
achievement of students is jeopardized (Pigott, Stearns, and Khey 2018). As Pigott, Stearns, and
Khey (2018) note, zero-tolerance policies in the areas of drugs and weapons are at the root of the
school to prison pipeline. To the extent that bullying victimization leads to drug use and weapon
carrying at schools, administrators and school resource officers must be aware and implement
policies to reduce bullying and understand behavioral and emotional reactions to it. Second,
males and females may have different responses to strainful experiences, which suggests a need
to consider different prevention and intervention strategies. The attitudinal responses should be
carefully monitored as emotional responses may be more internalized and easier to miss
(Gresham and Kern 2004; Stanger and Lewis 1993). Within a school setting, the identification
and understanding of students exhibiting these behaviors is an important responsibility of
teachers and other school professionals. Particularly, understanding how these responses man-
ifest differently in males and females is necessary so that they can be identified and early
interventions put in place.
Intensive school-based anti-bullying programs that incorporate parental meetings and increased
supervision have shown high levels of effectiveness (Ttofi and Farrington 2011). Prevention of
bullying and cyberbullying may reduce the burden on the justice system and enhance the educational
experience of students. Based on the findings from the current study, both traditional bullying and
cyberbullying have an impact on delinquent behavior. Schools should be aware of both forms of
bullying and, when discussing the adverse outcomes associated with bullying, be sure to discuss both
safety and the online and in-person environment.
14 C. M. BRADY ET AL.

Notes on contributors
Caitlin M. Brady is a doctoral candidate in the Department of Criminal Justice at the University of Central Florida.
She received her Master’s in Forensic Psychology from John Jay College of Criminal Justice. Her research interests
include: juvenile justice, mental health, gender differences, and risk assessment.

Thomas Baker is an associate professor in the Department of Criminal Justice at the University of Central Florida. His
research interests include perceptions of crime and the criminal justice system, institutional corrections, and life-
course criminology.

William V. Pelfrey Jr. is a Professor in the Criminal Justice Program and Chair of the Homeland Security/Emergency
Preparedness Program in the Wilder School of Government at Virginia Commonwealth University. He received his
PhD in criminal justice from Temple University and his MS in clinical psychology from Radford University. His
primary research areas are: homeland security, police use of force, cyber bullying, and the psychology of the offender.

References
Agnew, Robert. 1992. “Foundation for a General Strain Theory of Crime and Delinquency.” Criminology 30 (1):47–88.
Agnew, Robert. 2001. “Building on the Foundation of General Strain Theory: Specifying the Types of Strain Most
Likely to Lead to Crime and Delinquency.” Journal of Research in Crime and Delinquency 38 (4):319–61.
Agnew, Robert. 2002. “Experienced, Vicarious, and Anticipated Strain: An Exploratory Study on Physical
Victimization and Delinquency.” Justice Quarterly 19 (4):603–32.
Agnew, Robert. 2006. “General Strain Theory: Current Status and Directions for Further Research.” Pp. 101–23 in
Advances in Criminological Theory: Vol. 15. Taking Stock: The Status of Criminological Theory, edited by
F. T. Cullen, J. P. Wright, and K. R. Blevins. Piscataway, NJ, US: Transaction Publishers,
Agnew, Robert, Timothy Brezina, John Paul Wright, and Francis T. Cullen. 2002. “Strain, Personality Traits, and
Delinquency: Extending General Strain Theory.” Criminology 40 (1):43–72.
Aseltine, Robert H., Jr, Susan Gore, and Jennifer Gordon. 2000. “Life Stress, Anger and Anxiety, and Delinquency: An
Empirical Test of General Strain Theory.” Journal of Health and Social Behavior 41 (3):256–75.
Baker, Thomas and William V. Pelfrey Jr. 2016. “Bullying Victimization, Social Network Usage, and Delinquent
Coping in a Sample of Urban Youth: Examining the Predictions of General Strain Theory.” Violence and Victims 31
(6):1021–43.
Baron, Stephen W. 2009. “Street Youths’ Violent Responses to Violent Personal, Vicarious, and Anticipated Strain.”
Journal of Criminal Justice 37 (5):442–51.
Broidy, Lisa and Robert Agnew. 1997. “Gender and Crime: A General Strain Theory Perspective.” Journal of Research
in Crime and Delinquency 34 (3):275–306.
Cullen, Francis T., James D. Unnever, Jennifer L. Hartman, Michael G. Turner, and Robert Agnew. 2008. “Gender, Bullying
Victimization, and Juvenile Delinquency: A Test of General Strain Theory.” Victims and Offenders 3 (4):346–64.
Gresham, Frank M. and Lee Kern. 2004. “Internalizing Behavior Problems in Children and Adolescents.” Pp. 262–81
in Handbook of Research in Emotional and Behavioral Disorders, edited by R. B. Rutherford, M. M. Quinn, and
S. R. Mathur. New York, NY, US: Guilford Press,
Hay, Carter and Ryan Meldrum. 2010. “Bullying Victimization and Adolescent Self-Harm: Testing Hypotheses from
General Strain Theory.” Journal of Youth and Adolescence 39 (5):446–59.
Hay, Carter, Ryan Meldrum, and Karen Mann. 2010. “Traditional Bullying, Cyber Bullying, and Deviance: A General
Strain Theory Approach.” Journal of Contemporary Criminal Justice 26 (2):130–47.
Higgins, George E., Nicole L. Piquero, and Alex R. Piquero. 2011. “General Strain Theory, Peer Rejection, and
Delinquency/Crime.” Youth & Society 43 (4):1272–97.
Hinduja, Sameer and Justin W. Patchin. 2007. “Offline Consequences of Online Victimization: School Violence and
Delinquency.” Journal of School Violence 6 (3):89–112.
Hinduja, Sameer and Justin W. Patchin. 2008. “Cyberbullying: An Exploratory Analysis of Factors Related to
Offending and Victimization.” Deviant Behavior 29 (2):129–56.
Hu, Li-tze and Peter M. Bentler. 1999. “Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional
Criteria versus New Alternatives.” Structural Equation Modeling: A Multidisciplinary Journal 6 (1):1–55.
Iacobucci, Dawn. 2010. “Structural Equations Modeling: Fit Indices, Sample Size, and Advanced Topics.” Journal of
Consumer Psychology 20 (1):90–98.
Jennings, Wesley G., Nicole L. Piquero, Angela R. Gover, and Deanna M. Perez. 2009. “Gender and General Strain
Theory: A Replication and Exploration of Broidy and Agnew’s Gender/Strain Hypothesis among A Sample of
Southwestern Mexican American Adolescents.” Journal of Criminal Justice 37 (4):404–17.
Kaufman, Joanne M. 2009. “Gendered Responses to Serious Strain: The Argument for a General Strain Theory of
Deviance.” Justice Quarterly 26 (3):410–44.
 DEVIANT BEHAVIOR 15

Keith, Shelley. 2018. “How Do Traditional Bullying and Cyberbullying Victimization Affect Fear and Coping among
Students? an Application of General Strain Theory.” American Journal of Criminal Justice 43 (1):67–84.
Lee, Yeungjeom and Jihoon Kim. 2018. “Examining the Gendered Effect of Experienced and Vicarious Victimization:
A General Strain Theory Perspective.” American Journal of Criminal Justice 43 (2):181–96.
Mazerolle, Paul and Jeff Maahs. 2000. “General Strain and Delinquency: An Alternative Examination of Conditioning
Influences.” Justice Quarterly 17 (4):753–78.
Mazerolle, Paul and Alex Piquero. 1997. “Violent Responses to Strain: An Examination of Conditioning Influences.”
Violence and Victims 12 (4):323–43.
McGrath, Shelly A., Catherine D. Marcum, and Heith Copes. 2012. “The Effects of Experienced, Vicarious, and
Anticipated Strain on Violence and Drug Use among Inmates.” American Journal of Criminal Justice 37 (1):60–75.
Messias, Erick, Kristi Kindrick, and Juan Castro. 2014. “School Bullying, Cyberbullying, or Both: Correlates of Teen
Suicidality in the 2011 CDC Youth Risk Behavior Survey.” Comprehensive Psychiatry 55 (5):1063–68.
Moon, Byongook, Merry Morash, and John D. McCluskey. 2012. “General Strain Theory and School Bullying: An
Empirical Test in South Korea.” Crime & Delinquency 58 (6):827–55.
Nellis, Ashley. 2015. A Return to Justice: Rethinking Our Approach to Juveniles in the System. Lanham, MD: Rowman &
Littlefield.
Ousey, Graham C., Pamela Wilcox, and Christopher J. Schreck. 2015. “Violent Victimization, Confluence of Risks and
the Nature of Criminal Behavior: Testing Main and Interactive Effects from Agnew’s Extension of General Strain
Theory.” Journal of Criminal Justice 43 (2):164–73.
Patchin, Justin W. and Sameer Hinduja. 2011. “Traditional and Nontraditional Bullying among Youth: A Test of
General Strain Theory.” Youth & Society 43 (2):727–51.
Paternoster, Raymond and Paul Mazerolle. 1994. “General Strain Theory and Delinquency: A Replication and
Extension.” Journal of Research in Crime and Delinquency 31 (3):235–63.
Peck, Jennifer H., Kristina K. Childs, Wesley G. Jennings, and Caitlin M. Brady. 2018. “General Strain Theory,
Depression, and Substance Use: Results from a Nationally Representative, Longitudinal Sample of White,
African-American, and Hispanic Adolescents and Young Adults.” Journal of Child & Adolescent Substance Abuse
27 (1):11–28.
Pew Research Center. 2017. “Mobile Fact Sheet.” Retrieved May, 5 2019. www.pewinternet.org/fact-sheet/mobile/.
Pigott, Christina, Ami E. Stearns, and David N. Khey. 2018. “School Resource Officers and the School to Prison
Pipeline: Discovering Trends of Expulsions in Public Schools.” American Journal of Criminal Justice 43 (1):120–38.
Piquero, Nicole Leeper and Miriam D. Sealock. 2004. “Gender and General Strain Theory: A Preliminary Test of
Broidy and Agnew’s Gender/GST Hypotheses.” Justice Quarterly 21 (1):125–58.
Quinn, Susan T. and Megan C. Stewart. 2018. “Examining the Long-Term Consequences of Bullying on Adult
Substance Use.” American Journal of Criminal Justice 43 (1):85–101.
Stanger, Catherine and Michael Lewis. 1993. “Agreement among Parents, Teachers, and Children on Internalizing and
Externalizing Behavior Problems.” Journal of Clinical Child Psychology 22 (1):107–16.
Steiger, James H. 2000. “Point Estimation, Hypothesis Testing, and Interval Estimation Using the RMSEA: Some
Comments and a Reply to Hayduk and Glaser.” Structural Equation Modeling 7 (2):149–62.
Stubbs-Richardson, Megan, H. Colleen Sinclair, Rebecca M. Goldberg, Chelsea N. Ellithorpe, and Suzanne C. Amadi.
2018. “Reaching Out versus Lashing Out: Examining Gender Differences in Experiences with and Responses to
Bullying in High School.” American Journal of Criminal Justice 43 (1):39–66.
Tharp-Taylor, Shannah, Amelia Haviland, and Elizabeth J. D’Amico. 2009. “Victimization from Mental and Physical
Bullying and Substance Use in Early Adolescence.” Addictive Behaviors 34 (6–7):561–67.
Ttofi, Maria M. and David P. Farrington. 2011. “Effectiveness of School-Based Programs to Reduce Bullying:
A Systematic and Meta-Analytic Review.” Journal of Experimental Criminology 7 (1):27–56.
Vivolo-Kantor, Alana M., Brandi N. Martell, Kristin M. Holland, and Ruth Westby. 2014. “A Systematic Review and
Content Analysis of Bullying and Cyber-Bullying Measurement Strategies.” Aggression and Violent Behavior 19
(4):423–34.

View publication stats