Children and Youth Services Review 129 (2021) 106178

Contents lists available at ScienceDirect

Children and Youth Services Review

journal homepage: www.elsevier.com/locate/childyouth

Bullying patterns and their associations with child maltreatment and

adolescent psychosocial problems
Dalhee Yoon a, *, Stacey L. Shipe a, b, Jiho Park c, Miyoung Yoon d
a
Department of Social Work, Binghamton University-State University of New York, USA
b
Child Maltreatment Solutions Network, Pennsylvania State University, USA
c
Jack, Joseph and Morton Mandel School of Applied Social Sciences, Case Western Reserve University, USA
d
School of Social Work, University of Michigan, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Introduction: This study aimed 1) to identify underlying heterogeneous patterns of bully-victim; 2) to examine
Bullying whether the different types of child maltreatment predict the patterns of bully-victim; and 3) to investigate the
Peer victimization association between patterns of bully-victim and adolescent psychosocial problems (depression, trouble at
Peer aggression
school, and substance use).
Child maltreatment
Adolescent psychosocial problems
Methods: This study included a sample of 1139 (48.7% girls, 53.4% Black) drawn from the Fragile Families and
Adolescence Child Wellbeing Study. Children’s self-reported bullying victimization at age 9 was used using the Panel Study of
Income Dynamics Child Development Supplement III. Teacher’s reported bullying perpetration at age 9 was used
using Social Skills Rating System. Child maltreatment types were assessed at age 5 using the Parent-Child Conflict
Tactics Scale Coding. At age 15, adolescent depression was measured using modified Center for Epidemiologic
Studies Depression Scale; trouble at school was measured using modified Add Health In-School Questionnaire;
and self-reported substance use was used.
Results: Latent class analysis produced four classes: bully-victim (19.8%), victim (16.3%), no bully-victim
(38.9%), and bully (24.9%). Individuals who have been neglected are more likely to be in the victim class
compared to all other classes. Physical abuse to be at heightened risk of involvement in the bully-victim,
compared to victim class. Additionally, individuals in the victim group are greater risk for depression, prob­
lems at school, and alcohol, as compared to those in the other classes.
Conclusions: This study augments the knowledge base on bully/victim, child maltreatment, and behavioral health
outcomes and elucidates several suggestions for research and policy.

1. Introduction sexual violence perpetration (Espelage, Basile, & Hamburger, 2012;

Farrington, Loeber, Stallings, & Ttofi, 2011; Hemphill et al., 2011),
According to a recent nationally representative survey, the rate of along with psychological distress (Klomek, Sourander, & Gould, 2010).
bullying victimization was 23% and the rate of bullying perpetration Given the deleterious consequences of bullying, prior research has
was 6% among U.S. children and adolescents aged 6–17 years (Lebrun- strived to understand the patterns of bullying and their consequences,
Harris, Sherman, Limber, Miller, & Edgerton, 2019). Bullying—either but few studies have examined them within the context of child
victimization or perpetration experiences in childhood—has been maltreatment. This is specifically a glaring gap in the research because
recognized as a trigger to developing maladaptive behaviors. Bullying bullying has been identified as a common outcome of child maltreat­
victimization had detrimental effects on psychological health in later ment (Hong, Espelage, Grogan-Kaylor, & Allen-Meares, 2012; Wolfe,
life, including depression and suicidality (Barzilay et al., 2017; Taki­ Crooks, Chiodo, & Jaffe, 2009). Understanding of child maltreatment
zawa, Maughan, & Arseneault, 2014; Ttofi, Ttofi, Farrington, Lösel, & and its association with bullying patterns would offer meaningful
Loeber, 2011). Similarly, bullying perpetration reported behavioral guidelines for practice by suggesting specific points for interventions
health issues, such as subsequent violent behavior, binge drinking, and and treatments.

* Corresponding author at: Department of Social Work, Binghamton University-State University of New York, P.O. Box 6000, Binghamton, NY 13902-6000, USA.
E-mail address: dyoon@binghamton.edu (D. Yoon).

https://doi.org/10.1016/j.childyouth.2021.106178
Received 21 December 2020; Received in revised form 11 July 2021; Accepted 20 July 2021
Available online 23 July 2021
0190-7409/© 2021 Elsevier Ltd. All rights reserved.
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

1.1. Patterns of bullying between traumatic experiences (e.g., child maltreatment) and bullying.
Child maltreatment experiences have frequently identified as a strong
Bullying refers to intentional, repetitive, and aggressive behaviors risk factor for problematic peer relationships among at-risk of
among school aged children in which a power imbalance exists between maltreatment populations (Alto, Handley, Rogosch, Cicchetti, & Toth,
victims and perpetrators (Olweus, 1994; Rettew & Pawlowski, 2016). 2018; O’Hara, 2020; Wang et al., 2020; Yoon, 2020). Thus, more
Individuals either try to obtain or are deprived of power in peer re­ research on bullying patterns in middle childhood within the context of
lationships determine the role in bullying such as perpetrator, victim, or child maltreatment is needed.
both. Social capital theory helps to explain diverse patterns of bullying
by addressing different functioning and dynamics of social groups due to 1.2. Child maltreatment types and bullying
interpersonal relationships (Evans & Smokowski, 2015). According to
the social capital theory, victims have minimal social capital, which The developmental psychopathology perspective explains a diverse
hinders them to gain high social status, whereas bullies acquire and range of functioning throughout developmental periods and emphasizes
expand their social capital by relegating peers with low social status. the consequences of early adverse experiences (Cicchetti & Rogosch,
Due to the nature of social capital that could be wavering, individuals 1996; Toth & Cicchetti, 2013). In line with the perspective, several
can be involved in both bully and victim experiences. Those who have empirical studies have identified that child maltreatment is associated
both experiences tend to find weaker peers and mimic the perpetrators’ with bullying involvement in both victimization (Bowes et al., 2009;
behaviors in order to obtain social capital and end their victimization. O’Hara, 2020; Yoon, Yoon, Park, & Yoon, 2018) and perpetration
Informed by social capital theory, previous bullying studies have (Bowes et al., 2009; Hsieh et al., 2020; Wang et al., 2020). Although
demonstrated children and adolescents’ bullying experiences using the previous research has reported the strong association between all types
variable-centered approach. However, the variable-centered approach of child maltreatment and bullying, findings have been mixed across
does not capture the variability within and across subgroups of bullying studies depending on either type of child maltreatment or the form of
(Giang & Graham, 2008; Turns & Sibley, 2018). This, in turn, obscures bullying (Cunningham et al., 2019; Lo et al., 2021). For example,
an understanding of the complexity of bullying patterns. In addition, Nicholson, Chen, and Huang (2018) examined the effects of physical
prior research using the variable-centered approach was limited because abuse and neglect on bullying victimization, and showed that physically
these studies used arbitrary cutoff scores, such as mean, median, or abused children were more likely to be bullied by their peers, whereas
standard deviation (Jansen et al., 2012; Solberg & Olweus, 2003). A neglected children were not associated with bullying victimization.
person-centered approach such as latent class analysis (LCA), on the However, Yoon et al. (2018) found no relationships between physical
other hand, is an effective method to understand heterogeneity in abuse and bullying victimization. In regard to perpetration experiences,
bullying involvement experiences (etterncourt, Farrell, Liu, and Sulli­ Duke and colleagues (2010) examined the effects of adverse childhood
van, 2013). experiences—including physical and sexual abuse experiences—on
Building upon the strength of person-centered approach, a growing violence perpetration outcomes including bullying. They identified that
body of research has used the person-centered approach to identify both physical and sexual abuse experiences were at heightened risk for
distinct patterns and subgroups of bullying behaviors. Betterncourt et al. becoming bullying perpetrators. Similarly, Hsieh et al. (2020) examined
(2013) and Lovegrove, Henry, and Slater (2012) examined the patterns the effects of diverse child maltreatment (i.e., physical neglect, psy­
of bullying using perpetration and victimization indicators and found chological neglect, physical abuse, psychological abuse) on perpetration
four distinct patterns: victim, bully, bully-victim, and non-involved. experiences, and found all types of abuse were associated with bullying
Both the victim and non-involved patterns had low levels of aggres­ perpetration experiences. Although previous studies have supported the
sion and perpetration experiences, although the victim pattern had high influences of child maltreatment on bullying either victimization and
levels of victimization experiences. The other two groups, bully and perpetration, most studies have lumped the types of child maltreatment.
bully-victim patterns, had high levels of perpetration and aggression, Additionally, the studies that have specified the types of maltreatment
while only the bully pattern reported low levels of victim experiences. were limited in that it included selective maltreatment types and
Similarly, Giang and Graham (2008) examined the bullying subgroups bullying experiences.
and identified more specific patterns: five classes of bullying involve­
ment (i.e., socially adjusted, victims, aggressors, highly-victimized 1.3. Bullying and adolescent psychosocial problems
aggressive-victims, and highly-aggressive aggressive victims). Howev­
er, Goldweber, Waasdorp, and Bradshaw (2013) identified three pat­ Substance use, trouble at school, and depressive symptoms are
terns: low involvement, victim, and bully-victim. Although these studies frequently associated with bullying. According to Tomczyk and col­
have significant implications by describing heterogeneous patterns of leagues (2015), youth who have committed bullying as perpetrators in
bullying adapting both victimization and perpetration, their samples middle childhood were more likely to use diverse types of substances in
were focused on general sample in adolescence. adolescence, whereas victimization experiences in middle childhood
Bullying is most common in elementary school-aged children (Jan­ were not associated with adolescent substance use. Similarly, Gaete
sen et al., 2012). In order to develop interventions that target these age et al. (2017) identified that victimization was not associated with
groups and mitigate negative developmental problems, a better under­ marijuana use, although all the other bullying experiences—regardless
standing of bullying in middle childhood is necessary. Only two of victim or perpetration—were positively associated with alcohol and
empirical studies, to our knowledge, have investigated the bullying tobacco use. As for trouble at school, all types of bullying involvement
patterns in middle childhood. Williford and colleagues (2011) explored were identified as risk factors for suspension or expulsion, feeling of not
the bullying patterns at grade 4th, 5th, and 6th and the bullying tran­ belonging at school, and to endorse cheating (Glew, Fan, Katon, Rivara,
sition from late elementary school to middle school. They revealed the 4 & Kernic, 2005). On the other hand, Stein and colleagues (2007)
patterns (i.e., aggressor, victim, aggressor-victim, and uninvolved). demonstrated that youth identified as bully-victim had no significant
Using 3th, 4th, and 5th grade students, Moses and Williford (2017) also differences with school problems as compared to those identified as
examined the bullying patterns. However, they failed to disentangle the bully. But these same youth had higher rates of school problems than
patterns of bullying by taking into account only victimization separately youth who were identified as victim and non-involved. In regard to
despite bullying victimization and perpetration are not mutually depressive symptoms, previous studies have commonly reported that
exclusive, involving both forms of bullying simultaneously (Solberg & bully-victim and victims were more likely to have higher levels of
Olweus, 2003). In addition, the study samples for both were public depressive symptoms (Lovegrove & Cornell, 2014; O’Brennan, Brad­
school students in urban setting and did not analyze the relationship shaw, & Sawyer, 2009; Özdemir & Stattin, 2011; Seeds, Harkness, &

2
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

Quilty, 2010). Table 1

Descriptive Statistics of Study Variables (N = 1139).
1.4. The current study M (SD) / %

Gender (Female) 48.7

Previous studies have shown various combinations of bullying Race (Black) 53.4
perpetration and victimization using general samples, however, scarce Race (White) 31.6
research has established whether the patterns of bullying differ by types Caregivers’ education (More than HS) 72.6
of child maltreatment. In addition, the existing literature has produced Bully
1. Fights with others 45.0
inconsistent findings on the bullying pattern and its association with 2. Threatens or bullies others 30.6
child maltreatment types and adolescent psychosocial problems, 3. Argues with others 60.8
including substance use, problems at schools, and depressive symptoms. Victim
Thus, this study aimed: 1) to identify underlying heterogeneous patterns 1. Kids picked on you or said mean things to you 54.3
2. Kids hit you 21.9
of bullying in a high-risk sample; 2) to examine whether the different
3. Kids take your things 13.1
types of child maltreatment predict the patterns of bullying; and 3) to 4. Kids purposely left you out of activities 31.7
investigate the association between patterns of bullying and adolescent Child maltreatment
psychosocial problems. Given that LCA is a data-driven exploratory Physical assault 80.2
approach, we did not hypothesize the number of bullying patterns. As Psychological aggression 94.0
Neglect 11.5
for the second aim, we have hypothesized that each child maltreatment Adolescent behavioral health
type would be associated differently with patterns of bullying, guided by Alcohol 14.0
developmental psychopathology perspective. Based on the extant liter­ Marijuana 17.0
ature, we broadly hypothesized the association between bullying pat­ Cigarette 4.1
Depressive symptoms 8.03 (3.09)
terns and adolescent psychosocial problems: Individuals who have
Trouble at schools 7.24 (1.93)
experienced severe bullying would be associated with greater risk of
adolescent psychosocial problems. Note. M = mean; SD = standard deviation; HS = high school.

2. Method For each question, the response options were a five-point Likert scale,
ranging from 0 = not once to 4 = every day. All these items were
2.1. Sample dichotomized: 0 = not at all and 1 = at least one experience, given the
extreme distribution.
Data were drawn from the Fragile Families and Child Wellbeing
Study (FFCWS), which is a longitudinal study of families from 20 U.S. 3.2.2. Child maltreatment
cities to investigate the context and consequences of fragile families At age 5, primary caregivers were asked about their acts of physical
(Reichman, Teitler, Garfinkel, & McLanahan, 2001). Using a stratified assault, psychological aggression, and neglect in the past year using a
and multistage random sampling procedures, data were collected at modified version of the Parent-Child Conflict Tactics Scales Coding
hospitals between 1998 and 2000 at the time of the birth of a focal child. (Straus, Hamby, Finkelhor, Moore, & Runyan, 1998). Physical assault
The FFCWS over-sampled non-martial births, resulting in a relatively consisted of 5 items of physical punishment behaviors (e.g., “spanked
disadvantaged sample of families. Follow-up surveys were conducted child on the bottom with your bare hand,” “slapped child on hand, arm, or
from children, caregivers, and teachers when the focal child was 1, 3, 5, leg”). Psychological aggression consisted of 5 items of emotional painful
7, and 15 years by phone with select in-home assessments. For this verbal or symbolic acts towards child (e.g., “ever swore or cursed at child,”
study, we included a sample of 1139 children, their parents/caregiver, “shouted, yelled, or screamed at child”). Neglect consisted of 5 items of
and teachers who completed the main study variables across all three inadequate providing food or supervision (e.g., “left child home alone,”
time points at ages 5, 9, and 15 (48.7% female, 53.4% Black, 31.6% “was not able to make sure child got the food”). For each type of child
White, and 72.6% caregiver who completed more than high school de­ maltreatment, eight response options were used, from “this has never
gree, See table 1). Compared to baseline sample, this study sample is happened,” to “more than 20 times”. For this study, dichotomized prev­
more likely to report having high school education on caregiver, χ2 (1) alence of each child maltreatment—the most frequently used score—is
= 35.26, p < .001. In addition, children in this study are more likely to used, 0 = never happened and 1 = happened one or more times.
be picked on by peers, χ2 (1) = 5.60, p = .018 and left from the activities,
χ2 (1) = 3.29, p = .070. but less likely to use cigarette, χ2 (1) = 5.38, p = 2.2.3. Adolescent psychosocial problems
.020 and marijuana, χ2 (1) = 21.94, p < .001. No other significant dif­ Adolescent psychosocial problems (age 15) includes substance use (i.
ferences between baseline sample and the study sample were found. e., alcohol, cigarette, and marijuana), trouble at school, and depressive
symptoms. For substance use, adolescent reported whether they have
2.2. Measures ever use alcohol, cigarette, and marijuana use. Each response option was
dichotomous variable, 0 = no and 1 = yes. To assess trouble at school, a
2.2.1. Bullying modified version of the National Longitudinal Study of Adolescent
Both bullying perpetration and victimization were measured at age Health Wave I In-School Questionnaire was used (McNeely, Nonne­
9. For bullying perpetration, teachers were asked about three items of maker, & Blum, 2002). Adolescents were asked about 4 items regarding
children’s aggressive behaviors toward peers (i.e., “fights with others;” their problems at schools such as paying attention and getting along
“threatens or bullies others;” and “argues with others”) using Social with your teachers. For each item, three response options were used: 1
Skills Rating System (Gresham & Elliot, 1990). For each item, three = never to 3 = often. A total scale of trouble at school was created by
response options were used: 1 = never, 2 = sometimes, 3 = often, and 4 = summing the responses, with higher scores indicating higher levels of
very often. All three items were dichotomized as either (0) never or (1) at trouble at school. Cronbach’s alpha for the trouble at schools in this
least one time. For bullying victimization, children reported their victim sample is 0.63. Adolescent depressive symptoms were measured using a
experiences with kids at school and neighborhoods during the last modified version of the Center for Epidemiologic Studies Depression
month, such as “picked on you or said mean things to you” and “taken your Scale (Radloff, 1977). Adolescent responded to the 5 questions
things, like your money or lunch, without asking.” using four items from the regarding depressive symptomatology during the past 4 weeks: e.g., “I
Panel Study of Income Dynamics, Child Development Supplement III. feel life is not worth living,” and “I feel depressed.” Each item was ranged

3
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

from 1 = strongly agree to 4 = strongly disagree. The four items—except

Note. Note. B bully; V victim; PA physical assault; PSY psychological aggression; NEG neglect; ALC alcohol use; MAR marijuana use; CIG cigarette use; DEP depression; TS trouble at schools; FEM female; EDU caregiver’s
18
for “I feel happy”—were reverse-coded and a summative scale of

1
depressive symptoms were created; higher scores indicate higher levels
of depressive symptoms. Cronbach’s alpha for the depressive symptoms

− 0.05
in this sample is 0.78.

17

1
2.2.4. Covariates

− 0.02
0.01
Gender (1 = female, 0 = male), race (1 = Black, 0 = non-Black), and

16

1
primary caregiver’s education (0 = less than high school, 1 = high school
degree or more) were included as control variables.

− 0.11***
0.10**
− 0.02
2.3. Statistical analyses

15

1
To address the three aims of the present study, a series of analyses

− 0.08**
0.26***
0.10**
(LCA, multinomial logistic regression, logistic regression, and ordinary

0.02
14
least-squares multiple regressions) were employed using Mplus 8.3.

1
Before conducting the LCA, descriptive statistics for all study variables

0.16***
0.10***

− 0.07*
and bivariate correlations between bullying related variables were

− 0.03

− 0.04
produced using SPSS v. 25. LCA is a person-centered approach to

13

1
identify distinct patterns using categorical indicators (Collins & Lanza,
2010). To determine the optimal number of classes, two to six class

− 0.12***
0.31***
0.12***
0.13***
solutions were estimated iteratively (Aim 1). Several model fit indices

0.08**
− 0.01
including Akaike Information Criteria (AIC), Bayesian Information

12

1
Criteria (BIC), the sample-size-adjusted BIC (ABIC), the Vuong Lo-
Mendell-Rubin test (VLMR), and Entropy were used to determine the

0.41***
0.30***
0.18***
0.13***

− 0.06*
− 0.01

− 0.02
best fitting model, along with class size and interpretability. The model

11

1
with the lower AIC, BIC, and ABIC indicates a better fit to the data
(Collins & Lanza, 2010; Nylund, Bellmore, Nishina, & Graham, 2007).

0.08**
0.09**

− 0.04
0.07*
The significant p-value of VLMR yield whether the k class model

0.05
0.05

0.03
0.05
10

improve the model fit, compared to the k-1 model (Nylund et al., 2007).

1
The higher value of Entropy (range: 0–1) indicates better classification

0.08**

0.09**
accuracy (Celeux & Soromenho, 1996). After the best LCA model was

− 0.05

− 0.01
0.07*
0.03

0.05

0.03

0.00
selected, multinomial logistic regression was conducted to examine
9

1
whether the patterns of bullying differ by type of child maltreatment
(Aim 2). In addition, multiple logistic regressions and ordinary least-
0.39***

0.16***
0.10**

0.06**

− 0.01
0.06*
0.01
0.05

0.01

0.00
squares multiple regressions were conducted to investigate the associ­
8

ation between bullying patterns and adolescent psychosocial problems

(Aim 3). To handle missing data, Full Information Maximum Likelihood

0.14***
0.08**

0.08**

0.10**
estimation was used.
0.01
0.06
0.05
0.04

0.02
0.03
0.05
7

3. Results
0.27***

0.09**
0.08**
− 0.02
− 0.02
− 0.01
− 0.01

− 0.01
0.07*
0.02

0.04

0.02
3.1. Patterns of bullying
6

Table 2 shows bivariate correlations among study variables. Table 3

− 0.11***

presents multiple model fit indices for each LCA model. The 4-class so­
0.32***
0.27***

0.10**

0.09**
− 0.01
− 0.02

− 0.04
0.08*
0.02
0.02

0.05
0.03

lution has the lowest value of AIC, BIC, and ABIC, suggesting that the 4-
5

class solution was the best fit for these data. In addition, the p-value of
VLMR indicated that the 4-class solution significantly improved the 3-
0.32***
0.25***
0.36***

0.14***

class solution, whereas 5-class solution did not improve the 4-class so­
0.06*
0.07*
0.08*

0.08*
0.05
0.02
0.00
0.02

0.04

0.00

lution. Thus, we have opted the 4-class solution. Fig. 1 describes a plot
4

1
Correlations between study variables (N = 1139).

with item-probabilities for the 4-class solution. The first class, bully-
victim (n = 226, 19.8%), suggested a pattern of high endorsement on
0.14***
0.12***
0.12***

0.13***

0.24***
− 0.07*
0.10**
− 0.02
− 0.04

− 0.04
0.08*
0.07*
0.05

0.04
0.01

both bullying perpetration and victimization items. The second class is

3

named victim (n = 186, 16.3%) because this class comprised individuals

p < .05, ** p < .01, *** p < .001

who reported relatively low levels of perpetration, whereas item prob­

− 0.09**

− 0.09**
0.50***
0.13***
0.16***

0.17***

0.11***

0.22***
0.08**

abilities of victim were high. The third class, named no bully-victim (n =

− 0.01
0.08*
0.05

0.03
0.01

0.05
0.01

443, 38.9%), reported low involvement in both bullying perpetration

2

and victimization. The fourth class is named bully (n = 284, 24.9%).

Individuals in this class reported high engagement in bullying perpe­
− 0.09**
0.62***
0.61***
0.16***
0.15***

0.11***

0.11***

0.20***
− 0.07*
0.08**
0.10**

− 0.03
− 0.03
0.07*

tration, but low victimization.

0.05

0.02
0.00
1

3.2. Child maltreatment types and bullying patterns

education.
17. Black
12. MAR
10. NEG

16. FEM

18. EDU
14. DEP
11. ALC

13. CIG
Table 2

9. PSY

15. TS
8. PA
4. V1
5. V2
6. V3
7. V4
1. B1
2. B2
3. B3

Using multinomial logistic regression, the different effects of child

maltreatment types on bullying patterns were found after controlling for
*

4
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

Table 3 covariates, African American had lower odds of being in victim (OR =
Model fit indices with 2 to 6 latent classes. 0.44, 95% CI = 0.29–0.66, p < .001) and no bully-victim classes (OR =
Class AIC BIC ABIC Entropy VLMR Smallest class 0.40, 95% CI = 0.29–0.56, p < .001) compared to the bully-victim class.
% Individuals whose caregiver completed more than high school grades
2 8477.29 8552.86 8505.21 0.855 p< 45.1 had higher odds of being in victim class, compared to bully-victim class
.001 (OR = 1.69, 95% CI = 1.07–2.66, p = .026). Female had 1.67 times
3 8252.76 8368.63 8295.58 0.785 p< 18.7 higher odds of being in the victim (95% CI = 1.14–2.43, p = .008) and
.001 1.64 times higher odds of being in the no bully-victim (95% CI =
4 8087.86 8244.03 8145.57 0.748 p< 16.3
.001
1.21–2.23, p = .002) classes, compared to bully class. African American
5 8094.02 8290.50 8166.62 0.704 p= 10.2 had lower odds of being in victim (OR = 0.47, 95% CI = 0.32–0.69, p <
.432 .001) and no bully-victim classes (OR = 0.42, 95% CI = 0.31–0.58, p <
6 8097.39 8334.17 8184.89 0.722 p= 5.0 .001) compared to the bully class. Individuals whose caregiver
.223
completed more than high school grades had higher odds of being in
Note. AIC = Akaike Information Criterion; BIC = Bayesian Information Criteria; victim class, compared to bully class (OR = 1.60, 95% CI = 1.03–2.49, p
ABIC = Sample-adjusted BIC; VLMR = Vuong-Lo-Mendell-Rubin. = .038).

gender, race, and caregiver’s education, χ 2 = 85.68, p < .001 (Table 4).
Compared to the bully-victim class, individuals who have been neglected 3.3. Bullying patterns and adolescent psychosocial problems
had 1.86 times higher odds of being in the victim class (95% CI =
1.04–3.33, p = .036), while individuals who have been physically Fig. 2 shows the effects of the bullying patterns on adolescent psy­
assaulted had lower odds of being in the victim class (OR = 0.55, 95% CI chosocial problems (i.e., substance use, trouble at schools, and depres­
= 0.32–0.97, p = .039). Compared to the bully class, individuals who sive symptoms) after controlling for gender, race, and caregiver’s
have been neglected had 2.06 times higher odds of being in the victim education. More individuals in the victim class (19.2%) drank alcohol
class (95% CI = 1.18–3.62, p = .011). Compared to the no bully-victim than those in no bully-victim class (12.0%). As for cigarette use, more
class, individuals who have been neglected had 1.75 times higher odds individuals in the bully-victim (7.0%) and victim (6.3%) classes smoked
of being in the victim class (95% CI = 1.06–2.87, p = .028). As for the cigarette than those in the no bully-victim class (2.0%). In addition, more
individuals in the bully-victim (22.9%) and bully (19.5%) classes smoked

Fig. 1. Plot with item-probabilities for the 4-class solution.

Table 4
Effects of Child Maltreatment Types on Bullying Classes, OR (95% CI).
Reference group Bully-victim vs. Bully vs. No Bully-victim vs.

Victim No Bully-Victim Bully Victim No Bully-Victim Victim

Female 1.36 1.34 0.82 1.67 1.64 1.02

(0.92–2.02) (0.96–1.85) (0.57–1.16) (1.14–2.43) (1.21–2.23) (0.72–1.44)
African American 0.44 0.40 0.94 0.47 0.42 1.10
(0.29–0.66) (0.29–0.56) (0.65–1.37) (0.32–0.69) (0.31–0.58) (0.77–1.56)
Caregiver’s education 1.69 1.20 1.05 1.60 1.14 1.40
(1.07–2.66) (0.84–1.72) (0.72–1.54) (1.03–2.49) (0.81–1.60) (0.83–2.12)
Physical assault 0.55 0.66 0.68 0.81 0.97 0.84
(0.32–0.97) (0.41–1.08) (0.40–1.15) (0.49–1.34) (0.64–1.48) (0.54–1.31)
Psychological aggression 1.24 1.20 1.99 0.62 0.60 1.03
(0.51–3.01) (0.57–2.55) (0.81–4.89) (0.25–1.34) (0.28–1.30) (0.50–2.15)
Neglect 1.86 1.07 0.90 2.06 1.18 1.75
(1.04–3.33) (0.63–1.80) (0.51–1.60) (1.18–3.62) (0.71–1.96) (1.06–2.87)

Note. The significant association is indicated as bold.

5
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

Fig. 2. Adolescent Behavioral Health by Latent Classes.

marijuana than those in the no bully-victim class (12.4%). In regard to neglected children. Often these children have lower self-esteem and tend
trouble at school, individuals in the no bully-victim (M = 6.85, SE = 0.09) to internalize their issues (Juvonen & Graham, 2014; Trickett &
reported having less trouble at school compared to bully-victim (M = McBride-Chang, 1995). Therefore, their ability to speak up and defend
7.68, SE = 0.13), victim (M = 7.68, SE = 0.14), and bully (M = 7.22, SE = themselves may be compromised. Future research should continue to
0.11) classes. Additionally, individuals in the bully class reported having focus on the subtypes of bullying/victimization as it relates to neglect,
less trouble at school compared to bully-victim and victim classes. As for with additional focus on the different subtypes of neglect (i.e., physical,
depressive symptoms, individuals in the bully-victim (M = 8.47, SE = psychological, environmental, or a combination). This level of speci­
0.20) and victim (M = 8.72, SE = 0.22) classes reported having more ficity might lend itself to targeted inventions that focus on individual
depressive symptoms than those in the no bully-victim class (M = 7.71, groups.
SE = 0.15) and bully class (M = 7.74, SE = 0.18). In regards to cova­ Interestingly, we identified physical assault to be at heightened risk
riates, African American had lower odds of alcohol use (OR = 0.69, 95% of involvement in the bully-victim class, compared to the victim class.
CI = 0.49–0.97, p = .033). African American had lower odds of cigarette Prior research indicated that individuals who have been physically
use (OR = 0.40, 95% CI = 0.21–0.75, p = .004). Individuals whose assaulted tend to externalize their behaviors through aggressive and
caregiver completed more than high school grades had lower odds of disruptive means (Teisl & Cicchetti, 2008). To be specific, physically
marijuana use (OR = 0.53, 95% CI = 0.38–0.74, p < .001). assaulted children may expand their social capital by expressing their
aggression to other peers with low social status. They may also be
4. Discussion victimized, however, by peers with high social status given their social
and emotional problems. Future research should consider the mecha­
This study was to identify the patterns of bullying among a high-risk nisms of how physical assault experiences lead to bully-victim
sample. The study also aimed to advance the understanding of which experiences.
specific types of child maltreatment predicted bullying patterns, and to Regardless of having perpetration experiences, in general, in­
determine what the associations were between bullying patterns and dividuals who have victim experiences are more likely to engage in
adolescent psychosocial outcomes. Our findings contribute to the psychosocial problems, such as using tobacco, exhibiting more trouble
existing research by identifying the distinct patterns of bullying in high at school, and having higher depressive symptoms. As for substance use,
risk sample. Overall, our findings with regard to empirical bullying the strong effects of bullying victimization are distinctly identified: in­
classification (i.e., bully-victim, bully, victim, and no bully-victim) are dividuals who have been victimized are more likely to use alcohol, even
in line with social capital theory explaining diverse patterns of bullying, compared to bully-victim groups. These findings corroborate the find­
and most previous studies suggesting four heterogeneous patterns of ings presented in previous work that being a victim is a known risk factor
bullying (Betterncourt et al., 2013; Lovegrove et al., 2012; Williford, for future substance use (Radliff, Wheaton, Robinson, & Morris, 2012;
Brisson, Bender, Jenson, & Forrest-Bank, 2011). Tharp-Taylor et al., 2009; Valdebenito, Ttofi, & Eisner, 2015). We posit
As predicted by developmental psychopathology perspectives, the that youth are more likely to engage in substance use as a way to self-
bullying patterns are differently influenced by child maltreatment types medicate. These youths need to reduce symptoms of stress, depression,
and show significant differences from previous research, providing new and anxiety and therefore may use multiple substances and engage in
insights. Our findings suggest that individuals who have been neglected risky behaviors as a way to cope. Although the overall effects of bullying
are more likely to be in the victim class compared to all other classes (i. patterns on alcohol and tobacco use are similar, while marijuana use
e., bully, bully-victim, and no bully-victim). This is inconsistent with differs from these other two substances: higher rates of marijuana use
other studies, which suggest that neglect leads to aggressive behaviors among the bully-victim and bully classes. This finding is similar to those
(Logan-Greene & Jones, 2015). O’Hara (2020) found that children with of previous studies which reported that adolescents being bullied were
a history of neglect were not at greater odds for bullying victimization, less likely to engage in marijuana use (Priesman, Newman, & Ford,
only children with physical assault reported victimization experiences. 2018). Perren and Alsaker (2006) showed that bullying victims tend to
One explanation for these divergent findings is two-fold: first, previous be less cooperative or sociable, and to have fewer friends. Considering
studies tend to separate bullying and victimization as opposed to looking that the most common source of marijuana for adolescents is their
at the issue as a continuum, while this study incorporated both bullying friends (King, Merianos, & Vidourek, 2016), those who are bullied may
and victimization. A second explanation lies in the behaviors of find it difficult to obtain marijuana. This finding can also be explained

6
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

by the process of obtaining power from peers using marijuana. Ac­ LCA allowed us to identify the patterns of bullying accurately.
cording to Tucker and colleagues (2014), using marijuana in adoles­
cence may be a strategy to attain social status. In other words, 5. Conclusions and implications
individuals who bully others or have both bully and victim experiences
may try to keep their status within their peer groups by using marijuana. Despite the results of our study, which augment the knowledge base
Future research should examine how important it is to attain social on bully/victimization, child maltreatment, and psychosocial problems,
status as a reason for using different types of substances. it is important to note that the remaining questions can be considered in
Our study found that no bully-victim performed better in school than future studies. Although this study identified the similar patterns of
any form of bullying (i.e., victim, bully, or bully-victim): a finding bullying regardless of different measure and data sources, researchers
consistent with what previous research has demonstrated (Menesini & should continue focusing on the association between the bullying pat­
Salmivalli, 2017; Nakamoto & Schwartz, 2009). What was unique in our terns and traumatic experiences because our findings are more focused
results was that the bully class had less trouble at school compared to the on the child maltreatment. Equally important is a continued focus on the
bully-victim and victim classes. These findings may be due to the self- long-term psychosocial problems of bully/victimization, particularly at
reports of trouble at schools. Given that individuals who bully others specific stages. Extant research has considered bullying only in middle
tend to exhibit low psychological distress (Lereya, Copeland, Zammit, & childhood, and its consequences only in adolescence. Future research
Wolke, 2015), they may consider themselves as “normal” at school. may consider the longitudinal changes of bullying patterns during the
More studies are needed to examine the role of psychological distress in developmental stages and their associations with adolescent psychoso­
the association between performance at school and different types of cial problems.
bullying involvement in order to illuminate how different bullying Our findings suggest that there is a level of risk for bullying/
involvement is associated with school life. victimization among maltreated youth, which may influence their psy­
It was not surprising that individuals in the bully-victim and victim chosocial problems. Although school-wide interventions are useful, their
classes reported having more depressive symptoms than those in the no effectiveness is mixed and what specifically works within each program
bully-victim class and bully class, as these are well documented outcomes differs considerably. What is known is that longer, more intensive pro­
(Garnefski & Kraaij, 2014). Although victimization is well known as a grams that involve parents and caregivers have some of the strongest
risk factor for internalizing problems including depression, these find­ effects (Menesini & Salmivalli, 2017). Thus, school administrators may
ings suggest a need for additional focus on the magnitude of a victim’s benefit from offering these types of interventions for higher-risk youth.
depression. Adolescent depression due to bullying victim experiences is Child welfare practitioners need, ongoing training and to recognize of
associated with a higher risk for suicidal ideation and suicidality (Kodish youth that may be in the bullying/victim categories. It is important to be
et al., 2016). Given the grave consequences, future research should aware that because these youths have a history of maltreatment, they
consider the relations of suicidality and depression among bully-victims are at higher risk for re-victimization. Child welfare practitioners addi­
and victims. tionally must forge collaborative relationships with school personnel,
namely school social workers, to ensure coordinated planning and
4.1. Strengths and limitations shared goal-setting that focuses both bullies and victims.

There are noted limitations of this study. First, the measure of child Funding
maltreatment relied on parent reports as opposed to using state
administrative data for verification. It is possible that there was under­ This manuscript uses data from the Fragile Families and Child
reporting or a gap between child or child protective services report. Wellbeing Study (FFCWS), which is funded by Eunice Kennedy Shriver
Second, sexual abuse information at age 5 were not gathered in the National Institute of Child Health and Human Development (NICHD),
FFCWS, which may differently affect the results of this study. Third, the the National Opinion Research Center (NORC), and Mathmatia Policy
bully-perpetration data came from teacher reports. There are many av­ Research. The collector of the original data, the funder, Princeton Uni­
enues research has taken to collect this information including self and versity’s Center for Research on Child Wellbeing (CRCW) and Center for
peer reports. Ideally, there would be a way to validate the reports made Health and Wellbeing (CHW), as well as Columbia University’s the
by the teacher. The fourth limitation was the inability to account for the Columbia Population Research Center (CPRC) and the National Center
different bullying types. Because bullying has been the subject of for Children and Families (NCCF) bear no responsibility for the analyses
research for over two decades, there has been some agreement on the or interpretations presented here. Dr. Shipe’s work was supported by a
subtypes of bullying (i.e., verbal, physical, and more recently, cyber grant from the Eunice Kennedy Shriver National Institute of Child
bullying); however, our dataset was unable to account for this. In Health and Human Development (T32 HD101390, Co-PIs: Jackson &
addition, the dichotomized bullying cannot capture the chronic or Noll).
severity of bullying. A final limitation was the timeframe for when the
adolescent’s behavior was measured. A large body of research has CRediT authorship contribution statement
focused on adult outcomes of bullying/victimization, yet having six
years between the event and outcome may not be enough time to fully Dalhee Yoon: Conceptualization, Data curation, Formal analysis,
account for future negative behaviors, particularly if the bullying/ Writing - original draft, Writing - review & editing. Stacey L. Shipe:
victimization began at a later age. Writing - original draft, Writing - review & editing, Validation. Jiho
Despite these limitations, there are multiple strengths in this paper Park: Writing - original draft, Writing - review & editing, Validation.
worth noting. First, this is the first study to examine the types of child Miyoung Yoon: Writing - original draft, Writing - review & editing,
maltreatment and their association with patterns of bullying using a Validation.
large number of high-risk children over an extended period of time.
Thus, we were able to capture behavior longitudinally. Multiple in­ Declaration of Competing Interest
terventions have been developed for bullying concerns, yet there remain
higher-risk groups that could benefit from targeted interventions. This The authors declare that they have no known competing financial
study contributes to the literature by identifying the unique effects of interests or personal relationships that could have appeared to influence
child maltreatment types on bullying patterns, as well as the role of the work reported in this paper.
bullying on further adolescent psychosocial problems. In addition, this
study used sophisticated analyses. As a person-centered approach, an

7
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

References Lereya, S. T., Copeland, W. E., Zammit, S., & Wolke, D. (2015). Bully/victims: A
longitudinal, population-based cohort study of their mental health. European Child &
Adolescent Psychiatry, 24(12), 1461–1471.
Alto, M., Handley, E., Rogosch, F., Cicchetti, D., & Toth, S. (2018). Maternal relationship
Lo, C. K. M., Ho, F. K., Emery, C., Chan, K. L., Wong, R. S., Tung, K. T. S., & Ip, P. (2021).
quality and peer social acceptance as mediators between child maltreatment and
Association of harsh parenting and maltreatment with internet addiction, and the
adolescent depressive symptoms: Gender differences. Journal of Adolescence, 63,
mediating role of bullying and social support. Child Abuse & Neglect, 113, 104928.
19–28.
https://doi.org/10.1016/j.chiabu.2021.104928
Barzilay, S., Brunstein Klomek, A., Apter, A., Carli, V., Wasserman, C., Hadlaczky, G., …
Lovegrove, P. J., & Cornell, D. G. (2014). Patterns of bullying and victimization
Wasserman, D. (2017). Bullying victimization and suicide ideation and behavior
associated with other problem behaviors among high school students: A conditional
among adolescents in Europe: A 10-country study. Journal of Adolescent Health, 61
latent class approach. Journal of Crime and Justice, 37(1), 5–22.
(2), 179–186.
Logan-Greene, P., & Jones, A. S. (2015). Chronic neglect and aggression/delinquency: A
Betterncourt, A., Farrell, A., Liu, W., & Sullivan, T. (2013). Stability and change in
longitudinal examination. Child Abuse & Neglect, 45, 9–20.
pattern of peer victimization and aggression during adolescence. Journal of Clinical
Lovegrove, P. J., Henry, K. L., & Slater, M. D. (2012). Examination of the predictors of
Child & Adolescent Psychology, 42(4), 429–441.
latent class typologies of bullying involvement among middle school students.
Bowes, L., Arseneault, L., Maughan, B., Taylor, A., Caspi, A., & Moffitt, T. E. (2009).
Journal of School Violence, 11(1), 75–93.
School, neighborhood, and family factors are associated with children’s bullying
McNeely, C. A., Nonnemaker, J. M., & Blum, R. W. (2002). Promoting school
involvement: A nationally representative longitudinal study. Journal of the American
connectedness: Evidence from the National Study of Adolescent Health. Journal of
Academy of Child & Adolescent Psychiatry, 48(5), 545–553.
School Health, 72(4), 138–146.
Celeux, G., & Soromenho, G. (1996). An entropy criterion for assessing the number of
Menesini, E., & Salmivalli, C. (2017). Bullying in schools: The state of knowledge and
clusters in a mixture model. Journal of Classification, 13(2), 195–212.
effective interventions. Psychology, Health & Medicine, 22(sup1), 240–253.
Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in developmental
Moses, M., & Williford, A. (2017). Individual indicators of self-reported victimization
psychopathology. Development and Psychopathology, 8(4), 597–600.
among elementary school-age students: A latent class analysis. Children and Youth
Collins, L. M., & Lanza, S. T. (2010). Latent class and latent transition analysis: With
Services Review, 83, 33–40.
applications in the social behavioral, and health sciences. Hoboken, NJ: Wiley.
Nakamoto, N., & Schwartz, D. (2009). Is peer victimization associated with academic
Cunningham, S., Goff, C., Bagby, R. M., Stewart, J. G., Larocque, C., Mazurka, R., …
achievement? A meta-analytic review. Social Development, 19, 221–242.
Harkness, K. L. (2019). Maternal-versus paternal-perpetrated maltreatment and risk
Nicholson, J. V., Chen, Y., & Huang, C.-C. (2018). Children’s exposure to intimate
for sexual and peer bullying revictimization in young women with depression. Child
partner violence and peer bullying victimization. Children and Youth Services Review,
Abuse & Neglect, 89, 111–121.
91, 439–446.
Duke, N. N., Pettingell, S. L., McMorris, B. J., & Borowsky, I. W. (2010). Adolescent
Nylund, K., Bellmore, A., Nishina, A., & Graham, S. (2007). Subtypes, severity, and
violence perpetration: Associations with multiple types of adverse childhood
structural stability of peer victimization: What does latent class analysis say? Child
experiences. Pediatrics, 125(4), e778–e786.
Development, 78(6), 1706–1722.
Espelage, D. L., Basile, K. C., & Hamburger, M. E. (2012). Bullying perpetration and
O’Brennan, L. M., Bradshaw, C. P., & Sawyer, A. L. (2009). Examining developmental
subsequent sexual violence perpetration among middle school students. Journal of
differences in the social-emotional problems among frequent bullies, victims, and
Adolescent Health, 50(1), 60–65.
bully/victims. Psychology in the Schools, 46(2), 100–115.
Evans, C. B. R., & Smokowski, P. R. (2016). Theoretical explanations for bullying in
O’Hara, M. A. (2020). Peer victimization of maltreated youth: Distinct risk for physically
school: How ecological processes propagate perpetration and victimization. Child
abused versus neglected children. Journal of School Health, 90(6), 457–464.
and Adolescent Social Work Journal, 33(4), 365–375.
Olweus, D. (1994). Bullying at school. In Aggressive Behavior (pp. 97–130). Boston, MA:
Farrington, D. P., Loeber, R., Stallings, R., & Ttofi, M. M. (2011). Bullying perpetration
Springer.
and victimization as predictors of delinquency and depression in the Pittsburgh
Özdemir, M., & Stattin, H. (2011). Bullies, victims, and bully-victims: A longitudinal
Youth Study. Journal of Aggression, Conflict and Peace Research, 3(2), 74–81.
examination of the effects of bullying-victimization experiences on youth well-being.
Garnefski, N., & Kraaij, V. (2014). Bully victimization and emotional problems in
Journal of Aggression, Conflict and Peace Research, 3(2), 97–102.
adolescents: Moderation by specific cognitive coping strategies? Journal of
Perren, S., & Alsaker, F. D. (2006). Social behavior and peer relationships of victims,
Adolescence, 37(7), 1153–1160.
bully-victims, and bullies in kindergarten. Journal of Child Psychology and Psychiatry,
Gaete, J., Tornero, B., Valenzuela, D., Rojas-Barahona, C. A., Salmivalli, C.,
47(1), 45–57.
Valenzuela, E., & Araya, R. (2017). Substance use among adolescents involved in
Priesman, E., Newman, R., & Ford, J. A. (2018). Bullying victimization, binge drinking,
bullying: A cross-sectional multilevel study. Frontiers in Psychology, 8, 1–14.
and marijuana use among adolescents: Results from the 2013 National Youth Risk
Giang, M. T., & Graham, S. (2008). Using latent class analysis to identify aggressors and
Behavior Survey. Journal of Psychoactive Drugs, 50(2), 133–142.
victims of peer harassment. Aggressive Behavior, 34(2), 203–213.
Radliff, K. M., Wheaton, J. E., Robinson, K., & Morris, J. (2012). Illuminating the
Glew, G. M., Fan, M. Y., Katon, W., Rivara, F. P., & Kernic, M. A. (2005). Bullying,
relationship between bullying and substance use among middle and high school
psychosocial adjustment, and academic performance in elementary school. Archives
youth. Addictive Behaviors, 37(4), 569–572.
of Pediatrics & Adolescent Medicine, 159(11), 1026–1031.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the
Goldweber, A., Waasdorp, T. E., & Bradshaw, C. P. (2013). Examining associations
general population. Applied Psychological Measurement, 1(3), 385–401.
between race, urbanicity, and patterns of bullying involvement. Journal of Youth and
Reichman, N. E., Teitler, J. O., Garfinkel, I., & McLanahan, S. S. (2001). Fragile families:
Adolescence, 42(2), 206–219.
Sample and design. Children and Youth Services Review, 23(4-5), 303–326.
Gresham, F. M., & Elliot, S. N. (1990). Social skills rating system manual. Circle Pines, MN:
Rettew, D. C., & Pawlowski, S. (2016). Bullying. Child and Adolescent Psychiatric Clinics of
American Guidance Service.
North America, 25(2), 235–242.
Hemphill, S. A., Kotevski, A., Herrenkohl, T. I., Bond, L., Kim, M. J., Toumbourou, J. W.,
Seeds, P. M., Harkness, K. L., & Quilty, L. C. (2010). Parental maltreatment, bullying, and
& Catalano, R. F. (2011). Longitudinal consequences of adolescent bullying
adolescent depression: Evidence for the mediating role of perceived social support.
perpetration and victimisation: A study of students in Victoria, Australia. Criminal
Journal of Clinical Child & Adolescent Psychology, 39(5), 681–692.
Behaviour and Mental Health, 21(2), 107–116.
Solberg, M. E., & Olweus, D. (2003). Prevalence estimation of school bullying with the
Hong, J. S., Espelage, D. L., Grogan-Kaylor, A., & Allen-Meares, P. (2012). Identifying
Olweus Bully/Victim Questionnaire. Aggressive Behavior: Official Journal of the
potential mediators and moderators of the association between child maltreatment
International Society for Research on Aggression, 29(3), 239–268.
and bullying perpetration and victimization in schools. Educational Psychology
Stein, J. A., Dukes, R. L., & Warren, J. I. (2007). Adolescent male bullies, victims, and
Review, 24(2), 167–186.
bully-victims: A comparison of psychosocial and behavioral characteristics. Journal
Hsieh, Y.-P., Shen, A.-T., Hwa, H.-L., Wei, H.-S., Feng, J.-Y., & Huang, S.-Y. (2020).
of Pediatric Psychology, 32(3), 273–282.
Associations between child Maltreatment, dysfunctional family environment, post-
Straus, M. A., Hamby, S. L., Finkelhor, D., Moore, D. W., & Runyan, D. (1998).
traumatic stress disorder and children’s bullying perpetration in a national
Identification of child maltreatment with the Parent-Child conflict Tactics Scales:
representative sample in Taiwan. Journal of Family Violence, 1–10.
Development and psychometric data for a national sample of American Parents.
Jansen, P. W., Verlinden, M., Berkel, A.-V., Mieloo, C., van der Ende, J., Veenstra, R., …
Child Abuse & Neglect, 22(4), 249–270.
Tiemeier, H. (2012). Prevalence of bullying and victimization among children in
Takizawa, R., Maughan, B., & Arseneault, L. (2014). Adult health outcomes of childhood
early elementary school: Do family and school neighbourhood socioeconomic status
bullying victimization: Evidence from a five-decade longitudinal British birth cohort.
matter? BMC Public Health, 12(1). https://doi.org/10.1186/1471-2458-12-494
American Journal of Psychiatry, 171(7), 777–784.
Juvonen, J., & Graham, S. (2014). Bullying in schools: The power of bullies and the
Teisl, M., & Cicchetti, D. (2008). Physical abuse, cognitive and emotional processes, and
plight of victims. Annual review of psychology, 65, 159–185.
aggressive/disruptive behavior problems. Social Development, 17(1), 1–23.
King, K. A., Merianos, A. L., & Vidourek, R. A. (2016). Characteristics of marijuana
Tharp-Taylor, S., Haviland, A., & D’Amico, E. J. (2009). Victimization from mental and
acquisition among a national sample of adolescent users. American Journal of Health
physical bullying and substance use in early adolescence. Addictive behaviors, 34
Education, 47(3), 126–135.
(6–7), 561–567.
Klomek, A. B., Sourander, A., & Gould, M. (2010). The association of suicide and bullying
Tomczyk, S., Hanewinkel, R., & Isensee, B. (2015). Multiple substance use patterns in
in childhood to young adulthood: A review of cross-sectional and longitudinal
adolescents—A multilevel latent class analysis. Drug and Alcohol Dependence, 155,
research findings. The Canadian Journal of Psychiatry, 55(5), 282–288.
208–214.
Kodish, T., Herres, J., Shearer, A., Atte, T., Fein, J., & Diamond, G. (2016). Bullying,
Toth, S. L., & Cicchetti, D. (2013). A developmental psychopathology perspective on
depression, and suicide risk in a pediatric primary care sample. Crisis, 37(3),
child maltreatment. Child Maltreatment, 18(3), 135–139.
241–246.
Trickett, P. K., & McBride-Chang, C. (1995). The developmental impact of different forms
Lebrun-Harris, L. A., Sherman, L. J., Limber, S. P., Miller, B. D., & Edgerton, E. A. (2019).
of child abuse and neglect. Developmental Review, 15(3), 311–337.
Bullying victimization and perpetration among US children and adolescents: 2016
Ttofi, M. M., Ttofi, M. M., Farrington, D. P., Lösel, F., & Loeber, R. (2011). Do the victims
National Survey of Children’s Health. Journal of Child and Family Studies, 28(9),
of school bullies tend to become depressed later in life? A systematic review and
2543–2557.

8
D. Yoon et al. Children and Youth Services Review 129 (2021) 106178

meta-analysis of longitudinal studies. Journal of Aggression, Conflict and Peace Williford, A. P., Brisson, D., Bender, K. A., Jenson, J. M., & Forrest-Bank, S. (2011).
Research, 3(2), 63–73. Patterns of aggressive behavior and peer victimization from childhood to early
Tucker, J. S., de la Haye, K., Kennedy, D. P., Green, H. D., & Pollard, M. S. (2014). Peer adolescence: A latent class analysis. Journal of Youth and Adolescence, 40(6),
influence on marijuana use in different types of friendships. Journal of Adolescent 644–655.
Health, 54(1), 67–73. Wolfe, D. A., Crooks, C. C., Chiodo, D., & Jaffe, P. (2009). Child maltreatment, bullying,
Turns, B. A., & Sibley, D. S. (2018). Does maternal spanking lead to bullying behaviors at gender-based harassment, and adolescent dating violence: Making the connections.
schools? A longitudinal study. Journal of Child and Family Studies, 27, 2824–2832. Psychology of Women Quarterly, 33(1), 21–24.
Valdebenito, S., Ttofi, M., & Eisner, M. (2015). Prevalence rates of drug use among school Yoon, D. (2020). Peer-relationship patterns and their association with types of child
bullies and victims: A systematic review and meta-analysis of cross-sectional studies. abuse and adolescent risk behaviors among youth at-risk of maltreatment. Journal of
Aggression and Violent Behavior, 23, 137–146. Adolescence, 80, 125–135.
Wang, X., Zhao, F., Yang, J., Gao, L., Li, B., Lei, L.i., … Wang, P. (2020). Childhood Yoon, D., Yoon, S., Park, J., & Yoon, M. (2018). A pernicious cycle: Finding the pathways
maltreatment and bullying perpetration among Chinese adolescents: A moderated from child maltreatment to adolescent peer victimization. Child Abuse & Neglect, 81,
mediation model of moral disengagement and trait anger. Child Abuse & Neglect, 106, 139–148.
104507. https://doi.org/10.1016/j.chiabu.2020.104507