You are on page 1of 13

J Youth Adolescence

DOI 10.1007/s10964-016-0570-7

EMPIRICAL RESEARCH

A Three-Step Latent Class Analysis to Identify How Different


Patterns of Teen Dating Violence and Psychosocial Factors
Influence Mental Health
1
Hye Jeong Choi ●
Rebecca Weston2 Jeff R. Temple3

Received: 17 June 2016 / Accepted: 7 September 2016


© Springer Science+Business Media New York 2016

Abstract Although multiple forms (i.e., physical, threa- the homogenous subgroups of teen dating violence and
tening, psychological, sexual, and relational abuse) and targeting adolescent teen dating violence accordingly.
patterns (i.e., perpetration and victimization) of violence can
co-occur, most existing research examines these experi-
ences individually. Thus, the purpose of this study is to Keywords Teen dating violence 3-Step latent class

investigate: (1) homogenous subgroups based on victimi- analysis Mental health Acceptance of couple violence
● ●

zation and perpetration of multiple forms of teen dating


violence; (2) predictors of membership in these subgroups;
and (3) mental health consequences associated with mem-
bership in each subgroup. Nine hundred eighteen adoles-
cents in the 9th or 10th grade at seven public high schools in
Texas participated in the survey (56 % female, White: 30 %, Introduction
Hispanic: 32 %, African American: 29 %, others: 9 %).
A three-step latent class analysis was employed. Five latent Each year, approximately 10 % of adolescents experience
teen dating violence classes were identified: (1) non- severe physical or sexual abuse (Kann et al. 2014), with
violence; (2) emotional/verbal abuse; (3) forced sexual higher rates observed in regional and at-risk samples
contact; (4) psychological + physical violence; and (5) (Zweig et al. 2013). It is estimated that approximately
psychological abuse. Females, African Americans, and 400,000 youth in the US are victims of severe teen dating
youth who had higher acceptance of couple violence scores violence annually (Wolitzky-Taylor et al. 2008). Although
and whose parents had less education were more likely to mutuality (O’Leary et al. 2008; Swahn et al. 2010) and
members of dating violence classes compared with the different types of violent acts (Ybarra et al. 2016) play an
nonviolence class. Adolescents who experienced multiple important role in teen dating violence, most existing stu-
types of dating violence reported greater mental health dies are restricted by focusing on one or two type(s) of
concerns. Prevention programs may benefit by identifying abuse or by considering only perpetration or victimization.
Although recent studies have started to examine both teen
dating violence experiences (e.g., perpetration and victi-
mization) or multiple forms of abuse (Choi and Temple
* Hye Jeong Choi 2016; Ybarra et al. 2016), less is known about how mul-
hjchoi0810@gmail.com
tiple types and experiences of teen dating violence are
1
Department of Health Sciences, University of Missouri, 501 Clark related with mental health and other individual and con-
Hall, Columbia, MO 65211, USA textual factors. The purpose of this study is to simulta-
2
Department of Psychology, University of Texas at San Antonio, neously investigate how different forms (e.g., physical,
San Antonio, TX 78249, USA psychological) and experiences of teen dating violence
3
Department of Obstetrics/Gynecology, University of Texas at (e.g., perpetration and victimization) are related to psy-
Medical Branch, Galveston, TX 77555, USA chosocial factors and mental health.
J Youth Adolescence

Heterogeneous Patterns of Teen Dating Violence and (e.g., violence) by observing positive outcomes from others’
Psychosocial Factors use of this behavior (e.g., interparental violence). Children
exposed to interparental violence may adopt attitudes that
Teen dating violence can refer to threatening behavior, or are accepting of violence—for instance, as a way to resolve
abuse against a current or former dating partner that is conflict and a normal part of intimate relationships. Indeed,
physical, psychological, relational, or sexual in nature recent empirical studies indicate that acceptance of violence
(Saltzman et al. 2002). Physical violence occurs when a mediates the link between witnessing interparental violence
youth deliberately and physically hurts his/her partner and teen dating violence victimization (Karlsson et al. 2016)
(e.g., hitting, pushing, shoving, chocking); sexual violence and perpetration (Temple et al. 2013). The developmental
includes any abusive sexual contact against a partner’s will, system model (Capaldi et al. 2005) suggests that interactions
such as forced sex or kissing; threatening behaviors include among individual (e.g., psychopathology), relational (e.g.,
verbal and/or nonverbal communication intended to make a couple satisfaction), and contextual (e.g., age, gender, eth-
partner feel fearful (Saltzman et al. 2002); psychological nicity) factors are important in predicting teen dating vio-
abuse includes humiliating, insulting, ridiculing, alienating, lence. For example, Novak and Furman (2016) found that
and blaming a partner (Wolfe et al. 2001); and relational externalizing symptoms and a negative relationship with
violence is defined as acts intended to deliberately hurt or one’s partner were significantly related to victimization and
manipulate a dating partner’s social relationships (Linder perpetration.
et al. 2002). Wolfe et al. (2001) demonstrated that these five Different types or patterns of teen dating violence may
abuse were interrelated with one another, yet distinguishable. have different socio-psychological characteristics. Although
In addition, teen dating violence has different types of research is not conclusive, several studies have found dif-
experience such as victimization only, perpetration only or ferences in victimization and perpetration or type of violent
both victimization and perpetration (O’Leary et al. 2008). acts by gender, ethnicity, and parental education. With
Despite their differences and all possible combinations types respect to gender, adolescent females generally report per-
and experiences (e.g., physical victimization, physical victi- petrating physical violence more often than males (Wolfe
mization, and psychological perpetration), these abusive acts et al. 2003; Foshee and Matthew 2007; Foshee et al. 2009),
often co-occur (Haynie et al. 2013) and should not be con- whereas males report more perpetration of sexual abuse
sidered in isolation. Studies conducted with nationally (Molidor and Tolman 1998) and more victimization of
representative samples showed that almost one-third of physical violence (Novak and Furman 2016; Williams and
adolescents reported both psychological victimization and Frieze 2005). Notably, females are more often victims of
perpetration, 14 % of youth reported psychological victimi- injurious and severe physical and sexual abuse (Molidor
zation and physical victimization or perpetration (Ybarra et al. 2000). With respect to ethnicity, African American
et al. 2016), and nearly all perpetrators of physical violence youth are more likely to be involved in physical (Carver
also perpetrate psychological abuse (Haynie et al. 2013). et al. 2003; Halpern et al. 2001), psychological, and
Although these different types of teen dating violence physical/psychological (Haynie et al. 2013) abuse compared
often share predictors, developmental course, and con- to other ethnicities. Finally, low-income youth, as deter-
sequences (Wolfe et al. 2001), few studies have investigated mined by parental education or other means, are generally
all types of teen dating violence simultaneously, and even more likely to be involved in teen dating violence (Foshee
fewer have considered sexual and threatening behaviors et al. 2009). While additional research is warranted, it is
(Schiff and Zeira 2005). Instead, teen dating violence is clear that gender, ethnicity, and parental education/socio-
often defined as physical and psychological (Earnest and economic status (SES) may be differently associated with
Brady 2016) or physical and sexual (Buttar et al. 2013) in the type of and experiences with abuse. In addition, atti-
nature. Complicating matters is the fact that teen dating tudes accepting of violence have been shown to predict teen
violence is often reciprocal and mutual (Swahn et al. 2010), dating violence (Temple et al. 2013; Williams et al. 2008;
with different types of abuse co-occurring (Haynie et al. Kinsfogel and Grych 2004; Josephson and Proulx 2008).
2013; Ybarra et al. 2016). Generally, male-to-female violence has been positively
related to physical teen dating violence perpetration and
Teen Dating Violence and Psychosocial Factors threatening behaviors (Lichter and McCloskey 2004), psy-
chological perpetration (Josephson and Proulx 2008), and
Several theoretical models—from sociocultural to individual physical and sexual victimization (O’Keefe and Treister
behavior theories—have been proposed to account for the 1998). Acceptance of female-to-male violence (by both
relationship between psychosocial risk and protection factors males and females) has been shown to be positively asso-
and teen dating violence. For example, social learning theory ciated with physical teen dating violence perpetration
(Bandura 1973) suggests that youth learn a specific behavior (Josephson and Proulx 2008).
J Youth Adolescence

Teen Dating Violence and Mental Health: Depressive For instance, female victims of physical and sexual dating
Symptoms, Anxiety, and Hostility violence had significantly higher odds of having symptoms
of depression compared to non-victims and male victims
The association between mental health and teen dating (Ackard et al. 2007). Although several studies have exam-
violence varies by the type and direction (perpetration vs. ined the associations between teen dating violence,
victimization) of the violence (Haynie et al. 2013; Bonomi including patterns (Haynie et al. 2013; Bonomi et al. 2012),
et al. 2012). Victimization has been linked to anxiety, individuals, relational, and contextual factors (Temple et al.
depressive symptoms, and hostility (Holt and Espelage 2013; Lehrer et al. 2006; Howard and Wang 2003; Williams
2005; Exner-Cortens et al. 2013; Ackard et al. 2007; et al. 2008), and health outcomes (Exner-Cortens et al.
Williams et al. 2008; Brown et al. 2009; Lehrer et al. 2006) 2013; Ackard et al. 2007; Howard and Wang 2003; Call-
with victims of physical, sexual, or psychological abuse ahan et al. 2003; Brown et al. 2009), less is known about
reporting poorer psychological functioning relative to non- how gender, ethnicity, parental education, and acceptance
victims (Brown et al. 2009; Howard and Wang 2005; of dating violence influences anxiety, depressive symptoms,
Ackard et al. 2007). Hostility, on the other hand, is and hostility symptoms within different patterns of teen
positively linked to perpetration (Temple et al. 2016). dating violence. Thus, we asked Research Question 4:
Further, experiencing multiple acts of violence relative to Depending on latent classes based on patterns of teen dating
experiencing psychological, physical, or sexual abuse in violence, do risk and protective factors differently influence
isolation may lead to more severe depressive and anxiety mental health (e.g., depressive symptoms, anxiety, and
symptoms. For instance, Exner-Cortens et al. (2013) and hostility)?
colleagues found that victims of psychological and physical
abuse evidenced symptoms of depression whereas victims
of “only” psychological abuse did not evidence depression. Methods

Participants and Procedure


Current Study
A total of 1042 adolescents in the 9th or 10th grade at seven
The first purpose of this study is to identify homogenous public high schools in Texas completed surveys at Wave 1
subgroups in teen dating violence by using more compre- and Wave 2 (retention rate: 92 %). At baseline, we recruited
hensive definitions and indicators based on the literature participants in state mandated classes (response rate: 62 %)
(Wolfe et al. 2001). We ask Research Question 1: Are there and received parental consent and student assent prior to
distinct latent classes (e.g., subgroups) of teen dating vio- administering surveys. The study was approved by IRB at
lence based on experience and types of abusive acts? the last author’s institution. Although 1042 students
Indeed, it is likely that gender, ethnicity, SES (level of participated in the study, only those who endorsed the item,
parental education as a proxy), and attitudes about violence “I have begun dating, going out with someone, or had a
play important roles in how teen dating violence develops boyfriend/girlfriend” at baseline were included in current
over time (Temple et al. 2013; Ackard et al. 2007; Halpern analyses (N = 918). Slightly more than half were female
et al. 2001; Howard and Wang 2003). Given the hetero- (56 %) and average age was 15.1 years old. Participants
geneous patterns of violence (e.g., different experiences and self-identified their racial/ethnic background as Hispanic
types), and that latent classes may have different char- (32 %), White (30 %), African-American (29 %), and other
acteristics in terms of gender, ethnicity, parental education, (9 %). Parental consent and student assent were obtained.
and acceptance of dating violence, we also ask Research Project managers administered the pen-and-paper survey
Question 2: Are memberships in latent violence classes at schools, and participants received gift cards as
associated with gender, ethnicity, level of parental educa- compensation.
tion, and acceptance of couple violence? Because youth’s
patterns of teen dating violence (e.g., perpetration and vic- Measures
timization; types of abuse) can influence their well-being,
we ask Research Question 3: Do distinct classes of teen Lifetime dating violence (Wave 1)
dating violence have differential effects on depressive
symptoms, anxiety, and hostility? The 50-item Conflict in Adolescent Dating Relationships
Furthermore, gender, ethnicity, parental education, and Inventory (Wolfe et al. 2001) was used to measure lifetime
attitudes about violence can differently influence mental teen dating violence (TDV) perpetration (25 items) and
health depending on the patterns of teen dating violence victimization (25 items). Based on theory and empirical
(e.g., perpetration and victimization; types of abuse). studies (Wolfe et al. 2001), ten subscales were created;
J Youth Adolescence

sexual victimization (four items, e.g., “He/she touched me to increase internal consistency). Youth reported their
sexually when I did not want him/her to,”) and perpetration past week depressive symptoms scores by answering the
(four items, e.g., “I touched him/her sexually when he/she following items on a 4-point scale (1 = less than 1 day and
did not want me to,”), relational victimization (three items, 4 = 5–7 days): “I had trouble keeping my mind on what I
e.g., “He/she said things to my friends about me to turn was doing”; “I felt depressed”; “I felt fearful”; “My sleep was
them against me,”) and perpetration (three items), emo- restless”; “I was happy” (reverse coded); “I felt lonely”;
tional/verbal victimization (ten items, e.g., “He/she ridiculed “I could not ‘get going’ ”; “I was bothered by things
or made fun of me in front of others,”) and perpetration (ten that usually don’t bother me” (internal consistency = .80,
items), threatening behavior victimization (four items, e.g., M = 1.81, SD = 0.59).
“He/she destroyed or threatened to destroy something I
valued,”) and perpetration (four items), and physical abuse Anxiety (Wave 2)
victimization (four items; e.g., “He/she kicked, hit, or pun-
ched me,”) and perpetration (four items). Each subscale was Five items were used from the Screen for Child Anxiety-
dichotomized so that any positive response meant any Related Emotional Disorders (Birmaher et al. 1997), in
experience within each subscale. Because of the common- which adolescents responded to the following on a 3-point
ness of emotional/verbal abuse and our desire to avoid scale (1 = not true and 3 = very true): “I am nervous”; “I
over-representing adolescents involved in psychological worry about being as good as other kids”; “I worry about
TDV (Follingstad 2007), participants had to endorse things working out for me”; “I am a worrier”; “People
four or more (of ten) of these items to be considered as tell me that I worry too much” (internal consistency = .85,
victims/perpetrators of psychological abuse. Indeed, M = 1.90, SD = 0.47).
most participants (80 %) reported at least one instance of
emotional/verbal abuse. Hostility (Wave 2)

Demographic information (Wave 1) A subscale of the Symptom Checklist-90 (SCL-90;


Derogatis 1983) was used to measure hostility. Participants
Gender (Female = 1) and ethnicity were dummy-coded. Three reported their hostility scores by answering the following
dummy variables were created for ethnicity (i.e., Hispanic = 1 items on a 4-point scale (1 = never and 4 = most of the
vs. other ethnicities = 0, White = 1 vs. other ethnicities = 0, time): In general, how often do you (1) “have temper out-
and African-American = 1 vs. other ethnicities = 0). Partici- bursts you can’t control?” (2) “have urges to beat, injure, or
pants reported the highest level of their parent education by harm someone?” (3) “have urges to break or smash things?”
responding on a scale of 1–4, with 1 = less than high school (4) “get into frequent arguments?” (5) “feel easily annoyed
graduation, 2 = high school graduation, 3 = some college or or irritated?” and (6) “shout or throw things?” (internal
training after high school, and 4 = college graduation. The consistency = .83, M = 1.89, SD = 0.63).
mean for parental education was 2.86 (SD = 1.08).

Acceptance of couple violence (Wave 1) Analytical Plan

To assess acceptance of couple violence, six items were A 3-step latent class analysis (LCA; Asparouhov and
adapted from Foshee et al. (1992). Acceptance of male-to- Muthén 2014) was employed to identify meaningful sub-
female (three items) and female-to-male violence (three groups of adolescents with similar TDV patterns and to
items) were measured separately. Adolescents reported the examine the associations between these subgroups and
degree to which they accepted male-female violence (e.g., distal outcomes and predictors. LCA is used to identify
“a boy angry enough to hit his girlfriend must love her very homogenous subgroups based on participants’ patterns of
much”; Cronbach’s α = .71, M = 1.24, SD = 0.45) on a 4- response to indicators (e.g., physical victimization, psy-
point scale (1 = strongly disagree, 4 = strongly agree), with chological perpetration). The 3-step LCA approach has
a higher score indicating greater acceptance. The same scale several advantages over 1-step LCA, including not having
was used to measure acceptance of female-male violence; to re-calculate estimations of LCA when including
Cronbach’s α = .77, M = 1.57, SD = 0.68. co-variates or distal outcomes while also taking account
of a classification uncertainty rate (Asparouhov and Muthén
Depressive symptoms (Wave 2) 2014). To identify the optimal number of latent classes, the
Bayesian information criterion (BIC), the adjusted
Eight items were adapted from The Center for Epidemio- Lo–Mendell–Rubin likelihood ratio test (LMR; Lo et al.
logic Studies Short Depression Scale (Andersen et al. 1994; 2001), and theoretical interpretation were employed.
10 items were used, and in this study we removed two items Monte Carlo simulation studies showed that LRT and BIC
J Youth Adolescence

are good indicators to select the optimal number of classes outcomes (e.g., depressive symptoms) among classes, a
(Nylund et al. 2007; Yang 2006; Tofighi and Enders 2008). Wald test was employed (Asparouhov and Muthén 2014). A
A smaller BIC indicates a better fit of the model, while LMR manual 3-step estimation procedure with all auxiliary vari-
offers a p-value in terms of a model comparison between the ables was employed because it allowed us to examine the
k-1 and k classes. That is, when the p-value is less than .05, it moderator role of latent class in linear regression to predict
indicates that the k-class model is statistically better than the outcomes (Asparouhov and Muthén 2014). The full infor-
k-1 class model. Except for indicators of LCA, other vari- mation maximum likelihood estimation method (Graham
ables such as gender, ethnicity, parental education, and et al. 2003) was employed to handle missing data.
acceptance of couple violence are understood as auxiliary
variables in the 3-step LCA approach (Asparouhov and
Muthén 2014). Mplus 7.11 (Muthén and Muthén 1998– Results
2012) was employed to examine the models, and R3STEP
command was used for membership prediction of a latent The type of TDV most frequently perpetrated or experi-
class model. To compare the mean of the difference in distal enced was emotional/verbal abuse, followed by physical
abuse, threatening behavior, sexual abuse, and relational
Table 1 Number of adolescents experiencing dating violence in five abuse (see Table 1). Youth were more likely to report vic-
domains timization by sexual or relational abuse compared with
Role Domain n (%) perpetration of those categories, whereas they reported
similar percentages of victimization and perpetration of
Perpetration Sexual abuse 121 (11.1) emotional/verbal, physical abuse, and threatening behavior.
Relational abuse 86 (7.9)
Emotional/verbal abuse 463 (42.5)
Research Question 1: Latent Class Analysis
Threatening behavior 144 (13.2)
Physical abuse 216 (19.8)
A five-class model provided the best fit with good con-
Victimization Sexual abuse 222 (20.4)
vergence (see Table 1). Based on fit criteria, we identified
Relational abuse 195 (17.9)
five meaningful classes (see Fig. 1). Classes were: (1)
Emotional/verbal abuse 458 (42.0)
nonviolence (NV, n = 374, 40.7 %); (2) emotional and
Threatening behavior 161 (14.8)
verbal abuse (EV, n = 281, 30.6 %); (3) forced sexual
Physical abuse 216 (19.8) contact (FSC, n = 107, 11.7 %); (4) psychological +

Fig. 1 Response patterns for latent classes. EV emotional/verbal abuse class, FSC forced sexual contact, PPV psychological + physical abuse
violence class, PA psychological abuse class
J Youth Adolescence

Table 2 Model fit index


Model Loglikelihood AIC BIC Adjust BIC LRT

1-Class solution −4554.51 9129.02 9177.24 9145.48 N/A


2-Class solution −3940.75 7923.50 8024.77 7958.07 1227.52***
3-Class solution −3825.59 7715.17 7869.48 7767.86 230.33***
4-Class solution −3761.57 7609.15 7816.50 7679.90 128.03***
5-Class solution −3722.14 7552.28 7812.68 7641.18 78.87***
6-Class solution −3691.05 7512.10 7825.54 7619.11 62.18***
Note LRT: adjusted Lo–Mendell–Rubin likelihood ratio test after six classes, the model did not converge
***p < .001

physical violence (PPV, n = 80, 8.7 %); and (5) psycholo- those in the nonviolence class, respectively. That is, the
gical abuse (PA, n = 76, 8.3 %). Youth in the nonviolence more accepting youth were of female-to-male violence, the
class did not experience any type of TDV. Youth in the more likely they were to be members of forced sexual
emotional/verbal abuse class had high item-response contact, psychological + physical violence, or psychological
probabilities on emotional/verbal victimization (.76) and abuse classes compared to the nonviolence class.
perpetration (.78). Youth in the forced sexual contact class
had high item-response probability on sexual abuse victi- Research Question 3: Distal Outcomes of Classes
mization (.83). Youth in the psychological + physical vio-
lence class had high item-response probabilities on Anxiety, depressive symptoms, and hostility were included
emotional/verbal victimization (.93) and perpetration (.96), in the model as distal outcomes of classes to compare their
threatening behavior victimization (.60) and perpetration outcomes’ scores among classes (see Fig. 2). Youth in the
(.75), and physical victimization (.74) and perpetration emotional/verbal abuse and psychological abuse classes had
(.90). Finally, youth in the psychological abuse class had significantly greater scores in anxiety compared with the
high item-response probabilities on emotional/verbal victi- nonviolence and forced sexual contact classes. Youth in the
mization (.92) and perpetration (.88) and threatening emotional/verbal abuse, psychological + physical violence,
behavior victimization (.70), and relational victimization and psychological abuse classes had significantly higher
(1.00), and physical victimization (.69) (Table 2). scores in depressive symptoms and hostility than those in
nonviolence and forced sexual contact classes.
Research Question 2: Latent Class Analysis with
Psychosocial Factors Research Question 4: The Role of Latent Class as a
Moderator between Psychosocial Factors and Distal
Gender, ethnicity, parental education, acceptance of female- Outcomes
to-male violence, and acceptance of male-to-female vio-
lence were included as covariates in the five-class model To examine the moderator role of these five latent classes,
(Table 3). The nonviolence class was considered as a gender, ethnicity, parental education, and acceptance of
reference class. Except for the forced sexual contact class, male-to-female/female-to-male violence were included in
the proportions of female youth were significantly greater the LCA with a distal outcome model (see Table 4).
than in the no violence class across the three TDV classes Because the mental health of youth in the nonviolence class
(e.g., psychological abuse class). African-Americans were was least crucial in this study, interpretation of this class
more likely to be members of the emotional/verbal abuse was not considered in the paper.
and psychological + physical violence classes. Youth who
had less educated parents were more likely to be members Anxiety
of the psychological abuse class compared with the non-
violence class. While acceptance of male-to-female vio- Female youth in the emotional/verbal abuse class were more
lence did not significantly predict class membership, likely to evidence severe anxiety symptoms than male
acceptance of female-to-male violence significantly differed youth of this class. In addition, Hispanic youth in the
between youth in the forced sexual contact class and youth psychological + physical violence class had significantly
in the nonviolence class, between youth in psychological + higher scores on anxiety than non-Hispanics in the same
physical violence class and those in the nonviolence class, class. There were no significant interactions between
and between youth in the psychological abuse class and psychological abuse class membership and psychosocial
J Youth Adolescence

Table 3 Predictors of membership in latent class of dating violence


EV FSC PPV PA
B SE OR B SE OR B SE OR B SE OR

Female 0.50* 0.21 1.64 0.55 0.29 1.73 1.42*** 0.39 4.15 0.63* 0.31 1.87
Hispanic (vs. others) 0.28 0.39 1.32 0.78 0.51 2.18 0.72 0.65 2.05 0.24 0.64 1.27
White (vs. others) 0.11 0.37 1.12 0.29 0.53 1.34 −0.94 0.88 0.39 0.90 0.59 2.47
African-American (vs. others) 0.77* 0.37 2.16 0.29 0.57 1.33 1.29* 0.65 3.62 0.66 0.63 1.93
Parental education −0.20 0.11 0.82 0.04 0.13 1.04 −0.12 0.17 0.89 −0.30* 0.15 0.74
Acceptance of couple (male-to-female) violence −0.62 0.42 0.54 −0.30 0.40 0.74 0.25 0.36 1.29 −0.07 0.38 0.93
Acceptance of couple (female-to-male) violence 0.32 0.22 1.37 0.75** 0.25 2.13 1.04*** 0.26 2.83 1.08*** 0.26 2.94
Note The latent class of nonviolence was set as the reference group
EV emotional/verbal abuse class, FSC forced sexual contact class, PPV psychological + physical violence class, PA psychological abuse class
*p < .05, **p < .01, ***p < .001

Fig. 2 Distal outcomes of each latent class. EmoV emotional/verbal the same subscripted alphabet letter are statistically indistinguishable
abuse class, FSC forced sexual contact class, PPV psychological + at the .001 alpha level. P was set at .001 to avoid Type I error due to
physical violence class, PA psychological abuse class. Estimates with multiple comparisons

factors (e.g., parental education, acceptance of violence) in greater hostility scores than did their counterparts in the
this study. same class. On the other hand, youth with greater accep-
tance of female-to-male violence in this class had sig-
Depressive symptoms nificantly more hostility than their counterparts in the same
class. Female youth in the psychological + physical violence
Youth in the psychological abuse class with less educated class had significantly stronger hostility than did male youth
parents reported significantly more depressive symptoms in the same class. Further, Hispanic youth in this class were
than did youth in the same class with a parent having higher less likely to have stronger hostility when compared with
education levels. Interactions between other classes and non-Hispanic youth in the same class.
psychosocial factors were not significant.
Alternate Model Analysis
Hostility
We estimated LCA models with different cut points
Youth in the emotional/verbal abuse class who had less (e.g., 1+, 2+, 3+) of emotional/verbal abuse indicators.
acceptance of male-to-female violence had significantly Although the models with different cut points regarding
J Youth Adolescence

Table 4 Moderator effects of each class on anxiety, depression, and hostility


Nonviolence EV FSC PPV PA
b se b se b se b se b se

Anxiety
Female 0.20*** 0.06 0.28** 0.08 −0.01 0.14 −0.01 0.14 0.13 0.13
Hispanic −0.10 0.09 −0.04 0.14 −0.07 0.23 0.58* 0.23 −0.20 0.22
White −0.28** 0.08 −0.03 0.15 0.13 0.22 0.23 0.23 −0.32 0.20
African-American −0.23** 0.09 −0.04 0.13 −0.05 0.19 0.44 0.24 −0.19 0.18
Parental education 0.06* 0.03 −0.02 0.04 −0.08 0.08 0.04 0.07 −0.02 0.06
Acceptance of male-female violence −0.03 0.07 0.03 0.12 −0.19 0.19 −0.23 0.15 −0.20 0.16
Acceptance of female-male violence 0.06 0.05 −0.03 0.07 0.00 0.12 0.10 0.11 0.04 0.10
Depressive symptoms
Female 0.12 0.07 0.15 0.11 0.12 0.17 0.49 0.34 0.19 0.14
Hispanic −0.13 0.13 −0.21 0.19 −0.45 0.42 0.33 0.34 −0.07 0.31
White −0.31** 0.11 −0.12 0.20 0.10 0.43 0.60 0.32 0.06 0.32
African-American −0.25* 0.11 −0.16 0.19 −0.21 0.40 0.36 0.40 −0.01 0.29
Parental education 0.03 0.03 −0.05 0.05 0.01 0.10 0.03 0.10 −0.19** 0.06
Acceptance of male-female violence 0.03 0.10 −0.07 0.11 −0.06 0.28 0.15 0.34 −0.01 0.15
Acceptance of female-male violence 0.02 0.07 0.05 0.09 0.00 0.14 −0.26 0.29 0.12 0.11
Hostility
Female 0.03 0.08 −0.01 0.14 0.16 0.09 0.85* 0.35 0.31 0.18
Hispanic −0.20 0.13 −0.14 0.27 −0.10 0.13 −0.82* 0.39 −0.45 0.41
White −0.12 0.10 −0.25 0.24 −0.09 0.13 −0.40 0.38 −0.30 0.39
African-American −0.01 0.11 0.11 0.26 −0.10 0.15 −0.76 0.51 −0.11 0.37
Parental education −0.03 0.04 −0.09 0.06 0.05 0.05 0.04 0.10 −0.01 0.09
Acceptance of male-female violence 0.02 0.12 −0.31* 0.13 −0.04 0.13 0.46 0.34 −0.06 0.20
Acceptance of female-male violence 0.06 0.07 0.22* 0.11 0.13* 0.06 −0.35 0.35 0.19 0.15
EV emotional/verbal abuse class, FSC forced sexual contact class, PPV psychological + physical abuse abuse class, PA psychological abuse class
*p < .05, **p < .01, ***p < .001

these indicators provided slightly different probabilities victimization by emotional/verbal abuse, physical violence,
across all indicators, the 5 class provided the best model fit threatening behavior, sexual abuse, and relational abuse.
based on fix criteria (e.g., lowered BIC) across all cut Indeed, while we know that certain psychosocial factors
points. Further, these models resulted in similar classes such (e.g., gender) increase risk for experiencing teen dating
as nonviolence, emotional/verbal abuse, forced sexual violence (Foshee, et al. 2009; Temple et al. 2013; Williams
contact, psychological + physical violence, and psycholo- et al. 2008), and that both victims (Exner-Cortens et al.
gical abuse class. 2013) and perpetrators (Temple et al. 2016) experience poor
psychological health, it is unclear how these variables relate
when examined jointly and simultaneously. Using 3-step
Discussion latent class analysis, the current study investigated the latent
classes based on experiences (e.g., perpetration and victi-
Due to the complex nature of teen dating violence (e.g., mization) and types of dating violence (e.g., physical,
different definitions, types, experiences, severity), previous psychological). In addition, we tested whether gender,
studies often focus on specific types (e.g., East and Hokoda ethnicity, parental education, and acceptance of couple
2015; Buttar, et al. 2013) and either perpetration (e.g., violence would predict class membership; we compared
Temple et al. 2016; Johnson et al. 2015) or victimization rates of anxiety, depressive symptoms, and hostility among
(e.g., Cascardi 2016; Vagi et al. 2015). Our goal was to classes; and we examined interaction effects on mental
provide a more comprehensive description of teen dating health as outcomes between latent class and membership
violence by collectively considering perpetration of and predictors.
J Youth Adolescence

Latent Classification of Teen Dating Violence and verbal abuse is an important component of teen dating
violence.
We identified five mutually exclusive between, yet homo- Finally, the psychological abuse class had asymmetrical
genous within latent classes: (1) nonviolence (40.7 %), (2) patterns of teen dating violence while the emotional/verbal
emotional/verbal abuse (30.6 %), (3) forced sexual contact abuse and psychological + physical violence classes
(11.7 %), (4) psychological + physical violence (8.7 %), and had almost symmetrical patterns as shown in Fig. 1. For
(5) psychological abuse (8.3 %). We confirmed the limited example, the emotional/verbal abuse class had high
previous research demonstrating the mutuality of violence response probabilities on emotional/verbal perpetration and
among adolescents (Ybarra et al. 2016; Haynie et al. 2013). victimization but low response probabilities on the
Also, this finding is partially consistent with previous stu- remaining indicators (e.g., relational abuse victimization).
dies using nationally representative samples but including However, the psychological abuse class had higher response
only some types of abuse (Haynie et al. 2013). In addition probabilities on only emotional/verbal perpetration but high
to identifying similar classes (nonviolence, emotional/ver- response probabilities on relational, emotional/verbal, and
bal abuse, and psychological + physical violence classes), threatening behavior victimization. Also, the psychological
we also identified classes reflecting psychological abuse and abuse class had a high response probability on physical
forced sexual contact. The emergence of these additional abuse victimization. This class illustrated that youth were
classes may be due to different operationalization of teen mutually or reciprocally involved in emotional/verbal dat-
dating violence and time frames used (Vezina and Hebert ing violence, yet they had experienced different types of
2007). Although teen dating violence has been categorized psychological forms of dating violence along with physical
as physical, psychological, or sexual (Ybarra et al. 2016) in abuse. Thus, these results indicate that emotional/verbal
the literature, almost 17 % of youth in our sample engaged abuse could be mutual which is consistent with previous
in multiple violent acts. Research should therefore extend literature (Haynie et al. 2013), but other forms of psycho-
the theoretical framework and empirical focus from only logical violence and physical abuse can be one-directional.
one type of violence or victimization vs. perpetration to a
more comprehensive examination of multiple types and Psychosocial Factors and Membership Predictions in
experiences of teen dating violence. For example, social Latent Classes and Mental Health
learning theory (Bandura 1973) or the developmental sys-
tem model (Capaldi et al. 2005) can guide research on how Our findings with regard to Research Question 3 replicate
and why victims can also be perpetrators, victimization and and extend existing research. Consistent with previous
perpetration can co-occur, and violence can be exacerbated. studies, gender (Foshee and Matthew 2007; Foshee, et al.
Notably, a third of adolescents engaged in emotional and 2009), ethnicity (Haynie et al. 2013; Foster et al. 2004),
verbal abuse. These two forms of abuse were important parental education (Foshee et al. 2009), and acceptance of
across almost all latent classes. Despite the high frequency couple violence (Karlsson et al. 2016; Temple et al. 2013)
of psychological abuse including these two forms (Fritz and were associated with teen dating violence. Our finding
Slep 2009), researchers generally focus on physical and indicated that female youth were more likely to be both
sexual abuse. This is an important limitation in the literature victims and perpetrators. Also, the finding showed that
as youth in the emotional/verbal abuse class reported nearly female youth were more likely to be a member of the
the same amount of anxiety, depressive symptoms, and psychological abuse class compared with male youth. This
hostility as did youth in the psychological abuse and psy- finding may explain inconsistent previous findings: females
chological + physical violence classes. Indeed, research on were more likely victims (Wolitzky-Taylor et al. 2008) or
intimate partner violence has consistently demonstrated that perpetrators (Foshee et al. 2009). That is, a gender
psychological abuse is as or more harmful than physical difference could occur depending upon the operationaliza-
violence (Mechanic et al. 2008; Pico-Alfonso 2005). tion of teen dating violence or research design. Also,
Although results from the current study do not allow us to females may be victims of different forms of psychological
infer whether physical violence and psychological abuse abuse (e.g., spreading rumors from a dating partner) and
occur simultaneously with the same partner, it is clear that physical abuse even though they are involved in some
youth who engage in physical violence are also engaging in degree of emotional/verbal abuse.
emotional/ verbal abuse—whether in different ways in Youth involved in dating violent relationships had poorer
different relationships or the same way in all relationships. mental health compared to adolescents without a history of
It is possible that emotional/verbal abuse serves as a dating violence, which is consistent with previous literature
“gateway” or risk factor for other types of violence. Addi- (Temple et al. 2016). Surprisingly, the forced sexual contact
tional research is needed to understand how violence class had levels of anxiety, depressive symptoms, and
develops, but it appears increasingly clear that emotional hostility most similar to the nonviolence class.
J Youth Adolescence

Given the fact that forcing a partner to have sex is illegal relationship between acceptance of couple violence and
(and reportable), it is possible that youth did not endorse hostility. Acceptance of couple violence and hostility may
their or their partner’s sexually violent behavior in this be related in a more complex way with teen dating violence
survey. This information can involve different mental health than other predictive factors.
outcomes. Indeed, in our sample, sexual abuse victimization Intervention programs should target victimization and
was made up primarily of youth who endorsed “unwanted perpetration of multiple types of abuse and target at-risk
kiss” (perpetration: 97 % and victimization: 91 %). One youth. Given our findings and those of others (Connolly
possible explanation is that youth may think this “unwanted et al. 2010; Stocker and Richmond 2007), special emphasis
kiss” is part of normal behavior in relationship and a sign of should be placed on recognizing and preventing the
love (Henton et al. 1983). However, this and other less occurrence of psychological abuse and characteristics of
severe forms of sexual abuse should be examined further to healthy and unhealthy romantic relationships. Because
determine whether youth perceive such behaviors as psychosocial factors such as gender and acceptance of
abusive. violence can be used to predict membership in teen dating
Findings in this study illustrate that individual and con- violence classes and may have different effects on mental
textual factors (e.g., gender, acceptance of violence) related health, targeted and tailored dating violence interventions
to mental health may vary depending on dating violence should be developed. For example, dating violence inter-
class based on their experience and violent acts. For ventions could be tailored to youth with low parental edu-
example, females in the emotional/verbal abuse class and cation/socio-economic status and for youth who have
Hispanics in the psychological + physical violence class permissive attitudes toward to the use of violence—perhaps
exhibited higher anxiety levels compared to other youth in those who have witnessed interparental violence (Karlsson
the same classes, respectively. Level of parental education et al. 2016). In addition, our findings indicated that African
was particularly important for youth in the psychological Americans were more likely to be involved in mutual
abuse class with respect to depressive symptoms. Female emotional/verbal abuse and physical abuse, and that His-
and non-Hispanic youth in the psychological + physical panics were more likely to experience greater anxiety when
violence class showed greater hostility compared with male involved in mutual emotional/verbal abuse and physical
and Hispanic youth in the same class, respectively. dating violence. Therefore, culturally targeted or tailored
These findings indicate that contextual factors such as programs may be useful for prevention and intervention.
gender, ethnicity, and parental education can be risk factors For example, our results suggest intervention programs
or be protective to mental health. Hispanic youth might not designed for Hispanic students should incorporate techni-
be differently involved in teen dating violence compared ques for coping with anxiety.
with other ethnicities; Hispanic youth could be at-risk in
terms of mental health once they are involved in physical Limitations
and psychological dating violence.
Acceptance of couple violence was related with poor Our findings should be interpreted in light of several lim-
mental health within the forced sexual contact and emo- itations. First, severity, duration, and recency of teen dating
tional/verbal abuse classes. Youth with higher scores of violence were not considered. Second, our categorization of
acceptance of female-to-male violence might have higher emotional/verbal abuse can influence identification of latent
hostility resulting in sexual abuse victimization. Because teen dating violence classes. However, this was necessary as
hostility is positively related with dating violence (Connolly approximately four-fifths of youth in our sample endorsed
et al. 2010), youth in the forced sexual contact class may be at least one emotional/verbal abuse item (e.g., of 10 items).
at a greater risk of re-victimization (Humphrey and White Moreover, Follingstad (2007) argued that psychological
2000). Counterintuitively, youth with less acceptance of abuse has a larger chance of happening than not happening
male-to-female violence in emotional/verbal violence abuse in romantic relationships. Consequently, increasing the
class were more likely to have high rates of hostility. This threshold for emotional/verbal abuse was necessary to
finding may be understood in how class membership was identify latent classes in this study. Third, our data were
predicted. Given our finding that females were more likely from self-reported survey with participants in southeast
to be members of the emotional/verbal abuse class, females Texas. Although the sample is ethnically diverse, our results
in this class may have higher acceptance of female-to-male may not be generalizable to other regions. Fourth, we did
violence and lower acceptance of male-to-female violence. not control for previous mental health status. Finally, our
That is, acceptance of couple violence would be positively measures did not allow us to distinguish whether youth
associated with hostility when youth involved in emotional engaged in mutual violence with the same partner or if
and verbal dating violence. A recent study (Temple et al. youth experienced violence from one partner and then
2016) showed that psychological perpetration mediated the engaged in violence with another partner.
J Youth Adolescence

Conclusion References

We extended past research on dating violence among ado- Ackard, D. M., Eisenberg, M. E., & Neumark-Sztainer, D. (2007).
Long-term impact of adolescent dating violence on the behavioral
lescents by examining their perpetration of and victimiza-
and tsychological health of male and female youth. Journal of
tion by five types of behaviors: emotional/verbal abuse, Pediatrics, 151(5), 476–481.
relational abuse, threats, physical violence, and sexual Andersen, G., Vestergaard, K., & Lauritzen, L. (1994). Effective
abuse. A latent class analysis of these variables provided treatment of poststroke depression with the selective serotonin
reuptake inhibitor citalopram. Stroke, 25(6), 1099–1104.
support for five distinct classes of dating violence repre-
Asparouhov, T., & Muthén, B. (2014). Auxiliary variables in mixture
senting nonviolence, emotional/verbal abuse, psychological modeling: three-step approaches using Mplus. Structural
+ physical violence, forced sexual contact, and psycholo- Equation Modeling: A Multidisciplinary Journal, 21(3),
gical abuse. In addition, we found that female gender was 329–341. doi:10.1080/10705511.2014.915181.
Bandura, A. (1973). Aggression: a social learning analysis.
associated with membership in dating violence classes
Englewood Cliffs, NJ: Prentice-Hall.
except forced sexual contact, supporting consistent recent Birmaher, B., Khetarpal, S., Brent, D., Cully, M., Balach, L.,
research that females are more likely than males to report Kaufman, J., & Neer, S. M. (1997). The screen for child anxiety
perpetration of less severe violence (Foshee et al. 2009). related emotional disorders (SCARED): scale construction and
psychometric characteristics. Journal of the American Academy
Lower levels of parental education were associated with
of Child & Adolescent Psychiatry, 36(4), 545–553.
membership in the psychological abuse class, and accep- Bonomi, A. E., Anderson, M. L., Nemeth, J., Bartle-Haring, S.,
tance of female to male violence specifically was associated Buettner, C., & Schipper, D. (2012). Dating violence victimiza-
with membership in the forced sexual contact, psychologi- tion across the teen years: abuse frequency, number of abusive
partners, and age at first occurrence. BMC Public Health, 12(1),
cal abuse, or psychological + physical violence classes
637–647.
relative to membership in the nonviolence class. Further- Brown, A. L., Testa, M., & Messman-Moore, T. L. (2009).
more, we found that mental health outcomes varied by class Psychological consequences of sexual victimization resulting
and, in some cases, outcomes were moderated by psycho- from force, incapacitation, or verbal coercion. Violence against
Women, 15, 898–919.
social factors. These results have implications for inter-
Buttar, A., Clements-Nolle, K., Haas, J., & Reese, F. (2013). Dating
vention (vs. prevention) programs. Targeted programming violence, psychological distress, and attempted suicide among
can be delivered for students based on screening for specific female adolescents in the juvenile justice system. Journal of
types of violence with psychosocial factors. Correctional Health Care, 19, 101–112.
Callahan, M. R., Tolman, R. M., & Saunders, D. G. (2003). Adoles-
cent dating violence victimization and psychological well-being.
Funding This research was supported by Award Number Journal of Adolescent Research, 18, 664–681.
K23HD059916 (PI: Temple) from the Eunice Kennedy Shriver Capaldi, D. M., Shortt, J. W., & Kim, H. K. (2005). A life span
National Institute of Child Health & Human Development (NICHD) developmental systems perspective on aggression toward a part-
and 2012-WG-BX-0005 (PI: Temple) from the National Institute of ner. In W. M. Pinsof, J. Lebow (Eds.), Family psychology: the art
Justice (NIJ). The content is solely the responsibility of the authors and of the science (pp. 141–167). New York: Oxford University
does not necessarily represent the official views of NICHD or NIJ. Press.
Carver, K., Joyner, K., & Udry, J. R. (2003). National estimates of
Authors’ Contributions HJC conceived of the study, participated adolescent romantic relationships. In P. Florsheim (Ed.),
in its design and interpretation of the data, performed the statistical Adolescent romantic relations and sexual behavior: theory,
analysis, and drafted the manuscript; RW participated in interpretation research, and practical implications (pp. 23–56). Mahwah, NJ:
of the data and helped to draft the manuscript; JT conceived of the Lawrence Erlbaum Associates Publishers.
study, participated in its design and coordination, interpretation of the Cascardi, M. (2016). From violence in the home to physical dating
data and drafted the manuscript. All authors read and approved the violence victimization: the mediating role of psychological dis-
final manuscript. tress in a prospective study of female adolescents. Journal of
Youth and Adolescence, 45, 777–792.
Choi, H. J., & Temple, J. R. (2016). Do gender and exposure to
Compliance with ethical standards interparental violence moderate the stability of teen dating
violence?: latent transition analysis. Prevention Science, 17,
367–376.
Conflict of interest The authors declare that they have no conflict of Connolly, J., Friedlander, L., Pepler, D., Craig, W., & Laporte, L.
interest. (2010). The ecology of adolescent dating aggression: attitudes,
relationships, media use, and socio-demographic risk factors.
Ethical Approval All procedures performed in studies involving Journal of Aggression, Maltreatment & Trauma, 19(5), 469–491.
human participants were in accordance with the ethical standards of Derogatis, L. R. (1983). SCL-90-R: administration, scoring, and
the institutional and/or national research committee and with the 1964 proce-dures manual–II. Towson, MD: Clinical Psychometric
Helsinki declaration and its later amendments or comparable ethical Research.
standards. Earnest, A. A., & Brady, S. S. (2016). Dating violence victimization
among high school students in Minnesota associations with
family violence, unsafe schools, and resources for support.
Informed Consent Informed consent was obtained from all indi- Journal of Interpersonal Violence, 31, 383–406.
vidual participants included in the study.
J Youth Adolescence

East, P. L., & Hokoda, A. (2015). Risk and protective factors intimate partner violence during adolescence and young adult-
for sexual and dating violence victimization: a longitudinal, hood. Journal of Youth and Adolescence, 44, 708–726.
prospective study of Latino and African American adolescents. Josephson, W. L., & Proulx, J. B. (2008). Violence in young
Journal of Youth and Adolescence, 44, 1288–1300. adolescents’ relationships a path model. Journal of Interpersonal
Exner-Cortens, D., Eckenrode, J., & Rothman, E. (2013). Longitudinal Violence, 23(2), 189–208.
associations between teen dating violence victimization and Karlsson, M. E., Temple, J. R., Weston, R., & Le, V. D. (2016).
adverse health outcomes. Pediatrics, 131(1), 71–78. Witnessing interparental violence and acceptance of dating
Follingstad, D. R. (2007). Rethinking current approaches to psycho- violence as predictors for teen dating violence victimization.
logical abuse: conceptual and methodological issues. Aggression Violence against Women, 22, 625–646.
and Violent Behavior, 12(4), 439–458. Kann, L., Kinchen, S., Shanklin, S. L., Flint, K. H., Kawkins, J.,
Foshee, V. A., Fothergill, K., & Stuart, J. (1992). Results from the Harris, W. A., & Zaza, S. (2014). Youth risk behavior surveil-
teenage dating abuse study conducted in Githens Middle School lance—United States, 2013. Morbidity and Mortality Weekly
and Southern High Schools. Chapel Hill, NC: University of North Report, 63(supplement 4), 1–168.
Carolina. Unpublished Technical Report. Kinsfogel, K. M., & Grych, J. H. (2004). Interparental conflict and
Foshee, V. A., & Matthew, R. A. (2007). Adolescent dating abuse adolescent dating relationships: integrating cognitive, emotional,
perpetration: a review of findings, methodological limitations, and peer influences. Journal of Family Psychology, 18(3),
and suggestions for future research. In D. J. Flannery, A. T. 505–515.
Vazsonyi, & I. D. Waldman (Eds.), The Cambridge Handbook of Lehrer, J. A., Buka, S., Gortmaker, S., & Shrier, L. A. (2006).
Violent Behavior and Aggression (pp. 431–449). New York: Depressive symptomatology as a predictor of exposure to inti-
Cambridge University Press. mate partner violence among US female adolescents and young
Foshee, V. A., Reyes, M. L., & Wyckoff, S. (2009). Approaches to adults. Archives of Pediatrics & Adolescent Medicine, 160(3),
preventing psychological, physical, and sexual partner abuse. In 270–276.
K. D. O’Leary & E. M. Woodin (Eds.), Psychological and Lichter, E. L., & McCloskey, L. A. (2004). The effects of childhood
physical aggression in couples: Causes and Interventions exposure to marital violence on adolescent gender‐role beliefs
(pp. 165–189). Washington, DC: American Psychological and dating violence. Psychology of Women Quarterly, 28(4),
Association. 344–357.
Foster, H., Hagan, J., & Brooks‐gunn, J. (2004). Age, puberty, and Linder, J. R., Crick, N. R., & Collins, W. A. (2002). Relational
exposure to intimate partner violence in adolescence. Annals of aggression and victimization in young adults’ romantic relation-
the New York Academy of Sciences, 1036(1), 151–166. ships: associations with perceptions of parent, peer, and romantic
Fritz, P. A., & Slep, A. M. S. (2009). Stability of physical and relationship quality. Social Development, 11(1), 69–86.
psychological adolescent dating aggression across time and Lo, Y., Mendell, N. R., & Rubin, D. B. (2001). Testing the number of
partners. Journal of Clinical Child and Adolescent Psychology, components in a normal mixture. Biometrika, 88(3), 767–778.
38, 303–314. Mechanic, M. B., Weaver, T. L., & Resick, P. A. (2008). Mental
Graham, J. W., Cumsille, P. E., & Elek-Fisk, E. (2003). Methods for health consequences of intimate partner abuse a multidimensional
handling missing data. In J. A. Schinka, W. F. Velicer (Eds.), assessment of four different forms of abuse. Violence Against
Research methods in psychology (pp. 87–114). New York: Women, 14, 634–654.
Wiley. Volume 2 of Handbook of Psychology (I. B. Weiner, Molidor, C., & Tolman, R. M. (1998). Gender and contextual factors
Editor-in-Chief). in adolescent dating violence. Violence against Women, 4(2),
Halpern, C. T., Oslak, S. G., Young, M. L., Martin, S. L., & Kupper, 180–194.
L. L. (2001). Partner violence among adolescents in opposite-sex Molidor, C., Tolman, R. M., & Kober, J. (2000). Gender and
romantic relationships: findings from the national longitudinal contextual factors in adolescent dating violence. Prevention
study of adolescent health. American Journal of Public Health, Research, 7(1), 1–4.
91(10), 1679–1685. Muthén, L. K., & Muthén, B. O. (1998–2012). Mplus user’s guide. 7th
Haynie, D. L., Farhat, T., Brooks-Russell, A., Wang, J., Barbieri, B., edn Los Angeles, CA: Muthén and Muthén.
& Iannotti, R. J. (2013). Dating violence perpetration and Novak, J., & Furman, W. (2016). Partner violence during adolescence
victimization among US adolescents: prevalence, patterns, and and young adulthood: individual and relationship level risk fac-
associations with health complaints and substance use. Journal of tors. Journal of Youth and Adolescence, 45, 1849–1861.
Adolescent Health, 53(2), 194–201. Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on
Henton, J., Cate, R., Koval, J., Lloyd, S., & Christopher, S. (1983). the number of classes in latent class analysis and growth mixture
Romance and violence in dating relationships. Journal of Family modeling: a Monte Carlo simulation study. Structural Equation
Issues, 4, 467–482. Modeling, 14(4), 535–569.
Holt, M. K., & Espelage, D. L. (2005). Social support as a moderator O’Keefe, M., & Treister, L. (1998). Victims of dating violence among
between dating violence victimization and depression/ high school students are the predictors different for males and
anxiety among African American and Caucasian adolescents. females? Violence against Women, 4(2), 195–223.
School Psychology Review, 34(3), 309–328. O’Leary, K. D., Slep, A. M. S., Avery-Leaf, S., & Cascardi, M. (2008).
Howard, D. E., & Wang, M. Q. (2003). Psychosocial factors asso- Gender differences in dating aggression among multiethnic high
ciated with adolescent boys’ reports of dating violence. Adoles- school students. Journal of Adolescent Health, 42(5), 473–479.
cence, 38(151), 519–533. Pico-Alfonso, M. A. (2005). Psychological intimate partner violence:
Howard, D. E., & Wang, M. Q. (2005). Psychosocial correlates of US The major predictor of posttraumatic stress disorder in abused
adolescents who report a history of forced sexual intercourse. women. Neuroscience & Biobehavioral Reviews, 29, 181–193.
Journal of Adolescent Health, 36(5), 372–379. Saltzman, L. E., Fanslow, J. L., McMahon, P. M., & Shelley, G. A.
Humphrey, J. A., & White, J. W. (2000). Women’s vulnerability to (2002). Intimate partner violence surveillance: uniform defini-
sexual assault from adolescence to young adulthood. Journal of tions and recommended data elements: Version 1.0. Atlanta, GA:
Adolescent Health, 27(6), 419–424. Centers for Disease Control and Prevention, National Center for
Johnson, W. L., Giordano, P. C., Manning, W. D., & Longmore, M. A. Injury Prevention and Control.
(2015). The age–IPV curve: changes in the perpetration of
J Youth Adolescence

Schiff, M., & Zeira, A. (2005). Dating violence and sexual risk conflict in adolescent dating relationships inventory. Psycholo-
behaviors in a sample of at-risk Israeli youth. Child Abuse & gical Assessment, 13(2), 277.
Neglect, 29(11), 1249–1263. Wolfe, D. A., Crooks, C. V., Lee, V., McIntyre-Smith, A., & Jaffe,
Silverman, J. G., Raj, A., Mucci, L. A., & Hathaway, J. E. (2001). P. G. (2003). The effects of children’s exposure to domestic
Dating violence against adolescent girls and associated substance violence: a meta-analysis and critique. Clinical Child and Family
use, unhealthy weight control, sexual risk behavior, pregnancy, Psychology Review, 6(3), 171–187.
and suicidality. Journal of the American Medical Association, Wolitzky-Taylor, K. B., Horowitz, J. D., Powers, M. B., & Telch, M.
286(5), 572–579. J. (2008). Psychological approaches in the treatment of specific
Stocker, C. M., & Richmond, M. K. (2007). Longitudinal associations phobias: a meta-analysis. Clinical Psychology Review, 28(6),
between hostility in adolescents’ family relationships and 1021–1037.
friendships and hostility in their romantic relationships. Journal Yang, C. C. (2006). Evaluating latent class analysis models in quali-
of Family Psychology, 21(3), 490–497. tative phenotype identification. Computational Statistics & Data
Swahn, M., Alemdar, M., & Whitaker, D. J. (2010). Nonreciprocal and Analysis, 50(4), 1090–1104.
reciprocal dating violence and injury occurrence among urban Ybarra, M. L., Espelage, D. L., Langhinrichsen-Rohling, J., &
youth. Western Journal of Emergency Medicine, 11, 264–268. Korchmaros, J. D. (2016). Lifetime prevalence rates and overlap
Temple, J. R., Shorey, R. C., Tortolero, S. R., Wolfe, D. A., & Stuart, of physical, psychological, and sexual dating abuse perpetration
G. L. (2013). Importance of gender and attitudes about violence and victimization in a national sample of youth. Archives of
in the relationship between exposure to interparental violence and Sexual Behavior, 45, 1083–1099.
the perpetration of teen dating violence. Child Abuse & Neglect, Zweig, J. M., Dank, M., Yahner, J., & Lachman, P. (2013). The rate of
37(5), 343–352. cyber dating abuse among teens and how it relates to other forms
Temple, J. R., Choi, H. J., Elmquist, J., Hecht, M., Miller-Day, M., of teen dating violence. Journal of Youth and Adolescence, 42,
Stuart, G. L., & Wolford-Clevenger, C. (2016). Psychological 1063–1077.
abuse, mental health, and acceptance of dating violence among
adolescents. Journal of Adolescent Health, 59, 197–202.
Tofighi, D., & Enders, C. K. (2008). Identifying the correct number of
classes in growth mixture models. In G. R. Hancock (Ed.), Hye Jeong Choi Ph.D. is an Assistant Professor at University of
Mixture models in latent variable research (pp. 317–341).
Missouri-Columbia. Her research focuses on adolescent social
Greenwich, CT: Information Age.
interactions and influence within health and prevention. She is
Vagi, K. J., Olsen, E. O. M., Basile, K. C., & Vivolo-Kantor, A. M.
particularly interested in adolescent informal communication
(2015). Teen dating violence (physical and sexual) among US
high school students: findings from the 2013 National Youth Risk regarding prevention and its influence.
Behavior Survey. JAMA Pediatrics, 169, 474–482.
Vezina, J., & Hebert, M. (2007). Risk factors for victimization in
romantic relationships of young women a review of empirical Rebecca Weston Ph.D. is an Associate Professor in the Department of
studies and implications for prevention. Trauma, Violence, & Psychology at the University of Texas at San Antonio. Her research
Abuse, 8(1), 33–66. focuses on interpersonal violence and influences of victimization on
Williams, S. L., & Frieze, I. H. (2005). Courtship behaviors, rela- women’s health.
tionship violence, and breakup persistence in college men and
women. Psychology of Women Quarterly, 29, 248–257.
Williams, J. R., Ghandour, R. M., & Kub, J. E. (2008). Female per-
petration of violence in heterosexual intimate relationships ado- Jeff R. Temple is an Associate Professor at the University of Texas
lescence through adulthood. Trauma, Violence, & Abuse, 9(4), Medical Branch. He received his doctorate in counseling psychology
227–249. from the University of North Texas. His major research interests
Wolfe, D. A., Scott, K., Reitzel-Jaffe, D., Wekerle, C., Grasley, C., & include the predictors, consequences, and development and evaluation
Straatman, A. L. (2001). Development and validation of the of interventions to reduce intimate partner violence.

You might also like