University of Nebraska - Lincoln University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln
Sociology Department, Faculty Publications Sociology, Department of
2010
Bidirectional, Unidirectional, and Nonviolence: A Comparison of Bidirectional, Unidirectional, and Nonviolence: A Comparison of
the Predictors Among Partnered Young Adults the Predictors Among Partnered Young Adults
Lisa A. Melander
Kansas State University
, lmeland@ksu.edu
HarmoniJoie Noel
University of Nebraska-Lincoln
, hnoel@nebraska.edu
Kimberly A. Tyler
University of Nebraska-Lincoln
, kim@ktresearch.net
Follow this and additional works at: https://digitalcommons.unl.edu/sociologyfacpub
Part of the Sociology Commons
Melander, Lisa A.; Noel, HarmoniJoie; and Tyler, Kimberly A., "Bidirectional, Unidirectional, and
Nonviolence: A Comparison of the Predictors Among Partnered Young Adults" (2010).
Sociology
Department, Faculty Publications
. 128.
https://digitalcommons.unl.edu/sociologyfacpub/128
This Article is brought to you for free and open access by the Sociology, Department of at
DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Sociology Department,
Faculty Publications by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.
Published in Violence and Victims 25:5 (2010), pp. doi: 10.1891/0886-6708.25.5.617
Copyright © 2010 Springer Publishing Company. Used by permission.
Bidirectional, Unidirectional, and Nonviolence:
A Comparison of the Predictors Among
Partnered Young Adults
Lisa A. Melander, PhD
Kansas State University
HarmoniJoie Noel, PhD
Kimberly A. Tyler, PhD
University of Nebraska-Lincoln
Corresponding author — Lisa A. Melander, PhD, Kansas State University, Department of Sociology,
Anthropology, and Social Work, 204 Waters Hall, Manhattan, KS 66506; email lmeland@k-state.edu
Abstract
In order to understand more fully the context and impact of intimate partner violence
(IPV), it is important to make distinctions between different types of relationship ag-
gression. As such, the current study longitudinally examines the differential effects
of childhood, adolescent, and demographic factors on three different partner violence
groups: those who experience bidirectional IPV, those who experience unidirectional
IPV, and those who do not experience either form of IPV. Multinomial logistic regres-
sion results reveal that depressive symptoms and lower partner education predict bi-
directional when compared to unidirectional IPV and nonviolence. In contrast, other
risk factors such as illicit drug use are found to be predictors of unidirectional violence
only, which reveals that the correlates of violence vary depending upon the type of IPV
examined.
Keywords: bidirectional partner violence, risk factors, young adults
S
ince its rise to popularity in the 1960s and 1970s during the second wave of the wom-
en’s movement, topics surrounding various issues of family violence have permeated
both scholarly publications and academic discourse in general (Gelles, 1997). Despite
the relative continuity of violence topics, the focus of these studies, especially with regard
to intimate partner violence (IPV), has changed drastically over the years. For example, the
literature regarding “battered woman syndrome” (Walker, 1979, p. 19) was quickly rebut-
ted with concerned discussions of an equally alarming and a possibly more silent epidemic
of the “battered husband phenomenon” (Steinmetz, 1977-1978, p. 503). More recently, part-
ner violence researchers have been charged with the task of moving in a new direction, to
“make distinctions among types of partner violence” (Johnson & Ferraro, 2000, p. 948).
To understand the nuances of partner violence research, scholars have broadened the fo-
cus of their studies to include various types of violence, considering both males and females as
victims and perpetrators of IPV (Anderson, 2002; Caetano, Vaeth, & Ramisetty-Mikler, 2008;
Prospero, 2008; Williams & Frieze, 2005). One of the more recent trends in the family violence
literature focuses around examining the prevalence and predictors of bidirectional violence,
1
617
618 Melander, noel, & Tyler in Vi o l e n c e a n d Vi c t i m s 25 (2010)
which generally refers to situations in which a respondent reports being both a victim and
perpetrator of violence in the context of an intimate relationship (Caetano, Ramisetty-Mikler,
& Field, 2005; Harned, 2002; Lewis, Travea, & Fremouw, 2002; Robertson & Murachver, 2007;
Straus, 2008; Tyler, Melander, & Noel, 2009). Because few studies have employed measures of
bidirectional IPV, little is known about the predictors that are specic to this form of violent
behavior and how they compare to individuals involved in unidirectionally violent relation-
ships (i.e., only one partner perpetrates violence) and nonviolent couples. Consequently, the
purpose of this article is to examine whether childhood, adolescent, and demographic predic-
tors differ across three partner violence groups: those who experience bidirectional partner vi-
olence, those who experience unidirectional violence, and those who do not experience either
form of violence. Identifying the risk factors associated with both bidirectional and unidirec-
tional violence will not only provide guidance for future research endeavors but may also as-
sist clinicians with designing appropriate treatment and intervention strategies.
LITERATURE REVIEW
Child Maltreatment
Because the family is often considered society’s most violent institution, it is important to
look at the different forms of family violence simultaneously (Gelles, 1997). One of the most
consistent predictors of partner violence is a history of child abuse. Physical abuse, sexual
abuse, and neglect have been found to predict partner violence perpetration and victimiza-
tion (Field & Caetano, 2005; Tyler et al., 2009; Whiteld, Anda, Dube, & Felitti, 2003). As
such, individuals who experience maltreatment within the family of origin may be vulnera-
ble to revictimization at the hands of an intimate partner.
Depressive Symptomology
Although previous research has found that those in violent intimate relationships are more
likely to have depressive symptoms in the aftermath of their IPV experiences (Anderson,
2002; Prospero, 2008; Zlotnick, Johnson, & Kohn, 2006), other studies have explored whether
depressive symptoms occur prior to the violent interactions (Keenan-Miller, Hammen, &
Brennan, 2007). For example, Lipsky, Caetano, Field, and Bazargan (2005) found that de-
pressive symptoms predicted both IPV perpetration and victimization. In contrast, Caetano
and colleagues (2008) found that rates of depression did not signicantly differ among their
perpetrator only, victim only, and mutual violence couples. Because of these contradictory
ndings it is important to consider whether depressive symptomology is a precursor to dif-
ferent types of IPV experiences.
Substance Use
Substance use has also been linked to IPV in both general population and clinical samples
(Caetano et al., 2005; Drapkin, McCrady, Swingle, & Epstein, 2005; Lipsky et al., 2005). Al-
though some researchers nd no differences between mutually violent and perpetrator or
victim-only couples in terms of alcohol use (Caetano et al., 2008; Lewis et al., 2002), oth-
ers report that drinking is a risk factor for bidirectional violence. Males and females who re-
ported binge drinking in the past month were at an increased risk for mutual IPV among a
national sample of cohabiting and married adults (Cunradi, 2007). El-Bassel, Wu, Go, and
Hill (2005) found that although frequent crack and marijuana use increased the likelihood of
BidirecTional, UnidirecTional, and nonviolence 619
subsequent physical and sexual IPV, the ndings were inconclusive for cocaine, heroin, and
frequent binge drinking. In other words, these ndings indicate that the relationship be-
tween frequent drug use and bidirectional IPV may vary by type of drug.
Relationship Status
Relationship status has an impact on being involved in a violent relationship, with cohab-
iters having the highest rates of violence followed by married and dating couples (Magdol,
Moftt, Caspi, & Silva, 1998; Stets & Straus, 1990). Brown and Bulanda (2008) examined the
association between relationship status and IPV perpetration and victimization, and found
that cohabiting women were the most likely to perpetrate or be the victims of IPV followed
by married and dating women, respectively. For men, those in dating relationships were
least likely to perpetrate or be victimized by IPV; cohabiting and married men did not sig-
nicantly differ in their rates of perpetration or victimization. As such, accounting for rela-
tionship status may be an important factor for bidirectional violence.
Demographic Characteristics
There have been mixed ndings with regard to whether males or females are more likely to
be perpetrators and/or victims of partner violence. Several researchers have found that fe-
males are victimized more often by an intimate partner than their male counterparts (Cat-
alano, 2007; Gover, Kaukinen, & Fox, 2008). According to Rennison and Welchans (2000),
women are victimized by intimates at approximately ve times the rate of men. Others (Ca-
paldi & Owen, 2001; Williams & Frieze, 2005), however, report that women victimize men
more often. For example, in their nationally representative sample of adolescents, Whita-
ker, Haileyesus, Swahn, and Saltzman (2007) found that women reported more partner vi-
olence perpetration and victimization than men. Alternatively, some researchers nd that
men and women use approximately equal levels of violence toward one another and re-
port similar levels of victimization (Anderson, 2002; Harned, 2002; Robertson & Murachver,
2007; Straus, 2008). Cunradi (2007), for example, found that approximately the same propor-
tion of men (3.1%) and women (3.2%) reported experiencing mutual IPV.
Other demographic factors that have been linked to IPV perpetration and victimiza-
tion include age, racial and ethnic background, and sociodemographic status. In general,
younger individuals are at higher risk for both perpetrating and becoming victims of IPV
(Anderson, 2002; Cunradi, 2007; Gelles, 1997; Prospero, 2008) with females aged 20 to 24
having the highest risk of nonfatal IPV (Catalano, 2007). In terms of race and ethnicity, Field
and Caetano (2005) found that African American and Hispanic couples reported higher
frequencies of mutual physical partner violence than their White counterparts. Similarly,
Weston, Temple, and Marshall (2005) found that African American women in mutually vi-
olent relationships experienced signicantly more sexual and physical violence than their
Euro-American and Mexican American counterparts. Finally, those with lower socioeco-
nomic attainment, as measured by the educational level of the respondents (Drapkin et al.,
2005; Frias & Angel, 2005), their parents (Foshee et al., 2008), or their partners (Brown & Bu-
landa, 2008), are at increased risk for partner violence.
THEORETICAL PERSPECTIVE: JOHNSON’S IPV TYPOLOGY
Johnson (1995, 2006) created a typology of violent couples that has been used by other social
science researchers to provide context to contradictory ndings in the dating violence litera-
620 Melander, noel, & Tyler in Vi o l e n c e a n d Vi c t i m s 25 (2010)
ture (Anderson, 2002; Prospero, 2008). In a series of articles beginning in the 1990s, Johnson
has revised and expanded his typology of violence, which is based on the degrees of con-
trol and violence that are present in the relationship to include the following categories: in-
timate terrorism, mutual violent control, violent resistance, and situational couple violence.
Previously referred to as patriarchal terrorism, intimate terrorism refers to a form of “terror-
istic control” whereby one partner systematically uses violence as well as a broad range of
other power and control tactics (Johnson, 1995, p. 284). In this way, intimate terrorism can
be understood as a general attempt to control an intimate partner by any means necessary
and is more likely to escalate over time compared to other forms of violence (Johnson & Fer-
raro, 2000; Johnson & Leone, 2005). Although little is known about this rare form of violence
(Johnson & Ferraro, 2000), mutual violent control occurs when both partners are violent and
controlling toward each other. Violent resistance, on the other hand, occurs when one part-
ner is violent and controlling and the other partner responds with violence in a manner
akin to self-defense. Finally, situational couple violence, which was previously referred to
as common couple violence (Johnson, 1995), occurs when either one or both partners are vi-
olent but is not typically connected to a general pattern of control (Johnson, 2006; Johnson &
Ferraro, 2000). This form of aggression usually occurs in the context of a particular situation
in which conict occasionally “gets out of hand” and rarely escalates to severe, life-threat-
ening violence (Johnson, 1995, p. 285). Although some researchers have not found support
for using these categories of relationship violence (Anderson, 2008), this typology postu-
lates that there are important differences between unidirectional and bidirectional violence.
As such, it is important to learn more about these distinct forms of aggression to inform pre-
vention and intervention efforts.
THE CURRENT STUDY
Despite the relatively large body of literature on IPV, there have been few studies that have
simultaneously addressed the relationship between child maltreatment experiences, inter-
nalizing behaviors, and personal characteristics and bidirectional violence. Furthermore, re-
searchers have not determined whether the predictors of IPV vary across different partner
violence groups. As such, it is largely unknown whether factors that are usually linked to uni-
directional violence are also associated with bidirectional IPV. Existing studies also generally
utilize cross-sectional data, which makes it difcult to establish temporal order. The current
study addresses each of these shortcomings by examining the following research question us-
ing a nationally representative, longitudinal sample: Do childhood, adolescent, and demo-
graphic factors (e.g., child abuse, substance use, and relationship status) vary between bidirec-
tional, unidirectional, and nonviolent relationships in young adulthood?
METHOD
Data
The analyses are based on Waves I and III of the National Longitudinal Study of Adolescent
Health (Add Health). The data are from the restricted access core sample of over 15,000 in-
dividuals who were rst interviewed in grades 7 through 12 and then completed a Wave III
in-home interview in young adulthood (18-27 years old; see Udry, 1997 for a more detailed
description of the methodological structure of this data set). The analytic sample was re-