Prior to 1970, studies focusing on family violence were virtually
nonexistent. The lack of research in this area seemed to imply that
violence within the family was "either rare, dysfunctional or a pathology
traceable to mental illness or psychopathology" (Gelles, 1979, p.169). Yet,
during the past two decades, abuse between intimate partners has become
recognized by social scientists as a serious social problem; impacting on all
levels of society.
A national U.S. survey on family violence was first conducted in 1975 by Straus, Gelles and Steinmetz (1980). This large scale investigation
marked an initial attempt at estimating the prevalence of family violence in
American families. In doing so, it has been responsible for raising our awareness
about the problem of family violence. Since the publication of its
findings, a plethora of literature including empirical and review works in this
area has followed, drawing the attention of policy makers, legislators and service
The approach to studying the problems associated with family violence has changed throughout the course of this literature's development.
Early research into family violence assumed that such behaviour could only be the
result of a deranged mind. Support for this notion was advanced by
society's belief that the family as an institution, is committed to nonviolence
among its members through the maintenance of benevolent and loving
relationships. Since that time, sociologically based research has focused
on establishing the prevalence, correlates, and social patterns of family
violence (Straus et al., 1980). As a result of these studies, we have come
to know that abuse within the family is anything but rare.
Distinguishing Between Male and Female Perpetrated Abuse
For the past two decades, while family violence literature has focused on
the abuse of children, dating partners, spouses, siblings, and the elderly, the
study of spousal violence has become synonymous with the term "wife
abuse" (Sommer, Barnes & Murray, 1992). The reason for this misnomer is
due to almost exclusive focus of research on husband-to-wife abuse because of the
high visibility of females as victims of family violence (Shupe & Stacey,
1987). The shelter movement has also made it possible for researchers to
have a ready made sampling base comprised of women who were willing to provide testimonies of the abuse they endured (Ptachek, 1988). The politics surrounding
the investigation of female perpetrated violence (Steinmetz, 1977) attests to the
controversial nature of this research problem. In a review article on
relationship violence by women, Flynn (1990) elaborated on the politics of family
violence research by stating "feminists....fear that drawing attention to battered husbands will impede attempts to battle the more serious problem of
wife abuse" (p.194). In the report of their findings on family
violence rates over two surveys, Straus and Gelles (1986) stated,
"Violence by wives has not been an object of public concern. There
has been no publicity, and no funds have been invested in ameliorating this
problem because it has not been defined as a problem. In fact, our 1975 study was
criticized for presenting statistics on violence by wives" (p.472).
The conclusion that males are more prone to violence may in fact be erroneous
(Straus, 1988) because of the existing bias in research to investigate male
perpetrated violence and the factors associated with it. Straus's (1988)
conclusion can neither be supported nor disputed till such time as a gender
neutral approach to investigations into partner abuse are conducted. Only
then can the rates, predictors, and outcomes of male and female perpetrated
partner abuse be appropriately assessed and compared.
A consequence of the prevailing bias against research directed at the investigation of violence perpetrated by females is an absence of
intervention programs for female abusers and male abuse victims.
The reasons behind the system's failure to identify female perpetrators
of partner abuse have not been clearly established by empirical studies. It
may be that men are too embarrassed to report being abused, or they fear they
will not be believed if they do.
Currently, studies have not investigated the factors associated
with men's reluctance to come forward. The reasons cited are, for the most
part speculative, and are based on testimonials of small numbers men willing to
relate accounts of abuse.
Due to the reliance upon data derived from crime surveys, police reports,
and medical, civil, and criminal court records, researchers and those involved in
providing social services have rationalized men's greater proneness to violence
by asserting that women are more likely to be seriously injured in domestic
disputes (Wilson, 1990).
However, this assertion, as well, has not been empirically tested
in non-clinical samples of male and female victims of abuse. As a result,
the abuse and injuries sustained by men go either unreported or accounted for by
Recent studies examining violence perpetrated by women against men in courtship,
marital or common-law relationships have nevertheless, suggested that the rate of
female perpetrated partner abuse parallels that of males (Malone, Tyree &
O'Leary, 1989; Marshall & Rose, 1990;
Straus & Gelles, 1986; Thompson, 1991). In spite of this,
the inherent bias favouring research directed at male perpetrated partner abuse
is most evident in policies and intervention programs concerning family violence
(Sommer, Barnes & Murray, 1991).
Defining Partner Abuse
An ongoing methodological concern common to social science research has
been measurement problems associated with inadequate operational definitions of
the variable in question. In order to provide a precise definition of
partner abuse in this research, it has been limited to the occurrence and factors
associated with "physical" partner abuse. Throughout this paper,
the terms "abuse" and "violence" have been used
interchangeably. Unless prefixed by the term "severe", these have
been considered synonymous. Both have been intended to refer to "an
act (or acts) carried out with intention, or perceived intention of causing
physical pain or injury to another person" (Straus & Gelles, 1988,
Kaplan (1988) noted that a problem in defining and measuring abusive
behaviours within intimate relationships has been that many of these behaviours
in their benign forms are considered normal within the context of family
interactions. The Conflict Tactics Scale (CTS) was developed by Murray
Straus (1979) as an objective measure of the class of behaviours deemed to go
beyond what may be considered "normal". In its entirety, the CTS
measures three factorially separate variables: reasoning, verbal aggression and violence or physical aggression (Straus, 1979). The scale's brevity and
ability to be administered in both interview and self-report format has made the
CTS the most widely used measure of intrafamilial violence. Nevertheless,
the CTS has been criticized by feminists.
According to Straus and Gelles (1990), they have charged that this quantitative methodological tool is "inherently male oriented and distorts
the reality of women's lives" (p. 11). In spite of objections with the CTS,
it has demonstrated both construct and content validity as well as provided
adequate reliability (Straus, 1979).
The CTS may be scored in a number of ways in order to provide a maximal
description of violent conduct. According to Straus and Gelles (1988), estimates
of violent acts can be derived through the analysis of individual scale items.
Individual scale items can then be combined in a number of ways to form different
indices. For example, when all items are included together, an overall
index of violence is achieved. This overall index can be further divided
into two other indices to measure "minor" and "severe"
Statement of the Problem
In spite of the progress made in the field of family violence, this area
is fraught by several serious methodological weaknesses and limitations.
Generally speaking, design and statistical strategies employed thus far, have
been primarily aimed at determining the prevalence and correlates of abuse.
A brief discussion of the methodological issues facing family violence
researchers that are relevant to this study follows.
To date, general population based research examining changes in the rates
of family violence have compared the findings of similar, but unrelated studies
conducted at two points in time (Straus & Gelles, 1986). Although
Straus and Gelles (1986) argued that the methodologies and sampling strategies
employed in these studies support the trends demonstrated by the data, it has
also been asserted that changes in rates of family violence are a reflection of a cohort effect rather than an actual change in behaviour (Egley, 1991).
Currently, researchers have demonstrated a number of associations linking
partner abuse to situational and individual factors. For example, studies
that examine alcohol consumption and personality variables as possible risk
factors in domestic violence are evident in family violence literature but have
been limited to small, clinical samples (Beckman, 1978; Fitch & Papantonio,
Frieze & Schafer, 1984). The initiation and consequences
of violent episodes also remain unclear.
Past research has alerted us to the seriousness and the pervasiveness of
this problem. In spite of this, determining what makes some individuals
vulnerable to victimization and places others at risk for perpetration still
remains to be established. Researchers have also been unsuccessful in
finding ways of preventing battering from occurring or stopping it once it has
begun. The lack of general population based research investigating
predictors of partner abuse has also limited policy makers' ability to
effectively address the problems of partner abusers found outside clinical
There is a need for researchers to go beyond present research strategies
through the use of longitudinal data and statistical techniques that can make
causal inferences possible. This investigation was designed to overcome
these limitations by analyzing follow-up data, making comparisons with our
earlier research, and examining the predictive factors related to male and female perpetrated violence.