You are on page 1of 21

Sex Roles, Vol. 52, Nos.

11/12, June 2005 (


C 2005)

DOI: 10.1007/s11199-005-4201-0

Understanding Gender and Intimate Partner Violence


from an International Perspective

Barbara Krahé,1,2 Steffen Bieneck,1 and Ingrid Möller1

This paper reviews the international literature on intimate partner violence with a focus on
gender differences in perpetration and victimization rates. A total of 35 studies from 21 coun-
tries are discussed that report prevalence or incidence rates of men’s and women’s involve-
ment in physical and/or sexual aggression against an intimate partner. In addition, evidence
on risk factors as well as consequences of intimate partner violence for men and women is
presented. Conceptual and methodological differences between the studies and the lack of
comparable databases within countries are discussed as limitations of the evidence, and per-
spectives for future research are outlined in the framework of cross-cultural psychology.

KEY WORDS: intimate partner violence; gender differences; cross-cultural psychology.

Studies on intimate partner violence outside lished by the World Health Organization in 2002,
North America are few and far between, and studies acknowledges the possibility that women can be
examining gender differences in the perpetration or violent toward their partners. However, when re-
victimization by intimate violence in relationships viewing prevalence, risk factors, and consequences
are even rarer. The debate about equal or unequal of intimate partner violence worldwide, only men’s
involvement of men and women in violence against violence against women is considered because “the
an intimate partner has taken place almost exclu- overwhelming burden of partner violence is borne
sively on the basis of North American evidence by women at the hands of men” (Krug, Dahlberg,
(Currie, 1998; Dasgupta, 2002; Dobash & Dobash, Mercy, Zwi, & Lozano, 2002, p. 89).
2004; Kimmel, 2002; Straus & Gelles, 1990; White The present paper reviews evidence on both
& Koss, 1991). In Archer’s (2000) meta-analysis of physical and sexual partner violence from sources
sex differences in aggression between heterosexual outside North America. Physical partner violence
partners, no more than 7 out of the total 82 studies in- refers to actions based on the intention to inflict phys-
cluded were conducted outside North America. With ical harm on an intimate partner. Sexual violence
regard to women’s use of sexual aggression against refers to the use or threat of physical violence or
male partners, the most comprehensive collection of the exploitation of a partner’s inability to offer re-
studies so far does not contain any data from outside sistance to force her or him to engage in sexual acts.
North America (Anderson & Struckman-Johnson, More broadly, sexual aggression includes all behav-
1998). Across the world, the dominant perspective iors carried out with the intention to make a partner
on intimate violence casts men into the role of engage in sexual activity against her or his will (e.g.,
perpetrators and women into the role of victims. Krahé, Waizenhöfer, & Möller, 2003).
The “World Report on Violence and Health,” pub- The focus will be on studies that provide evi-
dence (a) comparing levels of men’s and women’s
1 University
violence against an intimate partner, or (b) looking
of Potsdam, Germany.
2 Towhom correspondence should be addressed at Department of
at gender differences in victimization by relationship
Psychology, University of Potsdam, Postfach 60 15 53, D-14415 violence. Bond (2004) has noted that cultures differ
Potsdam, Germany; e-mail: krahe@rz.uni-potsdam.de. in terms of constraints and affordances that impinge

807 0360-0025/05/0600-0807/0 
C 2005 Springer Science+Business Media, Inc.
808 Krahé, Bieneck, and Möller

on their members and channel their behavior into partner violence: the Conflict Tactics Scales (CTS,
similar patterns. In line with this view, intimate Straus, 1979; CTS2, Straus, Hamby, Boney-McCoy,
partner violence can be regarded as shaped by & Sugarman, 1996) providing a measure of physi-
socioeconomic variables, gender roles and the cal aggression in relationships, and the Sexual Ex-
normative acceptance of aggression that are char- periences Survey (SES; Koss, Gidycz, & Wisniewski,
acteristic of a given society (Vandello & Cohen, 1987; Koss & Oros, 1982) measuring sexual aggres-
2003). Indeed, feminist accounts of partner violence sion. These instruments provide behavioral descrip-
highlight the causal role of patriarchical power struc- tions of various forms of physical or sexual aggres-
tures in explaining men’s violence against women sion (rather than value-laden labels, such as “rape”).
(e.g., Marin & Russo, 1999), and historical analyses They are typically administered in a written for-
have shown that the very conceptualization of forced mat, usually under anonymous conditions, and can be
sexual intercourse as an offence against women’s used to obtain data both from the perpetrator and the
right to sexual self-determination is a relatively victim perspective. The CTS (or the slightly modified
recent development (Brownmiller, 1975). Analyses and extended version of the CTS2) and the SES have
of partner violence across cultures deal with a similar been widely used in international studies, as will be
relativity: whether or not a particular behavior, such apparent from the studies presented later. However,
as physically assaulting a woman for resisting her it is not always clear how carefully they were trans-
partner’s demands, is considered partner aggression lated or adapted for use in a foreign language (e.g.,
in a society depends to a significant degree on whether or not standard back translation procedures
socially shared norms concerning gender relations were used) and whether or not the reliability of the
and the acceptance of violence (Levinson, 1989). new versions was established (e.g., Krahé, Reimer,
It should be clarified at the outset that the Scheinberger-Olwig, & Fritsche, 1999). Besides the
present review does not aspire to provide a compar- CTS and SES, many ad hoc instruments and sets
ative cross-cultural analysis of intimate partner vio- of questions have been used to measure both phys-
lence or, indeed, the role of gender in the victimiza- ical and sexual aggression, their number and format
tion by and/or perpetration of partner violence. Such often prescribed by the particular methods of data
an endeavor is precluded (a) by the fragmented na- collection and the time and space constraints associ-
ture of the available database, (b) by methodological ated with them. Regardless of the instruments used,
problems that undermine the comparability of differ- an important methodological distinction refers to the
ent studies, such as differences in the definition and time period for which reports of partner violence are
measurement of partner violence or sample compo- elicited. In prevalence studies, respondents are asked
sition, and (c) by a lack of evidence linking data on to indicate whether or not they have ever experi-
partner violence to cultural parameters. These con- enced or committed acts of partner violence (in their
straints and limitations will be discussed in more de- lifetime, since adulthood, since getting married). In
tail in the final section of the paper. Instead, the aim contrast, incidence studies collect information about
of this review is a more modest one, namely to sum- whether a person has experienced or committed part-
marize the available evidence on the scale of partner ner violence in a defined period of time, such as the
violence and the role of gender differences in coun- last 12 months. It is obvious that this difference in
tries outside North America. the temporal frame of reference must be taken into
account when interpreting the findings.
Another important aspect to bear in mind when
THE MEASUREMENT OF INTIMATE looking at rates of partner violence is the pop-
PARTNER VIOLENCE ulation from which respondents are drawn. Com-
munity samples comprise members of the general
One of the difficulties in assessing the scope of public that can be drawn on the basis of random
partner violence, both within and across countries, sampling, representative stratification, or availability
is the heterogeneity of instruments used to measure (“convenience samples”). In contrast, clinical sam-
violent behavior against a partner. Methods vary in ples are composed of participants who represent a
terms of what questions are asked and how they risk group for intimate partner violence and/or seek
are asked, and findings are invariably affected by some form of professional help from counseling or
these differences. Nonetheless, two research instru- medical agencies. The two ways of sampling yield dif-
ments stand out as the most prominent measures of ferent results. Findings based on community samples
Intimate Partner Violence—An International Perspective 809

generally underestimate the problem, whereas fig- Guinea (69 and 67%, respectively, of ever mar-
ures from clinical samples can overestimate the ex- ried/partnered women reporting physical assault by
tent of intimate partner violence in the general pop- a partner). The lowest prevalence figures came from
ulation. In addition, the two methods may uncover Paraguay (10% of ever married/partnered women)
distinct forms of intimate partner violence, differing and the Philippines (10% in a sample of pregnant
both in terms of severity and in terms of the under- women). Twelve-months incidence was highest in the
lying motivational dynamics (cf. Johnson’s, 1995, dis- West Bank and Gaza Strip, Palestine (52% of cur-
tinction between common couple violence and patri- rently married/partnered women) and lowest in the
archal terrorism). However, both ways of sampling U.S. (1.3% of all women, including those who had
are important in order to investigate different facets never been in a relationship). However, as Krug
of violence in intimate partnerships. et al. (2002) acknowledge, the methodological dif-
ferences preclude any comparison of the absolute
incidence and prevalence rates across the different
PREVIOUS REVIEWS countries.
The other two reviews are restricted in scope
Three recent reviews were located in the inter- to European countries. Hagemann-White (2001) re-
national literature that provided information across viewed 11 studies from 9 European countries that de-
different countries about women’s victimization by scribed the level of physical and/or sexual violence in
male partner violence (Hagemann-White, 2001; Krug intimate relationships in community samples. As in
et al., 2002; Kury, Chouaf, Obergfell-Fuchs, & the Krug et al. (2002) review, we see a wide varia-
Woessner, 2004). They are relevant to the present re- tion in prevalence rates between 8% (Denmark; past
view in that they focus on the traditional perspective partner) and 45% (Finland; past cohabitation part-
of female victims and male perpetrators, providing a ner), attributable to a large extent to differences in
background against which findings concerning male methodology (Hagemann-White, 2001, p. 733). Dis-
victimization and female perpetration as well as gen- cussing the problems involved in defining and mea-
der differences can be interpreted. suring intimate partner violence in a way that re-
The worldwide review by Krug et al. (2002) for flects gendered representations of violence, she calls
the World Health Organization presented reports for “developing standards of measurement that re-
from women on physical victimization by a partner spect subjective meanings and thereby admit a plu-
from 32 countries (5 African, 11 Latin American rality of cultures and customs, rather than imposing a
and Caribbean, 6 European and 4 Eastern Mediter- standard template on the experience of what is or is
ranean, 8 from the Asian and Western Pacific region, not violence” (Hagemann-White, 2001, p. 755). Kury
along with Canada and the United States; Krug et al., et al. (2004) examined European prevalence rates
2002, Table 4.1) In addition, two studies were in- for partner violence from six studies, three of which
cluded from India and Thailand on men’s reported had already been included in the Hagemann-White
use of violence against a female partner. This com- (2001) review. Two of the new studies, from Spain
pilation reflects, first and foremost, vast differences and Sweden, only included women. In the Swedish
in methodology: sample size (ranging from N = 111 survey, prevalence rates for partner violence varied
in Norway to N = 89,199 in India), sample type between 8 and 20% in relation to respondents’ age.
(all women, married women only, pregnant women In the Spanish study, 12.9% of the surveyed women
only), time period covered (lifetime, last 12 months, reported physical or severe physical abuse from a
last 3 months), and age range of respondents. The au- partner during the last 12 months, and 16.2% re-
thors do not report the type of instrument or method ported sexual abuse or severe sexual abuse (Medina-
of data collection used in the individual studies, but Ariza & Barberet, 2003). Kury et al. (2004) discuss
there is every reason to assume that there was also the findings from their review in the context of so-
wide variation in that respect. cial and socioeconomic changes in Europe over the
Apart from showing that intimate partner vio- last decades, such as rising poverty rates that they re-
lence does occur on a substantial scale across the gard as conducive to an increase in family violence
world, the findings cannot do more than delin- generally and intimate partner violence in particular.
eate the upper and lower boundaries of the prob- However, in the absence of studies linking changes in
lem. For lifetime prevalence, the highest figures partner violence rates to economic or social parame-
reported came from Nicaragua and Papua New ters, these explanations remain tentative.
810 Krahé, Bieneck, and Möller

PRESENT SAMPLE OF STUDIES cal aggression”). Other studies discuss psychological


abuse as a potential risk factor for physical violence
Studies were included in the present review if (e.g. Bradley, Smith, Long, & O’Dowd, 2002). We
they covered samples outside North America and will look at the co-morbidity of different forms of in-
provided primary quantitative data on the incidence timate partner violence after reviewing the evidence
or prevalence of physical and/or sexual partner vi- on physical and sexual violence.
olence from one or more of the following perspec- From the perspective of gender differences, it
tives: (a) female victimization, (b) male perpetration, would have been interesting to include information
(c) female perpetration, and (d) male victimization. about intimate violence in gay or lesbian relation-
Studies limited to the traditional perspective of fe- ships and relate it to men’s and women’s involvement
male victimization by male partners were included in partner violence in heterosexual relationships.
only if they came from Non-European countries or— Unfortunately, the number of studies on same-sex
in the case of European studies—have not been cov- violence was too small for this purpose (see, how-
ered in the previous reviews by Hagemann-White ever, Krahé, Schütze, Fritsche, & Waizenhöfer,
(2001) and Kury et al. (2004). A total of 35 stud- 2000; Toro-Alfonso & Rodriguez-Madera, 2004).
ies from 21 countries were identified on the basis of Therefore, the present review is limited to partner
these criteria. With three exceptions (BMFSFJ, 2004; violence in heterosexual relationships.
Luedtke & Lamnek, 2002; Mirrlees-Black, 1999), Of the 35 studies included, 14 provide preva-
all studies were published in academic journals or lence data on intimate partner violence, 15 present
books. incidence data, and 6 provide both prevalence and in-
As mentioned earlier, it is difficult to evalu- cidence data. To make the figures more easily inter-
ate and compare studies that used different mea- pretable, prevalence and incidence findings are pre-
sures of intimate partner violence. Studies that ex- sented separately. Table I summarizes the prevalence
amined physical aggression using the Conflict Tactics findings, Table II summarizes the incidence findings.
Scales should be comparable with each other. How- Results from studies providing both prevalence and
ever, the findings are not always based on the orig- incidence rates are split up and feature separately in
inal CTS items but on modified versions. Compar- the respective tables.
isons between studies looking at sexual aggression The majority of studies (N = 29) included
are even more complicated. Haj-Yahia (2000a) used community samples, only 6 studies involved clinical
the CTS2, whereas the surveys by Krahé (Krahé, samples (patients presenting at emergency depart-
1998; Krahé & Scheinberger-Olwig, 2002) developed ments or counseling agencies, in one instance men
German versions of the SES (Krahé et al., 1999). convicted of intimate partner violence). Nineteen
The remaining studies developed their own ques- studies included only female respondents, 14 studies
tionnaires and interviews, sometimes based on focus included men and women, and 2 studies included all-
group discussions to establish the cultural meaning male samples. Eighteen studies examined physical
of intimate partner violence (e.g., Deyessa, Kassaye, violence only, 6 studies looked at sexual aggression
Demeke, & Taffa, 1998). only, and 11 studies reported both physical and sex-
The focus of our review is on physical and sex- ual aggression (some of these also reported verbal or
ual violence in relationships. Nevertheless, it is im- psychological aggression, which was not in the focus
portant to recognize that relationship aggression can of this review and is therefore not included in the
take the form of psychological abuse, also termed tables). About half of the studies (N = 19) used the
verbal or emotional aggression. Psychological abuse Conflict Tactics Scales (CTS or CTS2), sometimes
includes insults and humiliation or controlling be- with modifications. Five studies used the SES in the
havior by the partner. It is widely recognized that original version or in modification. The remaining
physical aggression is almost always accompanied by studies employed a variety of instruments loosely
psychological abuse, which may have equal or even connected to the CTS or designed on the basis of
more lasting emotional effects than physical aggres- piloting in the respective country.
sion. Some of the reviewed studies included psycho- As Tables I and II show, there is a considerable
logical abuse, but only a few reported prevalence heterogeneity between the studies in methodology
rates (e.g. Tang, 1999a, 1999b, for “verbal abuse”; and findings. To impose some structure on the ma-
Eisikovits, Winstock, & Fishman, 2004, for “shouting terial that may make it easier to arrive at conclusions
and humiliation”; Haj-Yahia, 2000a, for “psychologi- about this diverse body of evidence, the discussion
Table I. Prevalence Rates for Intimate Partner Violence (IPV)
Prevalence rates for
Measure of IPV Male Female Male Female
Author (Year) Country Sample and timeframe perpetration perpetration victimization victimization
Arscott-Mills Jamaica Clinical sample of 187 25-item questionnaire 90% physical assault,
(2001) women attending (experiences of violence, 59% sexual assault
Women’s Crisis Centre type of injury); lifetime
for IPV (18–59 years)
BMFSFJ Germany Representative sample of CTS, items from Sexual 23% physical violence,
(2004) 10,624 women (16–85 Experiences Survey (SES), 7% sexual violence
years) lifetime
Bradley et al. Ireland Community sample of 20-item domestic violence 39% violent behavior
(2002) 1,692 women with questionnaire (similar to (of these: 46%
relationship experience CTS); lifetime injured)
attending general
practice (16–84 years)
Carrado et al. UK Representative sample of 5 items on physical assault 10% physical 11% physical 18% physical 13% physical assault
(1996) 894 men and 971 derived from CTS; lifetime assault assault assault ever, ever, 5% in current
women (over 15 years) ever ever 11% in current relationship
relationship
Castro et al. Mexico Community sample of 12 items on physical violence, 12.1% physical
(2003) 914 pregnant women 3 items on sexual violence violence, 10.0%
Intimate Partner Violence—An International Perspective

attending prenatal based on Index of Spouse sexual violence


health clinics (mean Abuse and Severity of
age = 25; SD = 5.5) Violence Against Women
Scale; lifetime before
pregnancy
Deyessa et al. Ethiopia Representative sample of 6-item questionnaire; since 45% physical abuse; of
(1998) 673 married women getting married these: 76% slapped
(mean age = 35.6; with fist, 53% kicked
SD = 10.6) or hit, 34% hit or
beaten
Ellsberg et al. Nicaragua Representative sample of CTS, sexual abuse by partner; 52% any physical
(2000) 360 women ever lifetime violence, 31%
married or living with a physical violence
partner (15–49 years) during pregnancy,
21% sexual abuse
Grimstad Norway Community sample of CTS; lifetime 15% moderate physical
et al. 174 new mothers (age violence, 5% severe
(1999) not given) physical violence
Jackson et al. New Zealand Community sample of Questionnaire regarding 67.4% unwanted 76.9% unwanted sexual
(2000) 200 female and 173 violence in dating sexual activity, activity, 21.0%
male high school relationships; lifetime 19.0% physical physical violence
students (16–20 years) violence
811
Table I. Continued.
Krahé (1998) Germany Community sample of Sexual Experiences Survey 44.7% any 54.5% any sexual
812

adolescents, 197 men (SES); lifetime sexual victimization, 16.8%


(M = 19.2 years) and aggression, through physical
194 women (mean 14.7% force
age = 17.7 years) through
physical
force
Krahé and Germany Community sample of SES; lifetime 21.6% any 30.3% any sexual
Scheinberger- 456 women (M = 18.4 sexual victimization, 3.3%
Olwig years) and 524 men aggression, rape through
(2002) (mean age = 18.7 3.6% rape exploitation, 1.8%
years) through rape through
exploita- physical force
tion, 0.2 %
rape
through
physical
force
Krahé, Germany Study 1: community Men’s Non-consensual Sexual Study 1: 11.7%
Scheinberger- sample of 247 men Experiences Survey sexual
Olwig, & (mean age = 18.3 (adaptation of the SES); aggression;
Bieneck years); Study 2: lifetime Study 2: 17.6%
(2003) community sample of sexual
153 men (mean age = aggression
22.3 years)
Krahé, Germany Community sample of Women’s Sexual Aggression 6.0% sexual
Waizenhöfer, 248 women (mean Survey (adaptation of the aggression
& Möller age = 18.1 years) SES); lifetime
(2003)
Kury et al. Germany Student sample of 309 SES; lifetime 3.6% mild sexual
(2004) women, (mean age = victimization, 33.7%
23 years) moderate
victimization, 1.6%
severe victimization
Luedtke and Germany Random sample of 1,236 Telephone interviews; lifetime 6.0% physical 2.7% physical violence
Lamnek households, (age not violence
(2002) given)
Maman et al. Tanzania Sample of 245 women CTS; lifetime total sample: 37.6%
(2002) attending health clinics physical abuse,
for HIV test (18–55 16.3% sexual abuse;
years) HIV negative
(N = 172): 28.3%
physical violence,
23.0% sexual
violence, HIV
positive (N = 73):
52.2% physical
violence, 44.1%
Krahé, Bieneck, and Möller

sexual violence
Mirrlees-Black UK Representative sample of Crime survey; lifetime 14.9% physical 22.7% physical assault
(1999) 6,098 women and 5,146 assault
men (16–59 years)
Robbé et al. Australia Clinical sample of 522 Items from CTS; lifetime 8.5% total 19.3% total physical,
(1996) female and 475 male physical, sexual, or emotional
patients attending sexual, or abuse, of these:
hospital emergency emotional 17.5% weapon used,
services (age > 16 abuse, of 28.7% sexual abuse
years) these: 11.9%
weapon used,
10% sexual
abuse
Roberts et al. Australia Clinical sample of 553 CTS; lifetime 4.1% total 16.1% total physical,
(1996) female and 670 male physical, sexual, or emotional
patients attending sexual, or abuse: of these:
hospital emergency emotional 21.1% weapon used,
services (age > 16 abuse; of 28.8% sexual abuse
years) these: 36.8%
weapon used,
12.3% sexual
abuse
Römkens (1997) The Netherlands Representative sample of Semi-structured interviews; 10.6% 20.8% unilateral
1,016 women (20–60 lifetime unilateral violence, 9% mild
years) violence violence, 4.8%
Intimate Partner Violence—An International Perspective

moderate violence,
6.3% severe or very
severe violence;
7.4% sexual force,

Note. Studies given in the table are marked with asterisk in the reference list.
813
Table II. Incidence Rates for Intimate Partner Violence (IPV)
814

Prevalence rates for


Measure of IPV Male Female Male Female
Author (Year) Country Sample and timeframe perpetration perpetration victimization victimization
Castro et al. Mexico Community sample of 12 items on physical violence, 10.6% physical
(2003) 914 pregnant women 3 items on sexual violence violence, 8.1% sexual
attending prenatal based on Index of Spouse violence
health clinics (mean Abuse and Severity of
age = 25; SD = 5.5) Violence Against Women
Scale; during pregnancy
Deyessa et al. Ethiopia Representative sample of ad hoc questionnaire, 10% physical abuse; of
(1998) 673 married women last 3 months these: 60% slapped
(mean age= 35.6; with fist, 52% kicked
SD = 10.6) or hit, 29% hit or
beaten
Dobash and UK Clinical sample of 95 men Global frequency of violent 69.5% 53.7% 60% physical 78.9% physical
Dobash convicted of IPV and acts, 9 specific forms of IPV, physical physical violence, of violence, of these
(2004) their female partners similar to CTS; violence; of violence, of these 20% 88.4% slapped,
(total N = 190) (age last 12 months these: these: slapped, 78.9% kicked body
not given) 82.1% 29.5% 17.9% kicked
slapped, slapped, body
54% 23.2%
kicked kicked
body body
Eisikovits Israel Representative sample of Questions referring to physical 6% all physical
et al. 2,544 women living violence (breaking things, violence, 1.7%
(2004) with a partner or moderate violence, severe severe physical
having lived with a violence); last 12 months violence, 6.1%
partner for at least 1 threats of violence
year in the last 5 years
(mean age = 43.1;
SD = 16.0)
Ellsberg et al. Nicaragua Representative 360 CTS; last 12 months 27% any physical
(2000) women ever married or violence, 20% severe
living with a partner physical violence
(15–49 years)
Grimstad Norway Community sample of CTS; during pregnancy 3% moderate physical
et al. 174 new mothers (age violence, 1% severe
(1999) not given) physical violence,
Hadi (2000) Bangladesh Random sample of 500 Structured interview about 26.8% sexual violence
married women (mean forced sex by husband; last
age = 30.5) 12 months
Krahé, Bieneck, and Möller
Haj-Yahia Israel Random sample of 1,111 CTS2; engagement period 19% physical violence,
(2000a) engaged Arab women (maximum 2 years) 13% sexual coercion
(16–38 years)
Haj-Yahia Palestine Study 1: random sample Questionnaire consisting of First (second) survey:
(2000b) of 2410 married women items from: CTS 52% (54%) physical
(17–65 years); Study 2: Psychological Maltreatment violence, 37.6%
random sample of of Women Inventory (40%) sexual abuse
1,334 married women Measure of Wife Abuse
(17–69 years) Index of Spouse Abuse
Abusive Behavior
Inventory; last 12 months
Kim and Korea Representative sample of CTS; last 12 months 27.8% total 15.8% total
Emery 1,279 married or violence violence
(2003) cohabitation men (>20 27.4% 15.5% minor,
years) minor, 2.8% severe
8.0%
severe
Kim and Korea Random sample of 88 CTS; last 12 months 20.8% total 6.3% total
Sung men and 56 women physical physical
(2001) (>60 years) violence, violence,
20.2% 6.3%
minor minor,
5.6% 1.4 %
severe severe
Magdol et al. New Zealand Representative sample of CTS; last 12 months 21.8% any 37.2% any 34.1% any 27.1% any physical
Intimate Partner Violence—An International Perspective

(1997) 425 women and 436 physical physical physical violence, 26.1%
men (20–21 years) violence, violence, violence, minor, 12.7% severe
21.8% 35.8% 31.8% minor,
minor: minor, 21.2% severe
5.7% 18.6%
severe severe
Maman et al. Tanzania Sample of 245 women CTS; last 3 months Total sample: 11.4%
(2002) attending health clinics physical abuse, HIV
for HIV tests negative (N = 172),
(18–55 years) 9.3% physical
violence, HIV
positive (N = 73),
16.4% physical
Maziak Syria Community sample of Repeated beating of any form 23% physical violence,
(2002) 412 women (no age at least three times; last all women, 26%
provided) 12 months married women
Mirrlees- UK Representative sample of Crime survey; last 12 months 4.2% partner’s 4.2% partner’s use of
Black 6,098 women and 5,146 use of force force
(1999) men (16–59 years)
Piispa (2004) Finland Representative sample of Modified CTS; in current 7.5% physical violence,
4,955 women partnership during the last of these: 37% less
(18–74 years) 5 years severe violence, 41%
more severe violence
815
816

Table II. continued.


Romito, Italy Clinical sample of 510 Questionnaire on physical, 18.2% physical or
Saurel- women attending sexual, psychological sexual violence by
Cubizolles, emergency violence, last 12 months current or former
and Crisma departments, advisory partner
(2001) clinics and social
service centers
(18–92 years)
Swart et al. South Africa Community sample of CTS2; last 12 months 35.3% 43.5% 37.8% physical 41.7% physical
(2002) 434 male and 494 physical physical violence, 25% violence, 18.9% with
female secondary violence, violence, with injury injury
school students 26.8% with 16.3% with
(13–23 years) injury injury
Tang (1999a) Hong Kong Random sample of 518 CTS; last 12 months 7.6% minor 11.1% minor violence,
married Chinese men violence, 1.8% 1.5% severe violence
and 752 married severe
Chinese women violence
(18–79 years)
Tang (1999b) Hong Kong Random sample of 1,132 CTS; last 12 months 10% all physical
Chinese married violence, 9.8% minor
women (18–75 years) violence, 1.4%
severe violence
Wetzels et al. Germany Representative sample, CTS; last 5 years 4.9% physical 6.5% physical violence
(1995) 1,902 men and 2,104 violence
women, >16 years

Note. Studies given in the table are marked with asterisk in the reference list.
Krahé, Bieneck, and Möller
Intimate Partner Violence—An International Perspective 817

of prevalence and incidence rates is organized into It is impossible to derive comparative conclu-
three sections: First, we will briefly discuss the ev- sions about the scale of women’s victimization in re-
idence concerning women’s victimization by male lationships from these data due to methodological
partners. Even though our focus is on gender differ- differences in sampling procedures, measurement,
ences in intimate partner violence, these studies are data collection, and analysis. However, it should be
relevant to this review because they contribute to the noted that the majority of studies used relatively
internationalization of domestic violence research. large sample sizes, and many relied on random or
Secondly, we will discuss the studies that compare stratified sampling procedures to enhance the qual-
victimization rates by men and women. Finally, we ity of the data. Therefore, despite the limits in terms
will look at the small set of studies comparing men’s of comparability, they represent an important step
and women’s perpetration rates of partner violence. towards accumulating an international database on
which future studies can build.
Women’s Victimization by Intimate Partners

Thirty-three of the 35 studies included in this re- Studies Comparing Women’s


view present evidence on the rate of women’s expe- and Men’s Victimization Rates
rience of physical violence from an intimate partner,
19 are confined to this perspective. Not counting the Twelve of the 35 studies provided comparative
clinical sample of women in Jamaica seeking help data on the rate of women’s and men’s victimiza-
for intimate partner violence (Arscott-Mills, 2001), tion by intimate partner violence. To facilitate the in-
the highest prevalence rates were reported from terpretation of their main findings, these studies are
Nicaragua (52%; Ellsberg, Pena, Herrera, Lilestrand, summarized in Table III. Again, the first impression
& Winkvist, 2000) and Ethiopia (45%, Deyessa et al., is that of a substantial variation in both prevalence
1998). The lowest prevalence rates were reported and incidence rates across the studies. In terms of the
from a German (2.7%; Luedtke & Lamnek, 2002) debate on gender symmetries or asymmetries in vic-
and a Mexican sample (12.1%, Castro, Peek-Asa, timization, the results are also mixed. Three studies
& Ruiz, 2003). The highest incidence rates were
reported from Palestine (52 and 54%; Haj-Yahia,
Table III. Comparative Studies of Victimization by Women and
2000b) and South Africa (41.7%; Swart, Stevens, & Men Through Physical Partner Violence
Ricardo, 2002).
Physical violence
Data about women’s sexual victimization are victimization (%)
provided by 16 of the 35 studies. Figures vary con-
siderably across different samples. Lifetime preva- Women Men Study∗
lence rates range from 7% in a German sample Prevalence
(BMFSFJ, 2004) and 10% in the Mexican sample Australia 19.3a 8.5a Robbé et al. (1996)
Australia 16.1a 4.1a Roberts et al. (1996)
(Castro et al., 2003) to 76.9% in the New Zealand
Germany 2.7 6.0 Luedtke and
sample (Jackson, Cram, & Seymour, 2000). How- Lamnek (2002)
ever, the latter study used a broad definition of sexual New Zealand 21 19 Jackson et al. (2000)
victimization (“unwanted sexual activity”), whereas The Netherlands 20.8 7.4 Römkens (1997)
the former presented three specific items referring UK 13 18 Carrado et al. (1996)
UK 22.7 14.9 Mirrless-Black (1999)
to verbal pressure and use of force. Clinical stud-
Incidence
ies found lifetime prevalences from 44% for HIV- Germany 6.5 4.9 Wetzels et al. (1995)
positive women from Tanzania (Maman et al., 2002) Hong Kong 12.6 9.4 Tang (1999a)
to 59% for Jamaican women attending a Women’s New Zealand 27.1 34.1 Magdol et al. (1997)
Crisis Center (Arscott-Mills, 2001). As expected, in- South Africa 41.7 37.8 Swart et al. (2002)
UK 78.9b 60b Dobash and
cidence rates are somewhat lower, but also display a
Dobash (2004)
substantial range in magnitude. The lowest rate was UK 4.2 4.2 Mirrlees-Black (1999)
reported for the Mexican sample (8.1%; Castro et al.,
∗ More detailed information about the studies is provided in
2003), followed by Israel (16%; Haj-Yahia, 2000a),
Tables I and II.
Bangladesh (26.8%; Hadi, 2000) and Palestine (37.6 a Combined figures for any physical, sexual, or emotional abuse.
and 40% from the two surveys conducted by bSample consisted of men convicted of intimate partner violence

Haj-Yahia, 2000b). and their female partners.


818 Krahé, Bieneck, and Möller

showed substantially higher rates for women’s vic- Table IV. Comparative Studies of Perpetration of Intimate
timization compared to men’s: the two Australian Partner Violence by Men and Women
studies by Robbé et al. (1996; 19.3% vs. 8.5%) and Physical violence
by Roberts, O’Toole, Raphael, Lawrence, and Ashby perpetration (%)
(1996; 16.1% vs. 4.1%) as well as the Dutch study Women Men Study∗
by Römkens (1997; 20.8% vs. 7.4%). Five studies re- Prevalence
ported similar levels of victimization for women and UK 11 10 Carrado et al. (1996)
men: the New Zealand study by Jackson et al. (2000), Incidence
the German study by Wetzels, Greve, Mecklenburg, Korea 6.3 20.8 Kim and Sung (2001)
Bilsky, and Pfeiffer (1995), the study with Hong New Zealand 37.2 21.8 Magdol et al. (1997)
South Africa 43.5 35.3 Swart et al. (2002)
Kong Chinese by Tang (1999a), the South African UK 53.7a 69.5a Dobash and
study by Swart et al. (2002) and the UK study by Dobash (2004)
Dobash and Dobash (2004). Three studies found ∗ More detailed information about the studies is provided in Ta-
higher victimization rates for men than for women:
bles I and II.
the German study by Luedtke and Lamnek (2002; 6.0 a Sample consisted of men convicted of intimate partner violence
vs. 2.7%), the UK study by Carrado, George, Loxam, and their female partners.
Jones, and Templar (1996; 18 vs. 13%), and the New
Zealand study by Magdol, Moffitt, Caspi, Newman,
Fagan, and Silva (1997; 34.1 vs. 27.1%). Finally, the ies, from Korea (Kim & Sung, 2001) and the UK
study by Mirrlees-Black (1999) provided divergent (Dobash & Dobash, 2004), perpetration rates were
results for lifetime prevalence and 12-months inci- higher for men than for women, in the other two
dence: while lifetime victimization rates were higher studies from New Zealand (Magdol et al., 1997) and
for women (23%) than for men (15%), incidence fig- South Africa (Swart et al., 2002), perpetration rates
ures were exactly the same for both men and women. were higher for women than for men. With respect to
Only one study directly compared women’s and sexual aggression, the German data by Krahé (Krahé
men’s sexual victimization rates. In their sample & Scheinberger-Olwig, 2002; Krahé, Waizenhöfer, &
of New Zealand adolescents, Jackson et al. (2000) Bieneck, 2003) found a higher prevalence of sex-
found prevalence rates for experience of unwanted ual aggression toward an intimate partner for men
sexual activity of 76.9% for women and 67.4% by (21.6%) than for women (6.0%).
men. In contrast, two German studies looking at Overall, the international literature reflects the
women’s and men’s sexual victimization by an in- same degree of inconsistency concerning men’s and
timate partner on the basis of very similar mea- women’s equal vs. unequal vulnerability to and per-
sures (Krahé & Scheinberger-Olwig, 2002; Krahé, petration of intimate partner violence that is char-
Scheinberger-Olwig, & Bieneck, 2003) found higher acteristic of the North American body of evidence
victimization rates for women (30.1%) than for men (e.g., Saunders, 2002). An important question raised
(11.7% for an adolescent and 17.6% for a young by comparative analyses of men’s and women’s in-
adult sample). volvement in intimate partner violence refers to
the extent to which perpetration and victimization
reports can be related to each other. Only four
Studies Comparing Men’s and Women’s studies presented data from men and women on
Perpetration Rates both perpetration and victimization, providing a full
cross-classification of gender and perspective (victim
Five studies presented comparative information vs. perpetrator). These are the studies by Carrado
about men’s and women’s involvement as perpetra- et al. (1996; UK), Dobash and Dobash (2004; UK),
tors of intimate partner violence. Their findings, sum- Magdol et al. (1997; New Zealand), and Swart et al.
marized in Table IV, match the diversity observed in (2002, South Africa). However, with the exception
the studies comparing victimization rates. The only of the Dobash and Dobash study, even these studies
prevalence study (Carrado et al., 1996) found al- do not shed light on the reciprocal perception of inti-
most identical perpetration rates for men (10%) and mate partner violence because respondents reported
women (11%) in a UK sample. The remaining four only their personal perception of perpetration and
incidence studies showed asymmetries, but are di- victimization unrelated to the perceptions of their
vided in terms of their direction. In two of these stud- partners.
Intimate Partner Violence—An International Perspective 819

Dobash and Dobash (2004) studied 95 British add to a large and consistent body of evidence
couples where the husband had been convicted and from North America showing the co-occurrence of
arrested for some form of domestic violence. This different forms of partner violence (e.g., Barnett,
study is interesting not so much because of the ab- Miller-Perrin, & Perrin, 1997).
solute scale of partner violence uncovered (which
was bound to be high given that all male partici- Consequences of Physical and Sexual Victimization
pants had been convicted for a domestic violence of-
fence) but for providing truly reciprocal reports of The consequences of physical and/or sexual
perpetration and victimization from the two mem- abuse have been examined predominantly for fe-
bers of a couple. In their sample, 78.9% of women male victims. Many serious physical and mental
and 60% of men reported that they had been vic- health problems are considered to result from phys-
timized by their partner, and 69.5% of men as com- ical or sexual abuse by an intimate partner. Some
pared to 53.7% of women acknowledged at least of these problems are immediate, such as injuries
one violent incident against their partner in the last and shock, and some are long term in nature, such
12 months. These findings suggest two conclusions: as increased levels of anxiety or depression. The
(a) men’s violence against women was higher than Center for Health and Gender Equity (CHANGE,
women’s violence against men, as reflected in both 1999) provided the following classification of ad-
perpetrator and victim perceptions; and (b) victim- verse health outcomes of violence against women:
ization rates were higher than the corresponding per- physical health (e.g., injuries, functional impairment
petration rates (e.g., female victimization: 78.9% vs. or permanent disability), negative health behaviors
male perpetration: 69.5%). This means that perpe- (e.g., excessive alcohol consumption or drug abuse),
trators reported less violence than their partner felt chronic conditions (e.g., somatic complaints or
subjected to. However, this difference was signifi- chronic pain syndromes), reproductive health (e.g.,
cant only for men’s violence against women. When unwanted pregnancy, STIs/HIV, unsafe abortion or
asked to assess the seriousness of the violent inci- miscarriage, low birth weight), and mental health
dents, significantly more women than men perceived (post-traumatic stress, depression, anxiety, low self-
men’s violence against the female partner as very esteem, sexual dysfunction, phobias, or eating
serious (46.0% of women vs. 27.7% of men). No disorders).
such difference was found with respect to the pro- The adverse consequences of intimate partner
portion of men and women who regarded women’s victimization were also apparent in many of the sam-
violence as very serious, which in any case was much ples of the studies in this review. In terms of im-
lower (5.3% among men and 4.9% among women). mediate physical health consequences, a substantial
Dobash and Dobash (2004, p. 343) concluded that proportion of victims suffered physical injuries. Stud-
men and women “tend to agree about the nature, ies that recorded whether or not victims were in-
frequency, and impact of the violence perpetrated by jured by their partner’s acts of violence found in-
women, but disagree about men’s violence.” jury rates among victims from 25% (Swart et al.’s,
2002, study of South African adolescents) to 64%
(BMFSFJ, 2004, Germany; and Römkens, 1997, The
Co-Morbidity Netherlands). In the two studies using victim sam-
ples (Arscott-Mills, 2001, Jamaica; and Dobash &
The present review focused on physical and Dobash, 2004, UK), nearly all the victims had suf-
sexual violence between intimate partners. However, fered injuries.
it needs to be recognized that these forms of partner In the studies that compared men’s and women’s
violence tend to co-occur and are often accompanied victimization, injury rates were found to be higher
by psychological or emotional abuse. Studies includ- for women than for men (Mirrlees-Black, 1999; Swart
ing measures of psychological or emotional abuse et al., 2002). This finding supports a large body of
revealed a high degree of co-morbidity between evidence from the North American literature show-
the different forms of partner violence: physical ing that risk of injury as a consequence of intimate
and sexual violence were commonly accompanied partner violence varies as a function of gender. For
by psychological/emotional abuse (Eisikovits et al., example, Rennison and Welchan’s (2000) analysis of
2004; Ellsberg et al., 2000; Grimstad, Schei, Backe, data from the National Crime Victimization Survey
& Jacobsen, 1999; Römkens, 1997). These findings from 1993 to 1998 showed that a greater proportion
820 Krahé, Bieneck, and Möller

of women than men were injured in violent in the family. The last level describes risk factors
confrontations with an intimate partner. Archer’s at the level of the individual perpetrator, including
(2000) meta-analysis also revealed that significantly biographical experiences, such as witnessing mari-
more women than men were injured by their tal violence as a child, an absent or rejecting fa-
partners. ther, being abused as a child, and alcohol abuse
With respect to the psychological consequences (Heise, 1998).
of intimate partner violence, a strong association Several of the reviewed studies included risk
between physical abuse and mental health prob- markers (i.e., variables associated with an increased
lems was found in the Syrian sample analyzed by risk that may be correlated with or without being
Maziak (2002). In Haj-Yahia’s (2000a) Israeli study, causally related to physical and sexual violence). The
women victimized by both physical abuse and sex- most common risk markers addressed in the present
ual abuse showed significantly lower self-esteem as studies were demographic variables, such as age,
well as higher levels of depression, stress, and anxi- level of education, and economic situation. Rates
ety than did non-victimized women. Römkens (1997) for both victimization and perpetration declined with
also reported a significantly higher level of psychoso- age (Carrado et al., 1996, Mexico; Haj-Yahia, 2000b,
matic and depressive complaints among victims than Palestine; Mirrlees-Black, 1999, UK; Tang, 1999b,
among non-victims in her Dutch sample. Magdol Hong Kong), indicating that younger women are par-
et al. (1997) obtained a similar pattern for their Aus- ticularly at risk. However, studies also taking the
tralian study, but they also showed that the effect was age difference between partners into account showed
more pronounced for female than for male victims that relationships in which the man was considerably
(this gender difference was also found by Mirrlees- older than the woman were particularly vulnerable to
Black, 1999, in the UK). Piispa (2004) found that intimate partner violence (Tang, 1999b).
older female victims of partner violence (aged be- Low level of education and financial depen-
tween 30 and 64) in her Finnish study were more dence were identified as risk markers of victim-
negatively affected in terms of depression, loss of ization as well as perpetration in several of the
self-esteem, and fear than younger victims (aged un- studies (Castro et al., 2003, Mexico; Hadi, 2000,
der 30). Altogether, the findings from the present Bangladeshi; Haj-Yahia, 2000b, Palestine; Magdol
set of studies corroborate the large body of evidence et al., 1997, Australia; Maman et al., 2002, Tanzania;
from North American research by documenting se- Maziak, 2002, Syria). However, this relationship
vere physical and psychological problems as a result was not pervasive throughout the studies. Kim and
of intimate partner violence. Sung’s (2001) study of elderly couples in Korea
and Tang’s (1999b) study of Hong Kong Chinese
did not find higher victimization rates in relation
Risk Markers of Intimate Partner Violence to low income or unemployment. Unemployment
and low level of education were also identified
The need for a better understanding of the as risk markers of violence perpetration (Cas-
causes of intimate partner violence is stressed by tro et al., 2003, Mexico; Eisikovitz et al., 2004,
all researchers in this field. The “ecological frame- Israel; Haj-Yahia, 2000b, Palestine; Mirrlees-Black,
work” proposed by Heise (1998) is one of the most 1999, UK).
promising approaches to understand the interplay of With respect to biographical variables, different
personal, situational, and sociocultural factors that studies provided evidence that experiencing abuse as
combine to cause intimate partner violence. The a child and witnessing violence in one’s family of ori-
model tries to explain men’s violence against women gin was a risk marker of both perpetration (Castro
by specifying risk factors at four levels (CHANGE, et al., 2003, Mexico; Krahé, Waizenhöfer, & Möller,
1999). The society level includes norms granting men 2003, Germany; Swart et al., 2002, South Africa) and
control over female behavior, acceptance of violence victimization (Castro et al., 2003, Mexico; Ellsberg
as a means of conflict resolution, and rigid gen- et al., 2000, Israel; Krahé & Scheinberger-Olwig,
der roles. Risk factors at the community level are 2002, Germany). Again, these findings are in ac-
poverty, low socioeconomic status, and unemploy- cordance with the evidence from North American
ment, as well as isolation of women and family. As- sources, suggesting that the variables associated with
pects at the relationship level include marital con- an increased risk of intimate partner violence oper-
flict and male control of wealth and decision-making ate in a similar fashion in different cultural settings.
Intimate Partner Violence—An International Perspective 821

DISCUSSION acknowledged that there are probably many more


data sets out there available only as unpublished re-
The studies reviewed in this paper convey one ports and/or in languages inaccessible to us. It is
clear message: intimate partner violence, both phys- impossible to say to what extent these data would
ical and sexual, is prevalent in many societies, not make substantive contributions to the international
just in Western countries but across the world, and literature on intimate partner violence and how they
several of the risk markers as well as adverse conse- would affect the conclusions based on published
quences identified in the North American literature data.
were also found in a diverse sample of international A further limitation of the present database
studies. Beyond this broad statement, however, the refers to the comparability of the findings. The dif-
body of evidence reviewed in this paper raises more ficulties of comparing studies that differ widely in
questions than it answers. Prevalence and incidence terms of methodology have already been pointed
rates of both victimization by and perpetration of in- out. An additional problem lies in the presentation
timate partner violence were found to vary widely of the findings. All the studies we examined pre-
across the studies. Concerning the controversial issue sented information about prevalence or incidence
of men’s and women’s symmetrical vs. asymmetrical rates of intimate partner violence that were aggre-
involvement in intimate partner violence, the present gated across individual items, even when prevalence
evidence was unable to provide a clear answer. The or incidence rates were reported for individual items.
number of studies reporting similar rates of male and While this is a reasonable approach enabling readers
female perpetration or victimization, higher rates of to appreciate the general picture of partner violence
victimization for women or for men, and higher per- from the study, one should not be led to believe that
petration rates for men or for women were about similar classifications reflect similar underlying be-
equally divided. Rather than discussing the findings haviors. Some authors classified behaviors on the ba-
from individual studies, we will examine some of the sis of the approach provided by the authors of the
lessons to be learned from the studies included in this CTS (e.g., Kim & Sung, 2001; Magdol et al., 1997;
review and outline directions for future research. Römkens, 1997; Tang, 1999b). Other authors used
their own criteria (e.g., Piispa, 2004) or failed to ex-
plicate them altogether. In the present review, space
Limitations of the Database limitations precluded the presentation of findings
from the 35 studies at the individual item level. In-
A first point to be noted is that with the ex- stead categorical data (e.g., moderate or severe vio-
ception of three studies (BMFSFJ, 2004; Luedtke lence) were reported as provided by the different au-
& Lamnek, 2002; Mirrlees-Black, 1999), only stud- thors. This means, however, that moderate violence
ies were included in our review that were pub- as defined in one study may not map onto moder-
lished in academic journals. This decision was based ate violence reported from another study, even when
on two considerations: (a) that some form of qual- both used the CTS at the item level.
ity control was applied to papers that passed the Finally, the fact that a majority of studies in
standards of (generally peer-reviewed) journals; and this review used the CTS as a measure of intimate
(b) that these studies would be easily accessible to partner violence means that they are susceptible to
readers who wanted to follow up the information the same criticisms and concerns raised about this
provided in this review by consulting the original instrument in the North American literature (cf.
sources. This is important because our review con- summaries by Dobash & Dobash, 2004; Frieze, 2000;
centrated on the issue of prevalence or incidence Schwartz, 2000). The most fundamental criticism is
of intimate partner violence, whereas many of the that the CTS neglects the contextual framework in
studies went beyond this and looked at predictors or which partner violence occurs and does nothing else
moderators of prevalence/incidence rates, contribut- but “simply and crudely counting blows” (Schwartz,
ing a host of interesting findings over and above 2000, p. 821). By focusing on behavioral descriptions
those presented here. The decision to include the of individual incidents of partner violence, the CTS
three studies not published in academic outlets was is unable to capture the dynamics of violent episodes,
made because the available online summaries were such as whether an act of violence is proactive or un-
detailed enough to appreciate the soundness of the provoked as opposed to reactive or in self-defense.
methodological approach. However, it needs to be The original version furthermore neglected the
822 Krahé, Bieneck, and Möller

consequences of partner violence, i.e. the presence women than by men.” This criticism also pertains to
and severity of injuries, but this was addressed by most of the studies included in the present review.
including a separate injury severity scale in the
revised CTS2. However, even with this modification,
Generalizability within Countries
reliance on act-based frequency tallies runs the risk
of equating behaviors that vary widely in their poten-
A factor that precludes firm conclusions about
tial to harm. As Dobash and Dobash (2004) pointed
the international scale of intimate partner violence
out, the item “throwing an object at your partner”
from the present set of studies is the lack of a re-
may cover such diverse acts as throwing a lamp and
liable database within each country. With few ex-
throwing a pillow, which differ not only with respect
ceptions, each country is represented by only one
to their potential to cause injuries but also possibly
study. Therefore, it is unclear to what extent the re-
in terms of the intensity of the underlying intention
ported findings are reliable and valid reflections of
to harm.
the level of intimate partner violence in that partic-
Several critics have argued that the picture of
ular country. This problem is highlighted by look-
gender symmetry, i.e., men’s and women’s equal
ing at countries for which more than one study was
involvement in intimate partner violence, portrayed
available. For instance, the findings by Luedtke and
by studies using the CTS is largely due to the fact
Lamnek (2002) suggest that prevalence rates for inti-
that decontextualized instances of violence are
mate partner victimization are higher for men than
recorded that fail to distinguish motivationally dis-
for women in Germany. In contrast, the incidence
tinct forms of intimate partner violence in which men
rates found by Wetzels et al. (1995) for victimization
and women are differentially involved (Dobash &
by physical partner violence were higher for women
Dobash, 2004; Johnson, 1995). Johnson and Ferraro
than for men in the same country. Both studies re-
(2000) highlighted this problem in their distinction
lied on large and randomly drawn community sam-
between different types of partner violence. They
ples. Similar discrepancies are apparent in the two
argued that understanding partner violence requires
New Zealand studies. Jackson et al. (2000) found
a closer analysis of the motivational basis of the
slightly higher rates of victimization for women than
aggressor and proposed three distinct manifestations
for men, whereas Magdol et al. (1997) found sub-
of intimate partner violence: (1) Common couple
stantially higher victimization rates for men than for
violence (CCV), which arises in the context of a
women. Both studies used high school samples of
specific argument, is likely to be mutual, unlikely
similar age and sample size. These examples show
to involve serious injury or escalate over time, and
that we are still a far cry from being able to assess the
the most common form of intimate partner violence
scale of intimate partner violence within a particular
in the general population; (2) Intimate terrorism
country, which would be a precondition for examin-
(IT), based on the desire to exert control over the
ing and interpreting variations between countries. In
partner, is shown more frequently in a particular
addition, one has to deal with the problem of cultural
relationship, less likely to be mutual, and more likely
variability within countries that may render the task
to lead to serious injury and to escalate over time;
of providing overall rates questionable. Some of the
(3) Violent resistance (VR), shown in response to
studies in this review related differences within their
a violent attack from the partner, is perpetrated
samples to cultural parameters, such as religious af-
almost exclusively by women. Ignoring the different
filiation and power differentials in relationships (cf.
contextual features and motivational underpinnings
later sections).
of these forms of intimate partner violence leads to
a distorted picture of the roles of men and women
as perpetrators and victims. As Johnson and Ferraro Comparability Across Countries
(2000, p. 950) argued, almost all the studies claiming
gender symmetry in intimate partner violence “use Even if there were solid national databases on
the sort of general heterosexual sample in which intimate partner violence from a range of coun-
aggregated violence only appears to be gender tries, establishing patterns of worldwide variations
symmetric because it lumps together IT, which is or similarities would remain a difficult task. One
essentially perpetrated by men; CCV, which is per- problem characterizing the present state of research
petrated slightly more often by men than by women; is the fragmented and unsystematic nature of the
and VR, which is clearly perpetrated more often by data available for cross-cultural comparison. The
Intimate Partner Violence—An International Perspective 823

studies identified for this review on the basis of a per are based on the imposed etic approach. At a
careful search of the literature are best described as methodological level, this approach is confronted
a convenience sample. There was no comprehensive with the problems of “functional equivalence” and
database from which we could have drawn a random “measurement equivalence” (Berry et al., 2002).
sample, nor was it possible to select studies on the ba- Functional equivalence refers to the question of
sis of theoretical considerations, such as a country’s whether an instrument (e.g., CTS or SES) used to
standing on Hofstede’s (1980, 1998) dimensions of measure partner violence in one culture measures the
masculinity/femininity or individualism/collectivism. same underlying construct in another culture. For ex-
In addition, the task of comparing the scale of in- ample, the item “Have you had sexual intercourse
timate partner violence across cultures faces several when you didn’t want to because a man threatened
interrelated conceptual and methodological prob- or used some degree of physical force (twisting your
lems that are widely recognized by researchers in this arm, holding you down etc.)?” is treated as a rape
field, yet still far from being adequately addressed item in the SES (Koss et al., 1987). Using the same
(see also Krug et al., 2002, p. 92). A fundamen- item in a different culture and interpreting the re-
tal problem refers to the perspective to be adopted ported frequencies as prevalence figures of rape as-
by the research as a basis for studying psychologi- sumes that in that culture using force to obtain sexual
cal problems and processes in different cultures. In intercourse is considered unlawful in the same way
cross-cultural research, this issue is discussed in terms as in the Western world. This implicit assumption is
of the “etic” vs. “emic” distinction (Berry, Poortinga, problematic and rarely put to empirical test. In fact,
Segall, & Dasen, 2002; Smith & Bond, 1998). even within countries, this problem applies to the
Etic versus Emic. Structured measures like comparison of different racial or ethnic groups (e.g.,
the CTS and the SES present respondents with Sanchez-Hucles & Button, 1999). The second prob-
behavioral definitions of different forms of partner lem, measurement equivalence, refers to the way in
violence. Therefore, the use of such instruments which the data are collected. The standard format of
across different cultures should enable cross-cultural written questions used in Western cultures may be
comparisons by yielding figures on the prevalence or inappropriate in cultures with low literacy rates or in
incidence of partner violence unaffected by differen- which filling in forms is not part of everyday life. In
tial interpretations and evaluations of the behaviors Deyessa et al.’s (1998) Ethiopian study, for instance,
in question. In terms of methodological perspectives 60% of the sample was illiterate. Therefore, the items
in cross-cultural research, using the CTS or the SES had to be read out to them in a face-to-face interview,
to study physical and sexual violence in diverse creating a different measurement context from pro-
countries represents a prototypical case of the etic viding written material under anonymous conditions.
approach (Smith & Bond, 1998). This approach The Human Relations Area File (HREF), pro-
assumes that a given construct, such as wife battering viding an ethnographic archive of culture-specific in-
as a form of intimate partner violence, is universally formation, is an example of “derived etic” informa-
meaningful and the phenomenon universally given, tion. In his cross-cultural analysis of family violence,
albeit with different prevalence rates and, possibly, Levinson (1989) drew on the HREF to establish the
different determinants. On the basis of that assump- prevalence and typology of wife battering across 90
tion, research based on the etic approach seeks to cultural groups. As noted by Berry et al. (2002), a
identify similarities and differences between cultures problem with ethnographic archives like the HREF
with regard to the construct in question. To the is that the quality of the data is extremely variable.
extent that researchers transfer their own definitions The extent of this problem can be gauged by relat-
and measures into another culture, they adopt an ing findings across cultures to control variables, such
“imposed etic” approach. In contrast, research based as researchers’ language fluency and length of stay
on conceptualizations and measures developed in the respective culture. Levinson established that
within a particular culture represents a “derived reports of the frequency of wife battering in the 90
etic” approach. In studies using the “derived etic” cultural groups included in his analysis were uncorre-
approach, the aim is still to obtain evidence about lated with these control variables, increasing the con-
a particular construct that can be compared across fidence in the database.
cultures, but it allows for definitions and measures of The problems associated with the “imposed
that construct to be developed within each culture. etic” approach have long been pointed out by
The majority of studies reviewed in this pa- researchers calling for an emic approach to
824 Krahé, Bieneck, and Möller

complement (or sometimes more radically: to sub- (Levinson, 1989, pp. 54 and 72). Cultural variables
stitute) the etic perspective. In the emic approach, may affect not only the documented scale of part-
problems and issues for research are identified within ner violence but also the willingness to report it. As
a particular culture, and conceptual definitions as Sanchez-Hucles and Button (1999) point out, cultural
well as measurement tools are developed for use in groups with a collectivist orientation concerned pri-
that culture on the basis of the socially shared views marily with preserving the “face” of the family as a
of its members. The primary goal is to understand social unit may show seemingly lower rates of part-
the specific social constructions within a culture, not ner violence because of the unwillingness to disclose
the comparative analysis across cultures, which dis- instances of partner violence to third parties.
tinguishes the emic approach from the “derived etic” In the present sample of studies, Ellsberg et al.
approach described earlier. With regard to partner (2000) linked their finding that 52% of the women
violence, an emic approach would have to start by ever married or living with a partner experienced
analyzing whether or not a given culture has devel- physical abuse at some point in a relationship to the
oped a perception of forms of treatment of one’s dominant gender role conceptions in Nicaraguan so-
partner that are considered illegitimate (Tedeschi ciety. They describe two complementary concepts,
& Felson, 1994) and then go on to describe the Machismo and Marianismo, that define men’s iden-
specific behavioral manifestations falling within that tity in terms of dominance and aggression and
category. Very few such studies are reported in women’s identity in terms of dependency and sub-
the international literature, not least because they mission, respectively. Both concepts are conducive to
are less likely to pass the hurdles for acceptance in men’s violence against female partners in that they
mainstream international journals with their focus highlight men’s entitlement to women and women’s
on English language, quantitative methodology, and obligation to subject themselves to their husband’s
rigorous empirical control. Of the studies considered demands. They act against recognizing intimate part-
in this review, none is based on an emic analysis. ner violence as a social problem and lead to the
stigmatization of women who identify themselves as
victims. In a similar vein, Kim and Emery (2003)
Relating the Scale of Intimate Partner found that the rate of physical violence was consid-
Violence to Cultural Parameters erably higher in relationships characterized by male
dominance, i.e., unequal power structures favoring
A final limitation of the present body of ev- men, the traditional partnership ideology in Korean
idence to be acknowledged is that few studies at- society. The findings obtained by Tang (1999a) for
tempted to link the scope of physical and sexual vio- Chinese respondents in Hong Kong point in the same
lence to cultural variables that might provide cues to direction. Hadi (2000) also discussed the finding that
explaining cultural variation. This is due in large part 26% of married women in his Bangladeshi sample
to the fact that the studies included in this review pro- had experienced forced sex from their husband in
vided primary data from a particular country without relation to prevailing cultural norms that consider
a comparative focus. Levinson’s (1989) comparative forced intercourse by husbands for sexual gratifica-
analysis illustrates how cultural factors can be linked tion as legitimate. He concluded that “given the so-
to the prevalence of partner violence to examine dif- cial and cultural context of Bangladesh, it is too early
ferent causal explanations. He found substantial cor- to expect that sexual violence should be taken as a se-
relations between the extent of wife battering and the rious social and health problem” (p. 801). Eisikovitz
level of violence in other areas of social life, such as et al. (2004, p. 745) identified differences in the rate
male fighting and female fighting. These findings sup- of intimate partner violence in their Israeli sample
port the “culture patterns” model in which family vi- as a function of religious affiliation, with traditional
olence is seen as a reflection of a general tendency Jews and Muslims reporting higher rates than liberal
to use violence within a society. In terms of socio- Jews, concluding that “these groups share an attitudi-
structural variables, he found, for example, that di- nal orientation that legitimizes such behavior.” Haj-
vorce frequency was unrelated to wife beating across Yahia (2000b) found that victimization rates were
cultural groups. However, gender differences in di- higher among Muslim than among Christian women
vorce freedom, i.e. divorce being more difficult or im- in Palestine. In combination, these data support the-
possible for the wife compared to the husband, were oretical accounts of intimate partner violence de-
positively correlated with frequency of wife beating veloped and tested in Western societies that regard
Intimate Partner Violence—An International Perspective 825

asymmetrical decision-making power in relationships Bond, M. H. (2004). Culture and aggression—From context to
as a causal factor in intimate violence. coercion. Personality and Social Psychology Review, 8, 62–
78.
*Bradley, F., Smith, M., Long, J., & O’Dowd, T. (2002). Re-
ported frequency of domestic violence: Cross sectional survey
CONCLUSION of women attending general practice. British Medical Journal,
324 (7332), 271.
The aim of this paper was to bring together stud- Brownmiller, S. (1975). Against our will: Men, women and rape.
New York: Schuster.
ies from a wide range of countries outside North *Carrado, M., George, M. J., Loxam, E., Jones, L., & Tem-
America that examined the role of gender in inti- plar, D. (1996). Aggression in British heterosexual relation-
mate partner violence. The encouraging message to ships: A descriptive analysis. Aggressive Behavior, 22, 401–
415.
be drawn from the studies included in our review is *Castro, R., Peek-Asa, C., & Ruiz, A. (2003). Violence against
that the problem of intimate partner violence is be- women in Mexico: A study of abuse before and during
coming increasingly recognized across the world. Sys- pregnancy. American Journal of Public Health, 93, 1110–
1116.
tematic and large-scale research activities are gain- Center for Health and Gender Equity (CHANGE). (1999). Pop-
ing momentum, not only in industrialized countries ulation reports. Ending violence against women. Issues in
but also in economically deprived parts of the world. World Health, 17 (4).
Currie, D. H. (1998). Violent men or violent women? Whose def-
There is also a growing awareness that the traditional inition counts? In R. K. Bergen (Ed.), Issues in intimate vio-
perspective looking at women as victims and men as lence (pp. 97–111). Thousand Oaks, CA: Sage.
perpetrators of intimate partner violence needs to Dasgupta, S. D. (2002). A framework for understanding women’s
use of nonlethal violence in intimate heterosexual relation-
be expanded by considering the possibility of male ships. Violence against Women, 8, 1364–1389.
victimization and female perpetration. The findings *Deyessa, N., Kassaye, M., Demeke, B., & Taffa, N. (1998). Mag-
from these studies have been made available to the nitude, type and outcomes of physical violence against mar-
ried women in Butajira, Southern Ethiopia. Ethiopian Medi-
international research community through accessible cal Journal, 36, 83–92.
and quality-controlled publication outlets, contribut- *Dobash, R. P., & Dobash, R. E. (2004). Women’s violence in in-
ing to the development of an international database timate relationships: Working on a puzzle. British Journal of
Criminology, 44, 324–349.
on intimate partner violence. On the other hand, the *Eisikovits, Z., Winstok, Z., & Fishman, G. (2004). The First Is-
review has also revealed that the comparative analy- raeli National Survey on domestic violence. Violence against
sis of intimate partner violence across cultures needs Women, 10, 729–748.
*Ellsberg, M., Pena, R., Herrera, A., Liljestrand, J., & Winkvist,
to overcome a number of conceptual and method- A. (2000). Candies in hell: Women’s experiences of vio-
ological problems before conclusions can be drawn lence in Nicaragua. Social Science & Medicine, 51, 1595–
with confidence. There is still a long way to go before 1610.
Frieze, I. H. (2000). Violence in close relationships – Development
a more focused picture will emerge about the level of a research area: Comment on Archer (2000). Psychological
and cultural embeddedness of intimate partner vio- Bulletin, 126, 681–684.
lence, but it seems that researchers across the world *Grimstad, H., Schei, B., Backe, B., & Jacobsen, G. (1999). In-
terpersonal conflict and physical abuse in relation to preg-
are well positioned to take up the challenge. nancy and infant birth weight. Journal of Women’s Health &
Gender-Based Medicine, 8, 847–853.
*Hadi, A. (2000). Prevalence and correlates of the risk of marital
REFERENCES sexual violence in Bangladesh. Journal of Interpersonal Vio-
lence, 15, 787–805.
Hagemann-White, C. (2001). European research on the preva-
Anderson, P. B., & Struckman-Johnson, C. (Eds.) (1998). Sexually lence of violence against women. Violence against Women, 7,
aggressive women. New York: Guilford Press. 732–759.
Archer, J. (2000). Sex differences in aggression between hetero- *Haj-Yahia, M. M. (2000a). Patterns of violence against engaged
sexual partners: A meta-analytic review. Psychological Bul- Arab women from Israel and some psychological implica-
letin, 126, 651–680. tions. Psychology of Women Quarterly, 24, 209–219.
*Arscott-Mills, S. (2001). Intimate partner violence in Jamaica: A *Haj-Yahia, M. M. (2000b). The incidence of wife abuse and bat-
descriptive study of women who access the services of the tering and some sociodemographic correlates as revealed by
women’s crisis centre in Kingston. Violence Against Women, the two National Surveys in Palestinian Society. Journal of
7, 1284–1302. Family Violence, 15, 347–374.
Barnett, O. W., Miller-Perrin, C. L., & Perrin, R. D. (1997). Family Heise, L. (1998). Violence against women: An integrated, ecolog-
violence across the lifespan. Thousand Oaks, CA: Sage. ical framework. Violence Against Women, 4, 262–290.
Berry, J. W., Poortinga, Y. H., Segall, M.H., & Dasen, P. R. Hofstede, G. (1980). Culture’s consequences. Beverly Hills, CA:
(2002). Cross-cultural psychology: Research and applications Sage.
(2nd ed.). New York: Cambridge University Press. Hofstede, G. (Ed.). (1998). Masculinity and femininity: The taboo
*BMFSFJ (Bundesministerium für Familie, Senioren, Frauen dimension of national cultures. Thousand Oaks, CA: Sage.
und Jugend) (Ed.). (2004). Lebenssituation, Sicherheit und *Jackson, S. M., Cram, F., & Seymour, F. W. (2000). Violence and
Gesundheit von Frauen in Deutschland. Bonn: BMFSFJ. Re- sexual coercion in high school students’ dating relationships.
trieved 20-09-2004: www.bmfsfj.de. Journal of Family Violence, 15, 23–36.
826 Krahé, Bieneck, and Möller

Johnson, M. P. (1995). Patriarchial terrorism and common couple (Eds.), What causes men’s violence against women? (pp. 18–
violence: Two forms of violence against women. Journal of 35). Thousand Oaks, CA: Sage.
Marriage and the Family, 57, 283–294. *Maziak, W. (2002, July). Violence against women. The Lancet,
Johnson, M. P., & Ferraro, K. J. (2000). Research on domestic vi- 360 (27), 343–344.
olence in the 1990s: Making distinctions. Journal of Marriage Medina-Ariza, J., & Barberet, R. (2003). Intimate partner violence
and the Family, 62, 948–963. in Spain. Findings from a national survey. Violence against
*Kim, J., & Emery, C. (2003). Marital power, conflict, norm con- Women, 9, 302–322.
sensus, and marital violence in a nationally representative *Mirrlees-Black, C. (1999). Domestic violence: Findings from
sample of Korean couples. Journal of Interpersonal Violence, a new British Crime Survey self-completion questionnaire
18, 197–219. (Home Office Research Study 191). London: Home Of-
*Kim, J., & Sung, K. (2001). Marital violence among Korean el- fice. Retrieved 23–9-2004: www.homeoffice.gov.uk/rds/pdfs/
derly couples: A cultural residue. Journal of Elder Abuse and hors191.pdf.
Neglect, 13, 73–89. *Piispa, M. (2004). Age and meanings of violence: Women’s ex-
Kimmel, M. S. (2002). “Gender symmetry” in domestic violence. periences of partner violence in Finland. Journal of Interper-
A substantive and methodological research review. Violence sonal Violence, 19, 30–48.
against Women, 8, 1332–1363. Rennison, C. M., & Welchans, S. (2000). Intimate partner violence.
Koss, M. P., & Oros, C. J. (1982). Sexual experiences survey: A Bureau of Justice Statistics Special Report, May, 2000. Re-
research instrument investigating sexual aggression and vic- trieved 23–9-2004: www. ojp.usdoj.gov/bjs/cvict.htm.
timization. Journal of Consulting and Clinical Psychology, 50, *Robbé, M. de Vries, March, L., Vinen, J., Horner, D., &
455–457. Roberts, G. (1996). Prevalence of domestic violence among
Koss, M. P., Gidycz, C. A., & Wisniewski, N. (1987). The scope patients attending a hospital emergency department. Aus-
of rape: Incidence and prevalence of sexual aggression and tralian and New Zealand Journal of Public Health, 20, 364–
victimization in a national sample of higher education stu- 368.
dents. Journal of Consulting and Clinical Psychology, 55, 162– *Roberts, G. L., O’Toole, B. I., Raphael, B., Lawrence, J. M.,
170. & Ashby, R. (1996). Prevalence study of domestic violence
*Krahé, B. (1998). Sexual aggression among adolescents: Preva- victims in an emergency department. Annals of Emergency
lence and predictors in a German sample. Psychology of Medicine, 27, 747–753.
Women Quarterly, 22, 537–554. *Römkens, R. (1997). Prevalence of wife abuse in the Nether-
Krahé, B., Reimer, T., Scheinberger-Olwig, R., & Fritsche, I. lands: Combining quantitative and qualitative methods in
(1999). Measuring sexual aggression: The reliability of the survey research. Journal of Interpersonal Violence, 12, 99–
Sexual Experiences Survey in a German sample. Journal of 125.
Interpersonal Violence, 14, 91–100. *Romito, P., Saurel-Cubizolles, M. J., & Crisma M. (2001). The
*Krahé, B., & Scheinberger-Olwig, R. (2002). Sexuelle aggression. relationship between parents’ violence against daughters and
Göttingen: Hogrefe. violence by other perpetrators. Violence Against Women, 7,
*Krahé, B., Scheinberger-Olwig, R., & Bieneck, S. (2003). Men’s 1429–1463.
reports of nonconsensual sexual interactions with women: Sanchez-Hucles, J., & Dutton, M. A. (1999). The interaction
Prevalence and impact. Archives of Sexual Behavior, 32, 165– between societal violence and domestic violence. In M.
175. Harway & J. M. O’Neil (Eds.), What causes men’s vio-
Krahé, B., Schütze, S., Fritsche, I., & Waizenhöfer, E. (2000). lence against women? (pp. 183–203). Thousand Oaks, CA:
The prevalence of sexual aggression and victimization among Sage.
homosexual men. The Journal of Sex Research, 37, 142– Saunders, D. G. (2002). Are physical assaults by wives and girl-
150. friends a major social problem? A review of the literature.
*Krahé, B., Waizenhöfer, E., & Möller, I. (2003). Women’s sexual Violence Against Women, 8, 1424–1448.
aggression against men: Prevalence and predictors. Sex Roles, Schwartz, M. D. (2000). Methodological issues in the use of sur-
49, 219–232. vey data for measuring and characterizing violence against
Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & women. Violence Against Women, 6, 815–838.
Lozano, R. (2002). World report on violence and health. Smith, P. B., & Bond, M. H. (1998). Social psychology across cul-
Geneva: World Health Organisation. Available online at: tures (2nd ed.). New York: Harvester Wheatsheaf.
www.who.org. Straus, M. A. (1979). Measuring intrafamily conflict and violence:
*Kury, H., Chouaf, S., Obergfell-Fuchs, J., & Woessner, G. (2004). The Conflict Tactics (CT) Scales. Journal of Marriage and the
The scope of sexual victimization in Germany. Journal of Family, 41, 75–88.
Interpersonal Violence, 19, 589–602. Straus, M. A., & Gelles, R. J. (1990). Physical violence in Ameri-
Levinson, D. (1989). Family violence in cross-cultural perspective. can families: Risk factors and adaptations to violence in 8,145
Newbury Park, CA: Sage. families. Brunswick, NJ: Transaction.
*Luedtke, J., & Lamnek, S. (2002). Schläge in jeder dritten Fami- Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B.
lie. Agora, 1, 8–9. (1996). The revised Conflict Tactics Scales (CTS2): Develop-
*Magdol, L., Moffitt, T. E., Caspi, A., Newman, D. L., Fagan, J., & ment and preliminary psychometric data. Journal of Family
Silva, P. A. (1997). Gender differences in partner violence in a Issues, 17, 283–316.
birth cohort of 21-year-olds: Bridging the gap between clinical *Swart, L., Stevens, M. S. G., & Ricardo, I. (2002). Vi-
and epidemiological approaches. Journal of Consulting and olence in adolescents’ romantic relationships: Findings
Clinical Psychology, 65, 68–78. from a survey amongst school-going youth in a South
*Maman, S., Mbwambo, J. K., Hogan, N. M., Kilonzo, G., African community. Journal of Adolescence, 25, 385–
Campbell, J. C., Weiss, E., et al. (2002). HIV-positive women 395.
report more lifetime partner violence: Findings from a vol- *Tang, C. (1999a). Marital power and aggression in a community
untary counseling and testing clinic in Dar es Salaam, sample of Hong Kong Chinese families. Journal of Interper-
Tanzania. American Journal of Public Health, 92, 1331– sonal Violence, 14, 586–602.
1337. *Tang, C. (1999b). Wife abuse in Hong Kong Chinese families:
Marin, A. J., & Russo, N. F. (1999). Feminist perspectives of A community survey. Journal of Family Violence, 14, 173–
male violence against women. In M. Harway & J. M. O’Neil 191.
Intimate Partner Violence—An International Perspective 827

Tedeschi, J. T., & Felson, R. B. (1994). Violence, aggression, and violence. Journal of Personality and Social Psychology, 84,
coercive actions. Washington, DC: American Psychological 997–1010.
Association. *Wetzels, P., Greve, W., Mecklenburg, E., Bilsky, W., &
Toro-Alfonso, J., & Rodriguez-Madera, S. (2004). Domestic vi- Pfeiffer, C. (1995). Kriminalität im Leben alter Men-
olence in Puerto Rican gay male couples: Perceived preva- schen: Eine vergleichende Untersuchung von Opfererfahrun-
lence, intergenerational violence, addictive behaviors and gen, persönlichem Sicherheitsgefühl und Kriminalitätsfurcht.
conflict resolution skills. Journal of Interpersonal Violence, 19, Stuttgart: Kohlhammer.
639–654. White, J. W., & Koss, M. P. (1991). Courtship violence: Incidence
Vandello, J. A., & Cohen, D. (2003). Male honor and female in a national sample of higher education students. Violence
fidelity: Implicit cultural scripts that perpetuate domestic and Victims, 6, 247–256.

You might also like