What Puts Women at Risk of Violence From Their Husbands? Findings From a Large,Nationally Representative Survey in Turkey

Abstract

A large, nationally representative, cross-sectional survey was conducted in Turkey in 2008. In this survey, which used the WHO (World Health Organization) study module on violence, information about lifetime and current violence (past 12 months) was obtained using weighted, stratified, and multistage cluster sampling. This article describes factors associated with physical or sexual violence experienced by ever-married women, aged 15 to 49, from their current or most recent husbands in the 12 months before the survey. Logistic regression analysis is used to describe the risk and protective factors from a considerable range of explanatory variables. The findings confirm that many factors are similar to the experiences of other countries. The physical or sexual violence experienced by ever-married women from their husbands was 15.1%. The violence experienced by women is significantly positively associated with early childhood abuse experiences of both women and their husbands; marriages decided by families or others; husband’s behaviors such as drunkenness, adultery, controlling women’s behavior, and preventing contact with women’s family and friends. The age of the women, their contribution to the household income, support from women’s families, women’s acceptance of male authority, and nonpartner violence experience as well as regional differentials also affect the risk of violence. No significant associations were found with the employment status of women and men or education difference. This study, as one of the largest surveys ever conducted on the issue of domestic violence using face-to-face interviews, demonstrated how the patriarchal family structure still affects women’s lives in Turkey. This is particularly significant, given Turkey’s setting between traditional and modern values.

Keywords

domestic violence sexual assault

Introduction

Violence against women is a global problem that violates women’s human rights and adversely affects their health and well-being. Over the past three decades, the number of studies examining violence against women has increased dramatically as research into all forms of violence has slowly become a public health priority. Between 1999 and 2003, the World Health Organization (WHO) launched an initiative to collect comparable data on violence against women in 15 sites in 10 low-, medium-, and high-income countries. At about the same time, the Demographic and Health Surveys (DHS) began collecting data on gender-based violence in a range of developing countries via newly introduced modules on domestic violence and female genital cutting (Hindin, Kishor, & Ansora, 2008; Kishor & Johnson, 2004). This research revealed that violence by intimate partners is the most common form of violence against women globally (Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts, 2005; Heise, Ellsberg, & Gotteomeller, 1999).

The issue of domestic violence was first brought to the public agenda in Turkey by the feminist movement, with the “Stop Violence” campaign in March 1987. Since then, guesthouses run by the government and counseling centers and shelters run by feminist women’s groups have begun to offer support to women exposed to domestic violence. Moreover, “The Protection of Family Law” (No: 4320), aimed at combating domestic violence was passed in 1998 (Directorate General on the Status of Women [DGSW], 2008). Despite these achievements, many forms of gender inequality remain entrenched in Turkey, and many abused women still do not receive the services or support that they need. Turkey, with a population of 73 million, encompasses significant regional differences in terms of demographic and socioeconomic characteristics. The West is the most industrialized and socially and economically advanced region of the country, with lifestyles similar to those citizens in Western nations, whereas the East is the least developed part of Turkey. A tension also exists between traditional and modern values, and women’s daily existence is still circumscribed by many traditional expectations and religious and customary practices. Arranged, consanguineous, and religious marriages are still common, and the custom of bride price grants families the right to intervene in a couple’s marital life. In a patriarchal society, these factors influence power relationships within the family.

This article focuses on women’s experiences of current physical or sexual violence from their present or most recent husband (hereafter described as “husband”) by using data from the Turkey National Research on Domestic Violence conducted in 2008 (DGSW, 2009). In order to associate marital relationship and physical or sexual violence, only violence from husbands was taken into consideration in Turkey where marriage is widespread. The risk and protective factors for experiencing violence were determined for ever-married women aged 15 to 49 years.

Ecological Framework

Recently, the use of the ecological framework has become popular as a way to conceptualize the multilayered factors that contribute to the risks of experiencing intimate partner violence (Ellsberg, Pena, Herrera, Liljestrand, & Winkvist, 1999; Flake, 2005; Heise, 1998; Krug, Mercy, Dahlberg, & Zwi, 2002). Researchers use the ecological framework as a tool to explain the relationship between individual and contextual factors as well as to highlight the multiple causes of violence and the interaction of risk factors operating within the family and a broader community (Heise, 1998). In this study, a version of the Heise’s ecological framework is adapted to the Turkish context. The first circle contains the personal history factors of the man and woman, including their childhood experiences of abuse (Figure 1). The second circle covers the formation of marriage and marital relationships, including household factors and the current situation of the couple. The outer circle contains community-level factors (Figure 1).

Figure 1.

Ecological framework adapted from Heise (1998)

Personal abuse history

The “cycle of violence” hypothesis assumes a childhood history of abuse increases the likelihood of children experiencing or perpetrating violence in later years. Even though children respond in different ways to witnessing or experiencing violence in their homes, depending on the frequency and severity of exposure (Little & Kantor, 2002), these experiences increase their perception of violence as a normal behavior (Romito, Crisma, & Saurel-Cubizolles, 2001; Romito, Saurel-Cubizolles, & Crisma, 2001). Several studies have found that early childhood experiences of violence are risk factors for experiencing later partner violence (Ellsberg et al., 1999; Flake, 2005; Gelles, 1980; Hotaling & Sugarman, 1986, as cited in Heise, 1998; Jewkes, Levin, & Penn-Kekana, 2002; Kalmuss & Straus, 1982; Martin et al., 2002; Rivera-Rivera et al., 2003; Straus & Gelles, 1986). Husbands’ experience of physical violence in childhood is also positively associated with perpetration of partner violence; for example, witnessing one’s mother being beaten is a stronger predictor of partner violence than being abused oneself as a child (Abrahams & Jewkes, 2005; Naved & Persson, 2005).

Formation of marriage

Age and education differences between couples and young age at marriage are associated with power dynamics in the family. For example, age difference may lead to an increased exposure to violence, since an older husband may be perceived as having increased authority. Marriage at an early age is also associated with the woman having less education, bearing children at an early age, and having a lower position in the household (Singh & Samara, 1996). Not surprisingly, women who marry at an early age are also more likely to experience domestic violence (Jensen & Thornton, 2003).

Several marriage traditions in Turkey may also affect the risk of partner violence in marital relationships. The traditional involvement of the husband’s family in the marital life of young couples is often a source of marital conflict. Studies from a wide range of settings, including Mexico (Agoff, Herrera, & Castro,2007), Lebanon (El-Hadad & Balian, 1998; Keenan, El-Hadad, & Balian, 1998), Trinidad and Tobago (Hadeed & El-Bassel, 2006), Hong Kong (Chan et al., 2008), and Jordan (Clark, Silverman, Shahrouri, Everson-Rose, & Groce, 2010), have documented that involvement by in-laws is a potent risk factor for marital conflict and intimate partner violence.

Marriage within kinship networks (consanguineous marriage) is likewise high in Turkey as it is in many other Asian countries. In Asia as a whole, between 20% and 50% of women marry within their extended family system (Bittles, 1994). Although consanguineous marriage appeared to provide protection from violence in a study from Jordan, Clarke and colleagues recommended further investigating this phenomenon in the Jordanian context (Clark et al., 2010). As the authors point out, the interference of in-laws in the marital relationship is high in Jordan, especially in marriages with strong kin networks. In Turkey, elders also come into play via the berdel system—a marriage where two sisters are married into the same family. Traditionally, the berdel system was used as a way to avoid paying bride price and to strengthen ties between families.

Educational level

Completing secondary school has been found to be protective among women, reducing the likelihood of partner violence in a range of countries, including Uganda and Peru (Flake, 2005; Koenig et al., 2003). Studies conducted in the Middle East and North Africa have yielded similar results (Boy & Kulcyzki, 2008) as have previous studies in Turkey (Altinay & Arat, 2009; Balci & Ayranci, 2005; Kocacik, Kutlar, & Erselcan, 2007; Mayda & Akkus, 2004; Sahin & Sahin, 2003).

Family-related factors

Partner violence has also been linked to a number of factors related to the family and family dynamics. For example, there is a strong association between violence and the isolation of a woman from her natal family and close friends (Elsberg et al., 2000; Heise, 1998; Neilson et al., cited in Jewkes, 2002). Studies likewise suggest that violence is more prevalent in families where the husband independently makes the majority of the important decisions (Garcia-Moreno, 2005; Heise, 1998). Egalitarian relationships appear to have the lowest rates of relationship conflict and partner violence, an observation affirmed in studies undertaken in Haiti, the Philippines, and South Korea (Flake, 2005; Gage, 2005; Kim & Emerty, 2003)

Household characteristic

The common assumption is that women who live in poor households experience more violence than those who live in wealthier households (Hoffman, Demo, & Edwards,1994), although the reality has proven more complex (Vyas & Watts, 2009). Economic problems (Jewkes, 2002), insufficient household goods (Carlson, 1984, cited in Flake, 2002), and poverty (Elsberg et al., 1999) were found to be one of the main risk factors for domestic violence in South Africa, the United States, and Nicaragua, respectively. However, findings from other countries have varied. Among nine DHS countries, for example, there was no consistent relationship between risk of violence and the poverty/wealth status of the households in which a woman lives. In countries where there was a significant relationship (India, Egypt, Peru), women in the wealthiest quintile appeared to be protected (Kishor & Johnson, 2004). Besides the economic conditions of the household, the relative contribution of the woman and man to household income has also been linked to the prevalence of partner violence (Hindin & Adair, 2002; Mcloskey, Williams, & Larsen, 2005; Schuler, Hashemi, Riley, & Akhter, 1996; Vyas & Watts, 2009). For example, partner’s contribution of income to the household was negatively associated with the risk of domestic violence for women in Tanzania (Mcloskey et al., 2005), whereas women in Bangladesh who contributed more than nominally to the household expenses were 1.8 times more likely to experience violence than those who contributed very little or none (Bates, Schuler, Islam, & Islam,2004).

Current situation of couples

Some studies have indicated a positive relationship between women’s employment and domestic violence, and some others found either a negative relationship or no relationship at all (Kalmuss & Straus, 1990; Vyas & Watts, 2009). One study in rural South India found that women who had higher incomes than their partners experienced less domestic violence (Rao, 1997), whereas another study among young low-income women in Bangalore found they experienced more violence (Krishnan et al., 2010). A recent prospective study found that women whose partners become unemployed between the first and second visits of the study had 1.7 times the odds of experiencing violence compared to women whose partner remained employed (Krishnan et al., 2010). Panda and Agarwal (2005) found that possession of land or property reduced the risk of violence from one’s husband. They suggest that ownership of property gives women more security because it offers protection against fluctuations in the labor market and communicates that she has options outside of marriage. They recommend improving women’s property rights as one strategy to reduce partner violence. Patriarchal family structures that perpetuate male dominance in the family and women’s attitudes toward male authority are also related to domestic violence (Heise, 1998). Dibble and Strauss’s study of family violence asserts that the attitudes and behaviors are related to each other (Dibble & Strauss, 1980). Many studies have found attitudes accepting male dominance or partner violence to be a risk factor for experiencing violence. In particular, the studies mentioned in Boy and Kulzcycki’s paper on intimate partner violence in the Middle East and North Africa emphasized attitudes of women and men (Boy & Kulzcycki, 2008).

Drinking alcohol has been recognized as a risk factor for intimate partner violence in many studies (Fagan, 1993; Hotaling & Sugarman, 1986, as cited in Heise, 1998). When the husband is unfaithful and also frequently comes home drunk, women are at increased risk of partner violence (Flake, 2005; Jewkes et al., 2002; Koenig et al., 2003; Rao, 1997). Alcohol use is seen as related to marital conflict and as a facilitator of violence, even though the degree to which alcohol is a causal factor in domestic violence is still debated (Gil-Gonzales, Vases-Cases, Dardet-Alvarez, & Latour-Perez, 2006).

Violence from nonpartners, in particular from families (including both birth family and in-laws), was associated with increased domestic violence for women (Clark et al., 2010; Raj, Livramento, Santana, Gupta, & Silverman., 2006). It should be taken into consideration that the reporting of violence from family members increases reporting of partner violence (Clark et al., 2010). In one study, in-law conflict was found to be the strongest factor associated with increased violence (Chan et al., 2008). Obtaining support from family members was an important factor in reducing violence from intimate partners. Women who obtain a higher level of support from their families reported less intimate partner violence in some studies (Clark et al., 2010; Naved & Persson, 2008).

Community factors

Communities affect the likelihood of violence by protecting or insulating women. A few studies have looked at community factors. Community variation was measured by region in Flake’s study in Peru (Flake, 2005), and community characteristics were found to be principally determined by the social environment of that community, its beliefs, and its cultural factors. Rural women were found to be at more risk in the Middle Eastern countries (Boy & Kulczycki, 2008).

Method

Study Sample

This study is based on data collected as part of the National Research on Domestic Violence against Women in Turkey (Turkyilmaz, Ergocmen, Yuksel, Coskun, & Adali, 2009). The quantitative component of the research included a cross-sectional, household survey administered face to face to a nationally representative sample of 24,048 households. The study interviewed 12,795 women aged 15 to 59 years, using a stratified, multistage cluster sampling scheme designed to be representative at the national, urban-rural, and regional level (5 regions and 12 regions). More detailed information about the sampling can be found elsewhere (Turkyilmaz et al., 2009). After 2 weeks of training, 189 staff worked as interviewers, field editors, supervisors, and data entry staff.

Questionnaires

The survey included both household and individual respondent questionnaires. Women were asked about their educational and family background and also about their experiences of physical, sexual, emotional, or economic abuse from an intimate partner. In addition, the questionnaire collected information on sexual and physical violence from nonpartners, child sexual abuse, attitudes and norms on gender, violence during pregnancy, and coping strategies regarding violence. The response rates for household and individual interviews were 88% and 86.1%, respectively. The principles of the WHO ethical guidelines for domestic violence against women were strictly followed throughout the study (WHO, 2001). The survey used the innocuous title, “Turkey Women and Family Research,” to avoid alerting the wider community to the nature of the survey and to protect respondents and interviewers from possible harassment or violence. The sampling unit was households, and only one randomly selected woman was interviewed per household. The interviews were conducted in complete privacy after obtaining the respondent’s informed consent.

Outcome Variable: Physical or Sexual Violence by an Intimate Partner

The Turkish study adopted the same questions for measuring physical and sexual violence by an intimate partner as used in the WHO multicountry study. For physical violence, women were asked whether they had ever experienced any of seven specific acts perpetrated by an intimated partner, including having been slapped, pushed, or shoved; had something thrown at them that could hurt them; and punched, kicked, choked, or been threatened by or had a weapon used against them. For each act, they were asked how frequently it had happened within the past 12 months (once, a few times, many times) and how frequently they had experienced it prior to the past 12 months.

To measure sexual violence from an intimate partner, women were asked about acts of forced sexual intercourse, coerced sexual relationships, or forced participation in sexual acts that they found degrading or humiliating (Garcia-Moreno et al., 2005). They likewise were asked about the frequency of sexual violence within the past year and prior to the past year. Our primary outcome variable, as a subgroup of the main sample, was ever-married women aged 15 to 49 who have experienced one or more acts of physical and/or sexual partner violence by a current or former husband in the 12 months prior to the survey. The reference category for this analysis is ever-married women who have never experienced husband violence or who have experienced husband violence prior to the past 12 months. In addition, women who had experienced other forms of violence may appear in the reference category.

Exposure variables

Among the explanatory variables, the construction of country-specific and newly derived variables is described; other categories can be seen in Table 1. The data on childhood sexual abuse of women were obtained by directly asking the women about experiences of childhood abuse and also by giving women an opportunity to disclose anonymously. At the end of the interview, women were given a second opportunity to disclose child sexual abuse by marking either a “crying” or a “smiling” face to indicate their abuse status. For the purposes of the study, women were counted as sexually abused in childhood, if they indicated through either method a history of abuse.

Table 1.

Descriptive of Explanatory Variables and Logistic Regression Results for Ever-Married Women Aged 15 to 49

	Total	Experience of violence	Crude odds ratios (age adjusted)			Final model
Variables	%	%	OR		95% CI (confidence interval)	OR		95% CI
Personal abuse history
Childhood sexual abuse of women**				**
No	88.9	14.6	1.00			1.00
Yes	6.6	28.8	2.27	**	(1.82, 2.84)	1.39	*	(1.02, 1.90)
No answer	4.5	16.2	1.28		(0.94, 1.73)	0.96		(0.66, 1.40)
Witnessing of women#x02019;s mother#x02019;s physical abuse				**			**
No	68.7	11.0	1.00			1.00
Yes (see)	23.3	25.4	2.78	**	(2.35, 3.28)	1.59	**	(1.26, 1.99)
Yes (not see)	2.9	19.6	1.96	**	(1.27, 3.02)	1.01		(0.63, 1.63)
Don’t know	5.1	20.5	2.29	**	(1.75, 2.99)	1.40		(0.98, 2.00)
Physical abuse of husband during childhood				**			**
No	61.9	10.0	1.00			1.00
Yes	23.5	26.8	3.34	**	(2.83, 3.94)	1.76	**	(1.37, 2.26)
Don’t know	14.7	17.7	2.02	**	(1.65, 2.48)	1.27		(0.98, 1.65)
Husband*#x02019;s witnessing of his mother’s physical abuse				**			**
No	52.2	9.7	1.00			1.00
Yes (see)	19.6	26.6	3.41	**	(2.89, 4.01)	1.71	**	(1.34, 2.17)
Yes (not see)	6.9	19.1	2.12	**	(1.63, 2.75)	1.68	**	(1.27, 2.23)
Don’t know	21.2	16.2	1.83	**	(1.54, 2.19)	1.35	**	(1.10, 1.66)
Family formation
Educational difference				**
No difference	46.7	14.8	1.00			1.00
Woman has more education	15.3	11.9	0.72	*	(0.56, 0.93)	0.82		(0.60, 1.12)
Woman has less education	38.0	14.8	1.13		(0.97, 1.31)	0.94		(0.79, 1.11)
Age difference				**
About the same age	22.8	13.1	1.00			1.00
Woman is older	5.7	16.1	1.43	*	(1.04, 1.98)	1.34		(0.93, 1.94)
Man is older by 2-4 years	28.8	14.9	1.09		(0.90, 1.33)	1.00		(0.81, 1.24)
Man is older by 5-9 years	33.1	15.5	1.06		(0.86, 1.31)	0.91		(0.71, 1.16)
Man is older by 10 years	9.7	15.8	1.12		(0.86, 1.46)	0.83		(0.61, 1.13)
Age at marriage				**
< 7 years old	1.6	21.4	0.73		(0.38, 1.41)	1.00
17 > years old	98.4	14.9	1.00			0.64		(0.28, 1.44)
Marriage decision				**			**
Joint decision of couples	44.6	11.9	1.00			1.00
Decided by families or other	55.4	17.6	1.78	**	(1.53, 2.08)	1.41	**	(1.16, 1.72)
Marriage type				**
Both civil and religious	96.3	14.8	1.00			1.00
Only religious	3.7	23.6	1.47	*	(1.05, 2.05)	0.85		(0.56, 1.27)
Consanguinity
No	78.8	14.9	1.00			1.00
Yes	21.2	15.8	1.06		(0.90, 1.24)	0.89		(0.72, 1.10)
Bride price				**
No	87.7	12.4	1.00			1.00
Yes	12.3	22.2	1.73	**	(1.44, 2.07)	1.02		(0.81, 1.29)
Others at home at beginning of the marriage				**
No	40.1	11.4	1.00			1.00
Yes	59.9	17.5	1,69	**	(1.69-2.01)	1.11		(0.90, 1.38 )
Marital relationship
Husband’s affair with another woman				**			**
No	89.6	13.1	1.00			1.00
Yes	10.4	32.6	3.47	**	(2.88, 4.19)	1.82	**	(1.41, 2.37)
Preventing contact with family and friends				**			**
No	84.9	11.1	1.00			1.00
One of them	8.8	29.1	3.16	**	(2.50, 3.98)	1.97	**	(1.50, 2.58)
Both	6.3	49.8	7.92	**	(6.42, 9.76)	3.82	**	(2.87, 5.08)
Controlling her behaviors
No controlling behaviors	17.0	6.3	1.00	**		1.00
One or two	50.4	10.2	1.63	**	(1.22, 2.17)	1.19		(0.87, 1.64)
Three or more	32.6	27.2	5.07	**	(3.99, 6.45)	2.15	**	(1.61, 2.88)
Number of children				**			*
No live births	9.1	14.3	1.00			1.00
1 child	21.5	14.2	1.15		(0.88, 1.51)	1.15		(0.83, 1.58)
2 children	34.7	13.4	1.61	*	(1.12, 2.33)	1.28		(0.82, 2.01)
3 or more children	34.7	17.5	2.78	**	(1.99, 3.90)	1.62	*	(1.08, 2.42)
Marital duration				**
15 or more years	40.7	12.0	1.00			1.00
5-14 years	38.3	16.7	0.86		(0.68, 1.09)	1.05		(0.78, 1.41)
0-4 years	21.0	17.5	0.60		(0.83, 1.86)	1.09		(0.74, 1.61)
Current situation of couples
Employment status of women				**
Working	27.9	13.9	1.00			1.00
Not working	72.1	15.5	1.04		(0.89, 1.22)	0.97		(0.77, 1.24)
Ownership of house or land				**
Yes	20.1	16.2	1.00			1.00
No	79.9	10.7	1.39	**	(1.12, 1.71)	1.16		(0.92, 1.48)
Raising money for 4 weeks in case of need				**
Yes	45.5	11.5	1.00			1.00
No	52.5	18.3	1.78	**	(1.50, 2.09)	1.10		(0.89, 1.37)
Don’ know	2.0	11.4	1.04		(0.61, 1.78)	1.17		(0.61, 2.25)
Support from family members				**			*
Yes	70.1	12.6	1.00			1.00
No	29.9	20.8	2.07	**	(1.76, 2.43)	1.28	*	(1.04, 1.58)
Family members live nearby				**
Yes	69.2	17.2	1.00			1.00
No	30.8	14.1	1.25	**	(1.07, 1.46)	1.03		(0.85, 1.24)
Accepting traditional gender roles				**
No acceptance	14.2	14.7	1.00			1.00
Acceptance only one	38.3	15.5	1.05		(0.83, 1.34)	1.18		(0.91, 1.54)
High acceptance	47.5	14.9	1.08		(0.86, 1.36)	1.31		(0.99, 1.73)
Accepting male authority				**			**
No acceptance	26.8	11.4	1.00			1.00
Acceptance of 1-2 acts	44.9	14.3	1.28	**	(1.07, 1.53)	0.98		(0.78, 1.23)
Acceptance of 3 or more acts	28.4	19.9	1.99	**	(1.65, 2.41)	1.27	*	(1.01, 1.60)
Employment status of husband				**
Working	89.0	14.8	1.00			1.00
Not working	11.0	17.8	1.45	**	(1.15, 1.82)	1.01		(0.76, 1.33)
Drunkenness of husband				**			**
Never drunk	91.4	13.7	1.00			1.00
Rarely drunk	4.0	28.4	3.46	**	(2.41, 4.98	2.60	**	(1.67, 4.05)
Frequently drunk	4.6	31.5	2.54	**	(1.86, 3.47)	2.89	**	(1.96, 4.26)
Physical abuse from nonpartners				**
No one	83.5	12.9	1.00			1.00
Natal family	12.7	24.3	2.06	**	(1.73, 2.45)	1.27		(0.99, 1.63)
In-law family	2.0	40.3	5.19	**	(3.61, 7.48)	1.73	*	(1.03, 2.90)
Others	1.8	19.4	1.66	*	(1.01, 2.72)	1.24		(0.72, 2.15)
Sexual abuse from nonpartners				**			*
No one	97.5	14.4	1.00			1.00
Natal and in-law family	0.5	51.7	5.83	**	(2.51, 13.6)	2.13		(0.70, 6.50)
Others	2.0	35.0	2.96	**	(2.03, 4.32)	2.03	**	(1.21, 3.42)
Age							**
45-49	13.4	9.0				1.00
40-44	15.5	12.1				1.39		(0.97, 1.98)
35-39	16.4	12.7				1.48	*	(1.03, 2.13)
30-34	18.5	13.9				1.89	**	(1.33, 2.68)
25-29	19.5	18.9				2.72	**	(1.85, 3.98)
20-24	13.7	20.7				3.11	**	(1.94, 4.98)
15-19	3.0	24.2				4.92	**	(2.57, 9.43)
Household factors
Wealth level				**
High	21.7	9.1	1.00			1.00
Medium	42.5	13.9	2.30	**	(1.83, 2.91)	1.02		(0.78, 1.34)
Low	35.8	20.1	1.51	**	(1.18, 1.93)	1.07		(0.82, 1.39)
Contribution of family income				**			*
Almost the same	4.5	6.4	1.00			1.00
Women has no income	80.6	15.6	2.40	**	(1.39, 4.16)	2.53	*	(1.17, 5.51)
Woman only earner/earns more	3.7	11.0	1.90		(0.94, 3.82)	1.53		(0.61, 3.82)
Men earns more	11.3	16.1	2.74	**	(1.47, 5.09)	2.96	**	(1.31, 6.69)
Community factors
Region				**			**
West	43.8	10.6	1.00			1.00
South	12.8	16.1	1.64	**	(1.28, 2.09)	1.63	**	(1.25, 2.12)
Central	21.5	16.3	1.63	**	(1.35, 1.96)	1.22		(0.98, 1.53)
North	6.4	13.2	1.31		(0.97, 1.77)	1.20		(0.80, 1.79)
East	15.5	25.9	2.75	**	(2.33, 3.23)	1.99	**	(1.56, 2.54)
Type of place of residence				**
Urban	76.2	14.8	1.00			1.00
Rural	23.8	16.1	1.14		(0.98, 1.31)	0.88		(0.74, 1.06)
Total	9019	15.0
Wald						19.49	**
Nagelkerke R²						0.288

p < .05. **p < .01.

Age at marriage was measured as less than or greater than 17 years, since 17 is the minimum age of legal marriage for both women and men in Turkey. Age difference between spouses was considered “almost the same age” if the partners’ age difference was approximately an year. In the survey, women were asked about their experience of several different situations relating to their husband’s behavior. “Husband drunkenness” was coded as coming home drunk “rarely” (once a month or less) or “frequently” (once a week or more) in the 12 months before the survey. We also derived two new variables using factor analysis, one called “controlling her behavior” and a second called “preventing contact with family and friends.” The controlling behaviors of husbands included, to a large extent, the variables of “insisting on knowing where she is at all times; deciding what she wears; controlling when she may seek health care; getting angry when she speaks to another man, and suspecting her of infidelity.” The preventing contact with families and friends variable included the variables “try to keep her from her seeing her friends” and “try to restrict/prevent contact with her family of birth.”

The other variables formed by using the factor analysis technique were “acceptance of male authority” and “accepting traditional gender roles.” Acceptance of male authority means perceiving the following statements as acceptable: It is necessary to beat children for disciplinary purposes, it is right to beat women in some situations, a woman has an obligation to have sexual intercourse with her husband, a woman has an obligation to obey her husband, and a man is responsible for his wife’s behavior. The traditional gender roles variable was measured with the following questions: “Can a woman spend her money as she wishes?” and ‘Should men do housework?” These two variables were coded according to the number of accepted statements. To measure household socioeconomic position, we followed the practice outlined by Filmer and Pritchett for use in DHS surveys (Filmer & Pritchett, 2001), to derive a wealth index based on household assets and dwelling characteristics (DGSW, 2009).

Data Analysis

Because of the many factors affecting women’s lives, we explored a large the number of potential explanatory variables. These variables were sorted into six groups, and the analysis was performed in three phases. In the first phase, we checked collinearity among the explanatory variables, and no collinearity was detected. Next, we calculated the crude associations between each exposure variable and husband’s current violence, adjusted for the woman’s age. Although some variables were not independently associated with current violence, we nonetheless included them in the multivariate analysis, since multivariate analysis may yield different results. Finally, multivariate logistic regression analysis was performed by using PASW (SPSS)-18. The complex sample design of the survey was considered, and a “complex samples” module was used to perform the regression analysis. Table 1 presents the odds ratios and confidence intervals, taking into account the clustered nature of the data.

Results

Table 1 summarizes the primary findings of the study. The first column provides a breakdown of the population by response category for each of the exposure variables. The second column shows the percentage breakdown among women who experienced violence in the past 12 months by their current or most recent husband. Following the percentages, the odds ratios and confidence intervals for both the crude and final models are given. The final model is significant (with p < .001), and the Nagelkerke R² is 29%.

Overall, 15% of ever-married women aged 15 to 49 experienced physical or sexual violence by their current or most recent husband in the past 12 months; 6.6% of women reported childhood sexual abuse (as revealed through either interview reports or the anonymous question technique using the childhood faces). Consistent with the findings from many other studies, we found that prior experience of sexual abuse in childhood significantly increased the odds that a woman would experience current physical or sexual violence by her husband (OR = 1.39, 95% CI: 1.02-1.9). In the final model, the odds of current husband violence were also 1.76 times higher among women whose husband was physically abused during childhood.

Witnessing marital violence during childhood on the part of either the woman or her partner increased the odds of partner violence, and more than one fourth of the women and men were aware of their mothers being abused by their fathers. Women who witnessed their own mothers physically abused had 1.59 times the odds of current spouse abuse, and women whose husbands witnessed their mother physically abused in their childhood had 1.71 times the odds of being physically or sexually abused by their husbands in the past 12 months.

In the typical Turkish family, women have an educational level equal to or lower than their husband and most husbands are 2 or more years older than their wives. Both civil and religious marriages are common. In the crude analysis, all of the family formation variables were found significantly related to current partner violence when adjusted for the woman’s age. Being more educated than one’s husband actually protected women from violence in bivariate analysis, whereas the practices of bride price, abduction, and berdel increased women’s risk. In age-adjusted crude analysis, women were also at increased risk in marriages where there was only a religious (not a civil) ceremony, where the marriage was decided by family members, and where others lived with the couple at the beginning of a marriage. In the final multivariate model, however, only having one’s marriage decided by others remained a significant risk factor for current spouse abuse.

Partner-related factors such as infidelity, preventing her from seeing family and friends, and other controlling behaviors put women at increased risk of domestic violence in both crude and multivariate analysis. In the final model, women whose husbands were unfaithful had almost double the risk of violence (OR = 1.82, 95% CI: 1.41-2.37). Moreover, a woman who reported experiencing three or more controlling behaviors was 2.15 times more likely to experience current husband violence, and a woman whose husband prevented her seeing her family and friends was 3.82 times more likely to be abused.

In the age-adjusted bivariate analysis, women from poorer households had 2.30 times greater odds of husband violence than did women from better-off households, but this difference disappeared in adjusted analysis. By contrast, women who had no income experienced high odds of partner violence in the final model (OR = 2.53, 95% CI: 1.17-5.51) as did women whose husbands contributed more to the household budget then they did (OR = 2.96, 95% CI: 1.31-6.69).

Labor force participation and possession of land or house are uncommon among women in Turkey. Only 45.5% of the women reported that they could raise money to support themselves and their children for 4 weeks in case of emergency. Women’s ownership of property and their ability to raise money were not significantly associated with partner violence in the final model, although they were significant in bivariate analysis. About 70% of women reported that they could count on their family and friends for support when they require assistance, and an equal percentage have family members living nearby. Women without such support had 1.28 higher odds of experiencing current violence in the adjusted model.

High acceptance of male authority also increased women’s odds of partner violence in both bivariate and adjusted analysis. Less than 10% of women reported drunkenness by husbands in the past year, but those whose husbands came home drunk frequently or rarely were at substantially increased risk of violence (OR = 2.89, 95% CI = 1.96-4.26, for frequently; OR = 2.6, 95% CI: 1.67-4.05, for rarely). Husband’s employment status did not appear to affect risk.

Furthermore, 13% of women experienced physical violence from their families, and 2.5% of women experienced sexual violence from nonpartners after the age of 15. Physical and sexual abuse from nonpartners was significantly associated with husband violence. Physical violence from her husband’s family increased a woman’s risk of violence from her husband 1.73 times in the final model. Similarly, women who experienced sexual abuse from others were at enhanced risk of abuse by their husbands.

Where women lived in Turkey also affected their overall risk of spouse abuse. Women who lived in the Eastern and Southern regions of Turkey had 1.99 and 1.63 times higher risk of current physical or sexual violence, compared to women who lived in the Western region. As commonly found in other studies, younger women aged 15 to 19 and 20 to 24 were likewise at increased risk of current husband violence compared to women in the 45 to 49 age group.

Discussion

This study aimed to explore the factors that put women at increased risk of current physical or sexual husband violence as well as the factors that might be protective. In recent years, studies researching parental abuse as a predictor of violence have increased in Turkey (Altinay & Arat, 2009; Haj-Yahia & Uysal, 2008; Kocacik & Dogan, 2006; Kocacik et al., 2007; Ozcakir, Bayram, Ergin, Selimoğlu, & Bilge, 2008). The present study confirms the finding of many earlier studies, that if either the woman or her current husband witnessed parental violence in childhood, the woman is at increased risk of current physical or sexual violence from her husband. Qualitative study results reveal that parental abuse may cause early marriage in order to escape from violence at home; at this stage, the couple do not know each other well, and this may lead to marital conflict and violence in the husband’s home (Kardam & Yuksel, 2009). Similarly, physical abuse of the husband during childhood was positively associated with his risk of perpetration in adulthood. Childhood sexual abuse of women may be underreported in the survey, since talking about sexuality and sexual abuse are sensitive topics in Turkey. All these findings indicate the importance of the early experiences of abuse in the etiology of violence in adult relationships. However, the reason why some men become perpetrators of violence while others do not, even though they experienced and witnessed abuse in childhood, requires further research as suggested in other studies (Altinay & Arat, 2009; Heise & Garcia-Moreno, 2002). Longitudinal studies and further qualitative work may contribute to understanding these relationships.

Even though status inconsistency in educational level between a woman and her partner was not associated with current physical or sexual violence in the final model, the fact that women with more education than their husband had lower rates of violence in bivariate analysis suggests a potential role for education in empowering women and helping them to overcome violence. Employment, like education, can be a source of empowerment for women. The labor force participation in Turkey is around 23.5% for women and 69.1% for men in general (TURKSTAT, 2009). According to the findings of this study, women who did not have an income or those whose husbands earn more than them were at greater risk of violence, suggesting that economic dependence may play a role in elevating the risk of partner violence. Marriage is one of the most important socially encouraged institutions in Turkey, and arranged marriages still predominate. Even though the consent of woman is obtained in most of these marriages, women whose marriages were arranged by their family (with or without their consent), who had a berdel marriage (sisters marrying into the same family,) or who were abducted were more likely to experience violence than women who chose their husbands. It is possible that such arrangements influence the risk of violence by increasing marital conflict or by decreasing women’s autonomy, thus giving them lower status in the extended family system.

Of all the factors examined, it was the characteristics of the husband that were most strongly associated with women’s current risk of partner violence. Even though only civil marriages are legal in Turkey, some men have a civil marriage and several religious marriages. Religious wedding ceremonies socially legitimize a man’s affairs while he is still married to another. If a woman would behave similarly, however, she would risk being killed in the name of honor (Kardam, 2005).

Physical or sexual violence is widespread in the country, but most women do not disclose their abuse. If they do share their experience, they generally prefer to confide in family members even though they often do not receive support (Ergocmen, Uner, Abbasoglu, & Gökcen, 2009). Violence against women is still seen as a private family matter in Turkey even though it is publicly recognized by the state. This study underscores the importance of families in a couple’s life and highlights the significant role played by the patriarchal family in the experience of violence. Witnessing parental violence and experiencing violence before marriage, support from the natal family, and experiencing violence from the husband’s family are all positively associated with marital violence. These risks increase when women are young and living in less developed regions of the country.

Therefore, to reduce violence against women, transformations in families and society should be considered. In addition, women’s status should be strengthened through education and greater participation in the labor force, since even though women and men have equal rights in law, this is not translated into the daily life of women. Women’s social conditions, as well as their economic independence, are important factors in breaking the silence of women against violence. In Turkey, the fact that support from the wife’s own family is protective and violence from her husband’s family is a risk factor highlights the contradictory role that families can play in marital violence.

Most studies to date principally rely on information from women who have experienced violence, which is a necessary first step to exposing the issue. However, there is a need to investigate and understand men’s attitudes and points of view toward violence; further research should, therefore, also focus on men. Longitudinal studies should be conducted to understand the dynamics of domestic violence and the temporal relationship between risk factors. In addition, regional differences within the country should be investigated and differences between developed and less developed regions should be further analyzed.

This study represents the largest population-based survey ever conducted on partner violence, but it has some significant limitations. The data of the study are cross-sectional; therefore, it is impossible to establish the temporal ordering of events. Moreover, information on the husband’s behavior was obtained from their female partners so there may be bias in the reporting. On the other hand, this is the first national representative study that investigates the risk and protective factors for current physical or sexual violence in Turkey, and it provides important data for further research.

Footnotes

Acknowledgements

The authors of this study would like to thank the field workers for their dedicated efforts and also women who shared their experiences. The first analyses of this study were performed during İlknur Yuksel’s postdoctoral visit to the Gender Violence and Health Center at the London School of Hygiene and Tropical Medicine, and the authors would like to thank Charlotte Watts for her useful comments and support in LSHTM.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received the following financial support for the research, authorship, and/or publication of this article: The national survey on domestic violence against women in Turkey was funded by the European Commission at the request of the Directorate General on the Status of Women and implemented by a three-partner consortium that is composed of ICON Institute, Hacettepe University Institute of Population Studies, and BNB Consultant.

Bios

>İlknur Yüksel-Kaptanoğlu, PhD, is an instructor at the Institute of Population Studies of Hacettepe University in Ankara, Turkey. Her research interests include demography, domestic violence, reproductive health, poverty, qualitative research techniques, and marriage migration.

Ahmet Sinan Türkyılmaz (Associate Professor, PhD) is a lecturer at Hacettepe University, Institute of Population Studies in Ankara, Turkey. His research areas mainly include demography, sample and survey design, survey analysis, population projections, and domestic violence.

Lori Heise is a lecturer at the Gender Violence and Health Centre of the London School of Hygiene and Tropical Medicine. Her research focuses on social determinants of partner violence, and the intersection of gender relationships, economic and social power, and HIV.

References

Abrahams

Jewkes

(2005). Effects of South African men’s having witnessed abuse of their mothers during childhood on their levels of violence in adulthood, American Journal of Public Health, 95(10), 1811-1815.

Agoff

Herrera

Castro

(2007). The weakness of family ties and their perpetuating effects on gender violence: A qualitative study in Mexico, Violence Against Women, 13(11), 1206-1220.

Altinay

Arat

(2009). Violence against women in Turkey. İstanbul, Turkey: Punto.

Balci

Y. G.

Ayranci

(2005). Physical violence against women: Evaluation of women assaulted spouses, Journal of Clinical Forensic Medicine, 12, 258-263.

Bates

L. M.

Schuler

S. R.

Islam

M. K.

(2004). Socioeconomic factors and processes associated with domestic violence in rural Bangladesh. International Family Planning Perspectives, 30(4), 190-199.

Bittles

A. H.

(1994). The role and significance of consanguinity as a demographic variable. Population and Development Review, 20(3), 561-584.

Boy

Kulczycki

(2008). What do we know about intimate partner violence in the Middle East and North Africa. Violence Against Women, 14(1), 53-70.

Chan

K. L.

Brownridge

D. A.

Tiwari

Fong

D. Y. T.

Leung

W.-C.

P. C.

(2008). Understanding violence against Chinese women in Hong Kong: An analysis of risk factors with a special emphasis on the role of in-law conflict. Violence Against Women, 14, 1295-1312.

Clark

C. J.

Silverman

J. G.

Shahrouri

Everson-Rose

Groce

(2010). The role of the extended family in women’s risk of intimate partner violence in Jordan. Social Science & Medicine, 70, 144-151.

10.

Dibble

Strauss

A. M.

(1980). Some social structure determinants of inconsistency between attitudes and behavior: The case of family violence. Journal of Marriage and Family, 42(1), 71-80.

11.

Directorate General on the Status of Women [DGSW]. (2008). Kadına Yönelik Aile içi Şiddete İlişkin Hukuksal Durum ve Uygulama Örnekleri [Law and application of law on violence against women]. Retrieved from http://www.ksgm.gov.tr/Pdf/02_KYAIS_Hukuksal_Durum_Kitabi.pdf

12.

Directorate General on the Status of Women [DGSW]. (2009). Domestic violence against women in Turkey. Retrieved from http://www.ksgm.gov.tr/tdvaw/doc/Main_report.pdf

13.

Ellsberg

M. C.

Pena

Herrera

Liljestrand

Winkvist

(1999). Wife abuse among women of childbearing age in Nicaragua. American Journal of Public Health, 89(2), 241-244.

14.

Ergocmen

B. A.

Uner

Abbasoglu

Gökcen

(2009). Coping with violence against women. In Domestic violence against women in Turkey. Retrieved from http://www.ksgm.gov.tr/tdvaw/doc/Main_report.pdf

15.

Filmer

Pritchett

(2001). Estimating wealth effects of without expenditure data- or tears: An application to educational enrollments in states of India. Demography, 38, 115-132.

16.

Flake

F. D.

(2005). Individual, family, and community risk markers for domestic violence in Peru. Violence Against Women, 11, 353-373.

17.

Gage

A. J.

(2005). Women’s experience of intimate partner violence in Haiti. Social Science & Medicine, 61, 343-364.

18.

Garcia-Moreno

Jansen

H. A. F. M.

Ellsberg

Heise

Watts

(2005). WHO multi-country study on women’s health and domestic violence against women. Geneva, Switzerland: World Health Organization.

19.

Gelles

R. J.

(1980). Violence in the family: A review of research in the seventies. Journal of Marriage and Family, 42(4), 873-885.

20.

Gil-Gonzales

Vases-Cases

Dardet-Alvarez

Latour-Perez

(2006). Alcohol and intimate partner violence: Do we have enough information to act? European Journal of Public Health, 16(3), 278-284.

21.

Hadeed

L. F.

El-Bassel

(2006). Social support among Afro-Trinidadian women experiencing intimate partner violence. Violence Against Women, 12(8), 740-760.

22.

Haj-Yahia

M. M.

Uysal

(2008). Beliefs about wife beating among medical students from Turkey. Journal of Family Violence, 23, 119-133.

23.

Heise

(1998). Violence against women: An integrated, ecological framework. Violence Against Women, 4, 262-290.

24.

Heise

Ellsberg

Gotteomeller

(1999). Ending violence against women (Population Reports Vol. 27, No. 4). New York, NY: The Johns Hopkins University School of Public Health.

25.

Heise

Garcia-Moreno

(2002). Violence by intimate partners. In Krug

E. G.

Dahlberg

L. L.

Mercy

J. C.

Zwi

A. B.

Lozano

(Eds.), World report on violence and health (pp. 87-122). Geneva, Switzerland: World Health Organization.

26.

Hindin

M. J.

Adair

L. S.

(2002). Who is at risk? Factors associated with intimate partner violence in the Philippines. Social Science & Medicine, 55, 1385-1399.

27.

Hindin

M. J.

Kishor

Ansora

L. D.

(2008). Intimate partner violence among couples in 10 DHS countries: Predictors and health outcomes (DHS Analytical Studies, No. 18). Calverton, MD: Macro International Inc.

28.

Hoffman

K. L.

Demo

D. H.

Edwards

J. N.

(1994). Physical wife abuse in a non-Western society: An integrated theoretical approach. Journal of Marriage and Family, 56, 131-146.

29.

Jensen

Thornton

(2003). Early female marriage in the developing world. Gender and Development, 11(2), 9-19.

30.

Jewkes

(2002). Intimate partner violence: Causes and prevention. Lancet, 359, 1423-1429.

31.

Jewkes

Levin

Penn-Kekana

(2002). Risk factors for domestic violence: Findings from a South African cross-sectional study. Social Science & Medicine, 55, 1603-1617.

32.

Kalmuss

D. S.

Straus

M. A.

(1982). Wife’s marital dependency and wife abuse. Journal of Marriage and Family, 44(2), 277-286.

33.

Kardam

(2005). The dynamics of honor killings in Turkey: Prospects for action. Ankara, Turkey: Population Association, UNDP and UNFPA.

34.

Kardam

Yuksel

. (2009). Perceptions about violence against women: Qualitative research results. In Domestic violence against women in Turkey. Retrieved from http://www.ksgm.gov.tr/tdvaw/doc/Main_report.pdf

35.

Keenan

C. K.

El-Hadad

Balian

S. A.

(1998). Factors associated with domestic violence in low-income Lebanese families. Journal of Nursing Scholarship, 30(4), 357-362.

36.

Kishor

Johnson

(2004). Profiling domestic violence: A multi-country study. Calverton, MD: ORC Macro.

37.

Krishnan

Rocca

C. H.

Hubbard

A. E.

Subbiah

Edmeades

Padian

N. S.

(2010). Do changes in spousal employment status lead to domestic violence? Insights from a study in Bangalore, India. Social Science & Medicine, 70(1), 136-143.

38.

Kocacik

Dogan

(2006). Domestic violence against women in Sivas, Turkey: Survey study. Croatian Medical Journal, 47, 742-749.

39.

Kocacik

Kutlar

Erselcan

(2007). Domestic violence against women: A field study in Turkey. The Social Science Journal, 44, 698-720.

40.

Koenig

M. A.

Lutalo

Zhao

Nalugoda

Wabrie-Mangen

Kiwanuka

. . . Gray

(2003). Domestic violence in rural Uganda: Evidence from a community-based study. Bulletin of the World Health Organization, 81, 53-60.

41.

Krug

E. G.

Mercy

J. A.

Dahlberg

L. L.

Zwi

A. B.

(2002). The world report on violence and health. Lancet, 360, 1083-1088.

42.

Little

Kantor

G. K.

(2002). Using ecological theory to understand intimate partner violence and child maltreatment. Journal of Community Health Nursing, 19(3), 133-145.

43.

Martin

S. L.

Moracco

K. E.

Garro

Ong Tsui

Kupper

L. L.

Chase

J. L.

Campbell

J. C.

(2002). Domestic violence across generations: Findings from Northern India. International Journal of Epidemiology, 31, 560-572.

44.

Mayda

A. S.

Akkus

(2004). Domestic violence against 116 Turkish Housewives: A field study. Women and Health, 40(3), 95-108.

45.

Naved

R. T.

Persson

L. A.

(2005). Factors associated with spousal physical violence against women in Bangladesh. Studies in Family Planning, 36(4), 289-300.

46.

Ozcakir

Bayram

Ergin

Selimoğlu

Bilge

(2008). Attitudes of Turkish men toward wife beating: A study from Bursa, Turkey. Journal of Family Violence, 23, 634-638.

47.

Panda

Agarwal

(2005). Marital violence, human development and women’s property in India. World Development, 33(5), 823-850.

48.

Raj

Livramento

K. N.

Santana

M. C.

Gupta

Silverman

J. G.

(2006). Victims of intimate partner violence more likely to report abuse from in-laws. Violence Against Women, 12(10), 936-949.

49.

Rao

(1997). Wife-beating in rural South India: A qualitative and econometric analysis. Social Science & Medicine, 44(8), 1169-1180.

50.

Rivera-Rivera

Lazcano-Ponce

Salmeron-Castro

Salazar-Martinez

Castro

Hernandez-Avila

(2003). Prevalence and determinants of male partner violence against Mexican women: A population-based study. Salud Publica de Mexico, 46(2), 113-121.

51.

Romito

Crisma

Saurel-Cubizolles

M. J.

(2001). Adult outcomes in women who experienced parental violence during childhood. Child Abuse and Neglect, 27, 1127-1144.

52.

Romito

Saurel-Cubizolles

M. J.

Crisma

(2001). The relationship between parents’ violence against daughters and violence by other perpetrators: An Italian study. Violence Against Women, 7, 1429-1463.

53.

Sahin

H. A.

Sahin

H. G.

(2003). An unaddressed issue: Domestic violence and unplanned pregnancies among pregnant women in Turkey. The European Journal of Contraception and Reproductive Health Care, 8, 93-98.

54.

Schuler

S. R.

Hashemi

S. M.

Riley

A. P.

Akhter

(1996). Credit programs, patriarchy and men’s violence against women in rural Bangladesh. Social Science & Medicine, 43, 1729-1742.

55.

Singh

Samara

(1996). Early marriage among women in developing countries. International Family Planning Perspectives, 22, 148-157.

56.

Straus

M. A.

Gelles

R. J.

(1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and Family, 48(3), 465-479.

57.

TURKSTAT. (2009). Retrieved from http://www.tuik.gov.tr/PreHaberBultenleri.do?id=8424

58.

Turkyilmaz

A. S.

Ergocmen

B. A.

Yuksel

İ.

Coşkun

Adali

(2009). Methodology. In Domestic violence against women in Turkey. Retrieved from http://www.ksgm.gov.tr/tdvaw/doc/Main_report.pdf

59.

Vyas

Watts

(2009). How does economic empowerment affect women’s risk of intimate partner violence in low and middle income countries? A systematic review of published evidence. Journal of International Development, 21(5), 577-602.

60.

World Health Organization. (2001). Putting women first: Ethical and safety recommendations for research on domestic violence against women (WHO/FCH/GWH/01.1). Geneva, Switzerland: Author.