The study was designed to be primarily quantitative, although a qualitative analysis of the data for some questions was possible and necessary.* Information was gathered on the following issues:

the nature and extent of economic, emotional, physical and sexual abuse;



the nature of the abuse that women consider to have been the most serious and damaging;



details about the abusers;



the impact of the incident on survivors;



women’s experiences of both formal and informal service providers;



the extent and nature of women’s fear of crime generally; and



women’s suggestions about the kind of policies and changes required to serve their needs appropriately.

The United Nations Declaration on the Elimination of All Forms of Violence Against Women describes gender violence as:

"Any act of gender based violence that results in, or is likely to result in physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion or arbitrary deprivations of liberty, whether occurring in public or private life."13

The notion that violence is gender-based derives from the recognition that women and men do not experience the same forms of violence. In many instances, crimes almost exclusively experienced by women (such as sexual assault and violence within relationships) are strongly influenced by prevailing socio-cultural beliefs about women and men. Social ideas about gender thus play a significant role in creating societies that report high levels of violence towards women.

As the United Nations definition is a broad one, potentially encompassing a wide range of acts, this study specified four forms of abuse: economic abuse, emotional abuse, physical abuse, and sexual abuse.

These four categories of abuse are not mutually exclusive — many occur in combination with one another.

A total of 269 women were interviewed in this study, They were recruited from helping agencies in the metropolitan areas of Durban, Cape Town and Johannesburg. A stratified random sampling procedure was used to select participants. The sample was stratified according to race, age and type of abuse, in the following way:

The decision to select two-thirds of women who had experienced domestic abuse was based on the more pronounced shortage of data on domestic violence as opposed to non-domestic violence, as well as the availability of participants in each category.

Within each of these preselected categories, women were randomly chosen for interviews. This means that, although the study sought to investigate four types of abuse (economic, emotional, physical and sexual), participants were not selected according to these abuse categories, or any other criteria.

The most important limitation of the sample is the use of service providers for the selection of participants. This means that all participants will have sought assistance from such agencies in the past. The sample is thus skewed in favour of those women who seek help after abuse, and excludes those who either choose not to have counselling, are unaware of services, or are unable to reach service providers for whatever reason.

Since only women who had actually experienced abuse were interviewed, the findings cannot be used to indicate the extent of violence against women in the population. The stratification according to domestic and non-domestic abuse further means that the results do not indicate the varying risk that women face of becoming victims of domestic and non-domestic violence.

Table 1: Study sample

Age	Domestic 18-30 years	Domestic 31+ years	Non-domestic 18-30 years	Non-domestic 31+ years	Total
African	21	23	13	14	71
White	23	32	7	8	70
Coloured	22	26	10	6	64
Asian	16	30	11	7	64
Total	82	111	41	35	269

Table 2: Demographics of participants

Age 18 -30 45.7 30+ 54.3

Household size Living alone 8.2 2-5 members 72.1 6+members 19.7

Marriage contract In community of property 77.2 Out of community of property 22.8

Marital status Married 42.8 Single 28.6 Divorced 15.6 Separated 7.8 Cohabiting 3.3 Widowed 1.9

Race African 26.4 White 26.0 Coloured 23.8 Asian 23.8

Education < 8 years 3.4 7-9 years 24.7 10-12 years 43.1 Tertiary diploma/certificate 15.4 Degree/post-graduate 13.1

Type of marriage Legal (magistrate) 34.3 Religious(Christian) 22.4 Religious (Hindu, Jewish, Moslem) 20.1 Combination 17.2 Traditional Africa 6.0

Monthly income None 19.7 <R500 4.2 R500-R1 000 15.5 R1 000-R2 000 15.5 R2 000-R3 999 23.5 R4 000-R5 999 11.3 R6 000-R11 999 8.4 >R12 000 1.7

Children in care None 27.9 1 child in care 20.4 2 children 30.1 3 children 13.4 4+ children 8.2

Source of income Formal 40.5 Unemployed 17.5 Remittance from partner 12.6 Formal self-employed 11.5 Informal self-employed 9.7 Informal 5.6 Pension 1.1 Disablility grants 0.7

The sampling frame and methodology, however, provide useful new information on the prevalence of economic and emotional abuse among the women sampled. Expectations were that the more ‘serious’ types of abuse such as physical and sexual abuse, would predominate in the survey, since women are more likely to seek assistance from helping agencies after these incidents. However, the results show the extent to which different types of abuse co-occur, as well as the high number of women reporting emotional abuse.

At the time of being interviewed, the demographics of participants were as follows:

Demographic data was only collected about women at the time of the interview, demographics could not be used to indicate risk factors for abuse. No limitation was placed on when the abuse occurred in the woman’s life. Participants could have experienced the abuse many years ago (at which time their marital status, area of residence, income, level of education and household size, for example, may have been different). The experience of those who do not or cannot receive assistance may be different from those in this sample.

Participants were not asked about their general experiences with service providers, but only about impressions with regard to the incident of abuse they considered most serious. The many attempts made by women to receive help were therefore not captured. In many cases, women had not sought help for the incident they described as the most serious. As a result, the study did not record that individual’s perceptions of each service provider. One field worker commented that a participant who had sought help for years, received only very poor assistance. But because she had not sought help for the worst incident, her views were not recorded. The questionnaire was designed in this way to enable the gathering of accurate information about service providers. Many women visited helping agencies after being abused on more than one occasion. Asking women about every experience would have been impossible, given the length of the questionnaire, time constraints and fatigue issues.

Participants were also not always sure of the difference between district surgeons and other health workers, as well as the difference between a psychiatrist, a psychologist and a counsellor.

How questions and answers were interpreted, was also affected by the fact that the questionnaire was written in English. For those participants whose second language was English, questions had to be modified in order for them to be understood. Direct translation of some of the questions and concepts was not always possible, and the responses from participants were also difficult to translate to English. This process is likely to have affected both their understanding of the questions, as well as their answers.

Finally, the questionnaire was long and comprehensive. It is possible that fatigue set in towards the end, affecting the depth of responses. In covering such a wide range of questions, detail around responses may also have been sacrificed. However, interviewers were trained to handle fatigue and trauma appropriately.

Some participants remained traumatised by their experiences and were unable to focus on one single event — to the extent that they struggled to report adequately or fully on their experiences. Other participants, however, were able to share ideas and experiences and were eager to participate. Field workers commented on the strength displayed by participants and how women continued with the interview despite their excessively traumatic experiences.

Field workers also reported that it was difficult to obtain consent from some women to participate in the research. They also reported differences when interviewing various survivors of abuse. Some of the Asian participants refused to allow information about sexual abuse to be recorded as it was deemed inappropriate. This highlights the extent to which these incidents are underreported, even in this study.