Violence Against Women http://vaw.sagepub.

com/

Interviewer Training in the WHO Multi-Country Study on Women's Health and Domestic Violence
Henrica A. F. M. Jansen, Charlotte Watts, Mary Ellsberg, Lori Heise and Claudia Garca-Moreno Violence Against Women 2004 10: 831 DOI: 10.1177/1077801204265554 The online version of this article can be found at: http://vaw.sagepub.com/content/10/7/831

Published by:
http://www.sagepublications.com

Additional services and information for Violence Against Women can be found at: Email Alerts: http://vaw.sagepub.com/cgi/alerts Subscriptions: http://vaw.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://vaw.sagepub.com/content/10/7/831.refs.html

>> Version of Record - Jul 1, 2004 What is This?

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

VIOLENCE ARTICLE AGAINST WOMEN / JulyJansen 2004 et al. / INTERVIEWER 10.1177/1077801204265554 TRAINING

Interviewer Training in the WHO Multi-Country Study on Womens Health and Domestic Violence
HENRICA A. F. M. JANSEN
London School of Hygiene and Tropical Medicine World Health Organization

CHARLOTTE WATTS MARY ELLSBERG LORI HEISE

Program for Appropriate Technology in Health

CLAUDIA GARCA-MORENO
World Health Organization

The importance of a sound research strategy for measuring and understanding violence against women cross-culturally is well recognized. However, the value of specialized interviewer training to attain these data is not always fully appreciated. This article describes interviewer selection and training in the World Health Organization (WHO) Multi-Country Study on Womens Health and Domestic Violence and highlights their importance. Such training ensures high-quality data and cross-country comparability, protects the safety of respondents and interviewers, and increases the impact of the study. Moreover, women are not only willing to share experiences with trained and empathetic interviewers but also find the interview a positive experience. Keywords: interviewer selection; specialized interviewer training; violence against women; intimate partner violence; household survey; research methods

The Beijing Platform for Action from the 1995 United Nations World Conference on Women recommends the promotion of research on the prevalence of different forms of violence against women
AUTHORS NOTE: The authors wish to acknowledge the country researchers of the WHO Multi-Country Study on Womens Health and Domestic Violence: Bangladesh: Ruchira Tabassum Naved, ICCDR,B; Safia Azim, Naripokkho; Abbas Bhuiya, ICCDR,B, Lars Ake Persson. Brazil: Lilia Blima Schraiber and Ana Flavia Lucas DOliveira, University
VIOLENCE AGAINST WOMEN, Vol. 10 No. 7, July 2004 831-849 DOI: 10.1177/1077801204265554 2004 Sage Publications

831

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

832

VIOLENCE AGAINST WOMEN / July 2004

(VAW), especially domestic violence, and into the causes, the nature, and the consequences of violence against women. In 1996, the World Health Organization (WHO) held an Expert Consultation on Violence Against Women that specifically recommended that WHO undertake epidemiological research on these issues (WHO, 1996). As a result, the WHO Multi-Country Study on Womens Health and Domestic Violence was initiated in 1997. This study involved the design and implementation of crosscultural research on the prevalence, health implications, and risk factors for domestic violence. Between 2000 and 2003, large population surveys were conducted in eight culturally diverse countries: Bangladesh, Brazil, Japan, Namibia, Peru, Samoa, Thailand, and the United Republic of Tanzania. The study has also been replicated in its entirety or in amended form in other countries: Chile, Ethiopia, Indonesia, New Zealand, and Serbia and Montenegro. There is a growing number of other international research initiatives that also use population-based surveys to estimate the prevalence of different forms of violence against women across countries and cultures. These include the World Surveys of Abuse in Family Environments (WorldSafe) supported by the International Clinical Epidemiology Network (INCLEN) and the International Violence Against Women Surveys conducted by the European Institute for Crime Prevention and Control, affiliated with the United Nations (HEUNI), the United Nations Institute for Crime Research (UNICRI), and Statistics Canada. In addition,
of Sao PauloFaculty of Medicine; Ivan Franca-Junior, University of Sao PauloSchool of Public Health; Carmen Simone Grilo Diniz, Feminist Collective for Health and Sexuality; Ana Paula Portella, SOS Corpo, Genero e Cidadania. Japan: Mieko Yoshihama, University of Michigan; Saori Kamano, National Institute of Population and Social Security Research; Tamie Kaino, Ochanomizu University; Fumi Hayashi, Toyo Eiwa Womens University; Hiroko Akiyama, University of Tokyo; Tomoko Yunomae, Japan Accountability Caucus. Namibia: Eveline January, Hetty Rose-Junius, and Johan Van Wyk, Ministry of Health and Social Services; Alves Weerasinghe, National Planning Commission. Peru: Ana Gezmes, Centro de la Mujer Peruana Flora Tristan; Nancy Palomino and Miguel Ramos, Universidad Peruana Cayetano Heredia. Samoa: Tina Tauasosi, Secretariat of the Pacific Community. Thailand: Churnrurtai Kanchanachitra, Kritaya Archavanitkul, and Wassana Im-em, Mahidol University; Usa Lerdsrisanthat, Foundation for Women. Serbia and Montenegro (not in the first round of WHO VAW Study but replicating the study methodology): Stanislava Otasevic, Silvia Koso, and Katarina Bogavac, Autonomous Womens Center; Dragisa Bjeloglav, Strategic Marketing. United Republic of Tanzania: Jessie Mbwambo and Gideon Kwesigabo, Muhimbili University College of Health Sciences; Joe Lugalla, University of New Hampshire; Sherbanu Kassim, University of Dar es Salaam. The authors comprise the Core Research Team of the WHO study.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

833

the Demographic and Health Surveys (DHS) and the Centers for Disease Control (CDC)-supported International Reproductive Health Surveys increasingly contain a number of questions or a module on violence against women as part of larger household surveys on a range of issues. All of these surveys offer the advantages of large sample size, efficiency of data collection, standardization of measurement instruments, and generalizability to a general or specific population. However, relatively little attention has been placed on the role of interviewer training in improving data quality and in minimizing the risk to study participants. The amount of time spent on training interviewers on violence varies greatly among the different research initiatives. As Ellsberg and colleagues (Ellsberg, Heise, Pea, Agurto, & Winkvist, 2001) have illustrated, the careful selection and training of interviewers, as well as ongoing technical and emotional support, can substantially affect levels of disclosure on partner violence. This article describes the procedures used for interviewer selection and training by the WHO Multi-Country Study on Womens Health and Domestic Violence (WHO VAW Study). It discusses the lessons learned from experiences with interviewer training and their implications for research on violence against women. The article is based on experiences from all the participating countries but uses quantitative results from surveys conducted in Brazil, Namibia, Peru, Serbia, and Thailand. OBJECTIVES AND METHODOLOGY OF THE WHO VAW STUDY
MAIN OBJECTIVES

The WHO VAW Study was developed to obtain reliable and comparable data within and across culturally diverse countries on the following:
the prevalence and frequency of different forms of physical, sexual and emotional violence against women, particularly that inflicted by intimate partners; the association of violence by intimate partners with a range of health outcomes;

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

834

VIOLENCE AGAINST WOMEN / July 2004

factors that may protect or put women at risk for intimate partner violence; and strategies and services that women use to deal with the violence they experience.

In addition to these objectives, the WHO VAW Study also had a number of process-oriented objectives, as follows:
to develop and test new instruments for measuring violence crossculturally; to increase national capacity among researchers and womens organizations working on violence; to increase sensitivity to violence among researchers, policy makers, and health providers; and to promote a new model of research on domestic violence involving womens organizations with expertise on violence against women and fully addressing safety issues and safeguarding womens well-being.

A major focus of the WHO VAW Study has been the collection of rigorously sound and internationally comparable quantitative data on intimate partner violence. Qualitative data have also been collected to inform the development of the questionnaire and its country adaptations as well as the interpretation of results. In most participating countries, the quantitative component of the study consisted of a cross-sectional population-based household survey conducted in two sites: the capital (or other large city) and one province with rural and urban populations. In each of these sites, a representative sample of approximately 1,500 women aged 15 to 49 was selected to participate in face-to-face interviews. Women were asked about experiences of physical and sexual violence inflicted by intimate partners as well as by other people. To ensure that the study was policy- and action-oriented, it was essential to establish a close partnership between WHO, international research organizations, local research institutions, and womens health and human rights organizations. Each countrys research team included carefully chosen researchers with the technical skills required to conduct a large household survey as well as representatives from womens organizations addressing violence against women. While the study was ongoing, all the research teams met annually to share experiences and lessons

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING TABLE 1 Questionnaire Structure for the WHO Multi-Country Study on Womens Health and Domestic Violence Section 1: Characteristics of the respondent and her community Section 2: General health status Section 3: Reproductive health Section 4: Information regarding children Section 5: Characteristics of current or most recent partner Section 6: Attitudes toward gender roles Section 7: Experiences of partner violence Section 8: Injuries due to violence Section 9: Impact and coping mechanisms used by women who experience violence Section 10: Non-partner violence Section 11: Financial autonomy Section 12: Anonymous reporting of childhood sexual abuse; respondent feedback

835

learned in the field, to discuss and agree on the next steps, to build ownership, and to promote South-to-South collaboration. Members of the WHO Core Research Team from WHO, LSHTM, and PATH-DC gave technical assistance during critical stages of the survey, in particular during interviewer training, field pilot, and key stages in the process of data entry and processing. The study has been described in detail by Garcia-Moreno and colleagues (Garcia-Moreno, Watts, Jansen, Ellsberg, & Heise, 2003).
QUESTIONNAIRE

A uniform core questionnaire with structured questions was used as the basis for all country studies (WHO, 2003). Country modifications related to either adding questions to explore country-specific issues or ensuring appropriate response categories. The questionnaire consisted of 12 sections (see Table 1); early sections collected information on less sensitive issues. More sensitive issues, including the nature and extent of partner and nonpartner violence, were explored in later sections once rapport between interviewers and respondents had been established. As in many other surveys of partner violence, estimates of the prevalence of different forms of violence against women were obtained by asking female respondents direct questions about their experience of specific acts of physical and sexual violence from any partner. Follow-up questions were asked regarding the timing and frequency of violence. Women were also asked about experiences

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

836

VIOLENCE AGAINST WOMEN / July 2004

of emotionally abusive acts and controlling behaviors by their partners and about whether they had ever been physically abused while pregnant. In addition, the questionnaire included questions about womens experiences of physical and sexual violence by non-partners since the age of 15 and of sexual abuse by others before the age of 15. SELECTION AND TRAINING OF INTERVIEWERS
INTERVIEWER RECRUITMENT

International research indicates that womens willingness to disclose violence is influenced by a variety of characteristics such as the sex, age, marital status, attitudes, and interpersonal skills of interviewers (Ellsberg et al., 2001; Koss, 1993; WHO, 2001). The WHO VAW study used female interviewers and supervisors, and their careful selection and appropriate training were of paramount importance. Criteria for selection of interviewers were established and applied across countries. These included being able to engage with people of different backgrounds in an empathetic and nonjudgmental manner, emotional maturity, skills at building rapport, and ability to deal with sensitive issues. Criteria regarding the age and background of interviewers depended on the setting. Given the complexity of the questionnaire, the interviewers were required to have more than a primary-level education. In all countries, more potential interviewers and supervisors were recruited for training than the study required. This enabled the country research team to maintain some flexibility and have the option not to hire all of the interviewers. The final selection was done during or after the training itself. The recruitment in Namibia illustrates this process. More than 300 applicants applied for the position of interviewer in Namibia. Selection criteria included proficiency in English, having a Grade 12 certificate, preferably survey experience, and being unemployed. From these applicants, 100 were interviewed, and some 40 were selected for training. These were broadly representative of the many different cultural/language groups in Namibia. Their ages ranged from 17 to 50, and many

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

837

were single mothers aged 20 to 30. About half of the selected women had previous experience interviewing for the DHS. Supervisors were selected prior to or early in the first week of training, so that they could receive specific training, assist in the interviewer training, and start to work with and support their interviewer team. Five data entry persons and five field editors were selected in the third week of training from the trainees. Multicultural sensitivity was an important consideration in the training. Toward the end of the training, trainees wrote a onepage anonymous essay about the positive and negative aspects of the training. This exercise showed that a few trainees had not been able to overcome cultural/racial prejudices, which would make them unsuitable to interview women belonging to certain cultural groups or to work in a team with supervisors or interviewers from a background different from their own. This issue was dealt with in a sensitive way during a session when trainees could voluntarily read and explain their own views, and this could be discussed. One person who was unable to overcome her prejudices was asked to withdraw from the study. Final interviewer selection was done on the basis of technical and personal skills as well as written peer recommendations. At the end of the training, all participants, including those who were not selected, received a certificate.
TRAINING PROGRAM

The previous experiences of the WHO Core Research Team members and of the International Research Network on Violence Against Women had indicated that research on domestic violence required additional training and support to what is normally provided to survey research staff. For this reason, a standardized 3week training for interviewers was developed by the WHO Core Research Team for use in all settings (see Table 2). This included a framework for training and a set of accompanying materials: a training facilitators manual; a PowerPoint presentation on gender and violence against women; a manual with a question-byquestion explanation of the questionnaire; and specific procedural manuals for interviewers, supervisors, field editors, and data processors.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

838

VIOLENCE AGAINST WOMEN / July 2004 TABLE 2 Goals of Interviewer Training

To increase sensitivity of participants to gender issues at a personal as well as a community level To develop a basic understanding of gender-based violence, its characteristics, causes, and impact on the health of women and children To understand the goals of the Multi-Country Study on Womens Health and Domestic Violence To learn skills for interviewing, taking into account safety and ethical guidelines To become familiar with the questionnaire and protocol of the WHO Multi-Country Study

In each country, the training was conducted by the country research team, in most cases assisted by a member of the WHO Core Research Team. Certain sessions were conducted, as needed, by local or national psychologists, representatives of womens antiviolence groups, and census experts. An outline of the 3-week training schedule is in Garcia-Moreno et al. (2003). Much of the first week of the training was devoted to sensitizing trainee interviewers on gender issues, helping them to develop a basic understanding of gender-based violence, its dynamics and causes, and its impact on the health and well-being of women and children. During this period, field-workers were encouraged to confront and overcome any biases, fears, and stereotypes they held about women experiencing violence. A number of specific exercises were developed for this purpose. Stories of actual survivors of violence were used in some sessions, and in a number of sites, sessions were conducted by a counselor or a psychologist. In most countries, interviewers visited a shelter and/or had a session with women who gave testimony of their experiences with violence. This investment of time and energy is important for many reasons. As well as helping to ensure that respondents disclosing violence during the interview are treated appropriately, extensive training helps ensure data quality, as disclosure of abuse is sensitive to a respondents perception of judgment or blame on the part of interviewers (Jewkes, Watts, Abrahams, Penn-Kekana, & GarcaMoreno, 2000).

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

839

EMOTIONAL SUPPORT AND SKILLS

Given the prevalence of partner violence against women worldwide, the WHO VAW Study recognized that many interviewers would either have experienced partner violence themselves or have been affected by it in some way. For this reason, the training was explicitly designed to provide an opportunity for interviewers to reflect on their own experiences with abuse. During the training process, the subject of violence was openly discussed, and research team members were given opportunities to discuss their own experiences and feelings with others or with one of the main country researchers. This was used not only to identify who may need support during the study but also to help interviewers reflect on the challenges and personal emotions associated with disclosing experiences of violence and its implications for the survey. At all times interviewers were informed that they had the option of withdrawing from the project if they found the training or the subject matter too difficult or sensitive. Another focus of the training was on developing skills to minimize any possible distress caused to respondents during the interview. Domestic violence is a sensitive and stigmatized issue, and women are often blamed for the violence they experience. In the questionnaire, all questions about violence and its consequences were phrased in a supportive and nonjudgmental manner (Liss & Solomon, 1996). Interviewers received training on how best to ask questions and how to respond if a respondent did become distressed. This included how to be empathetic and supportive, allowing the respondent time to take a break and giving her the opportunity to reschedule or terminate the interview. Although interviewers were encouraged to be supportive, it was made clear that they should not raise unrealistic expectations about the study and that they had not received sufficient training to take on the role of counselor. Interviewers were trained to give all respondents, irrespective of whether they disclosed experiencing violence, a list of services and possible providers of support in the community (often a pocket-size information card that could be easily hidden). Interviewers also spent time discussing and role-playing how to respond appropriately to women who became upset or requested additional assistance during or following an interview.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

840

VIOLENCE AGAINST WOMEN / July 2004

The questionnaire was designed to end all interviews on a positive note (Parker & Ulrich, 1990). At the end of an interview in which a woman had disclosed violence, the interviewer was trained to reinforce the respondents own coping strategies and to remind her that the information she had shared was important and would help other women. As the following quote from an interviewer from Peru illustrates, interviewers developed their own ways to support women who disclosed violence:
I would tell a woman who lived with violence that she should have faith and courage to keep going on, to fight for her children if she had any, and if not, to have the courage to face things. QUESTIONNAIRE TRAINING AND PILOT

The second week of the training focused on familiarizing interviewers with the questionnaire and giving them opportunities to practice using it. This included holding discussions about different sections of the questionnaire, using the question-by-question description of the questionnaire to address any queries on meaning or interpretation of questions or discrepancies in the translation, and conducting role-plays and field practice of different sections of the questionnaire. When possible, this included at least one practice interview with volunteers who had experienced violence, so that interviewers would have a chance to see how well and appropriately they reacted to women disclosing violence. Role-plays were also used to play out specific situations, such as how to introduce the study to a household, how to deal with interruptions by a partner, and how to support a distressed respondent. The third week of the interviewer training was used to pilot test the questionnaire and all field procedures, including logistics, safety measures, and supervisory procedures. During the pilot testing, problems that may be encountered in the field were discussed, including strategies to minimize nonresponse. The pilot test provided another opportunity to assess trainees on their skills as interviewers and team workers and played a decisive role in determining the final composition of interviewer teams.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

841

TRAINING OF SUPERVISORY STAFF

Survey coordinators and team supervisors attended the full interviewer training. Supervisors and editors received additional training on their role as supervisors (including helping with access to households), the sampling techniques to be used in the field (including the selection of households and respondents), mechanisms for supporting interviewers, following up on nonresponse, quality control procedures (including questionnaire checking, spot checks, and disciplinary actions), and responding to women or children requiring assistance. In a number of countries, data entry clerks also attended all or some sections of the interviewer training. Participating in the training helped them understand the complexity of the questionnaire and made them feel part of the team. This also facilitated later interactions between the data entry team and field staff and helped ensure that the data entry team felt able to request support if they became distressed while entering questionnaires.
FIELD SUPPORT

During the fieldwork, regular meetings were held with interviewers for emotional debriefing. In contrast to more technical meetings that focused on evaluating progress with data collection and other aspects of survey logistics, these debriefing sessions aimed to provide interviewers with an opportunity to discuss their own feelings about the interviews and were conducted by the country research teams and, in some cases, by professional counselors. The following comments of an interviewer from Peru show the range and complexity of feelings that arise:
I started to cry when I came out of the interview. I was so depressed, not because it was a case of violence, on the contrary, I had just interviewed a woman; she and her partner were so in love and so respectful to each other, you could tell they werent faking it. I felt bad because my life wasnt like this.

In most countries, opportunities for individual counseling were also provided if needed. Given the potentially distressing nature of research on violence and the memories that it may

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

842

VIOLENCE AGAINST WOMEN / July 2004

awaken among field staff, the country teams found these sessions to be essential for maintaining the morale and emotional wellbeing of staff during fieldwork (Ellsberg & Heise, 2002). A final evaluation was held in most sites on finishing the fieldwork. A common theme mentioned by interviewers and other research staff was how participating in the study had had a tremendous impact on their own lives. As the quotes below illustrate, many interviewers felt that the training and field experiences had opened their eyes to the realities of womens lives and the types of violence that women face and had been a transforming experience. As a result, many have gone on to become involved in the antiviolence movement.
I grew a lot emotionally. I am much more secure and mature as a person. It gives me a sense of pride to have been part of the study. I feel we can give the government hard facts and statistics to create better services for women. (Interviewer from Namibia) After having lived an experience like this study, we will never be the same, not only because of what we heard but also for what we learned, for being recipients of many life stories, each one of them with different levels and degrees of violence. (Interviewer from Peru)

THE BENEFITS OF INTERVIEWER TRAINING The results of the studies have been disseminated nationally in most of the participating countries, and the cross-country comparisons will be formally launched by WHO in 2004. Even before the findings from the study are shared, evidence of the contribution of strong interviewer training can be seen from data on the duration of the interviews, the levels of respondent satisfaction with the interview, and the extent to which women who had previously not disclosed their experiences of violence to anyone reported violence during the study.
DURATION OF THE INTERVIEW

Results for four countriesBrazil, Namibia, Peru, and Thailand (in total, seven sites; two in each country, except for Namibia, where the survey took place in the capital only)show that

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

843

between 65% and 92% of interviews lasted less than 1 hour, and only 1% of interviews lasted more than 2 hours (4% in a Peru provincial site). Despite the relative length, once an interview was started, it would usually be completed (in six of the seven sites, less than 1% of eligible respondents did not complete the interview; this was higher only in the province in Peru where it was 1.6%). Interviews were generally completed in one session, with the exception of the capital of Peru, where a small proportion of interviews was conducted over two sessions. In response to a concern about the potential duration of the interview, in Peru additional information on the perceived duration of the interviews was collected. In the capital and the province, the median duration of an interview for women who did not report physical or sexual partner violence was 40 and 50 minutes, and for women who reported partner violence, 60 and 65 minutes, respectively. It is interesting to note that the median perceived duration of the interview was 20 and 30 minutes for women who did not report violence and 30 minutes (in both sites) for those who reported violence. This suggests that, even in the province, where the interviews were on average longer than in the capital, women who reported violence and, therefore, were asked more questions did not feel that the interview lasted much longer than women who had not been asked these extra questions.
DISCLOSURE OF VIOLENCE

Accurately estimating the rate of underreporting of violence is very difficult, if not impossible. Most researchers agree that, given the stigmatized nature of violence, overreporting of violence is not common, as women are unlikely to report acts of violence when they did not occur, and that, if anything, prevalence estimates tend to underestimate the true magnitude of partner violence (Koss, 1993). An indication of the extent to which the WHO VAW Study design and in-depth training and support of interviewers may have facilitated disclosure can be obtained by examining what proportion of women who disclosed violence during the survey reported never having previously spoken to anyone about their experiences.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

844

VIOLENCE AGAINST WOMEN / July 2004

0% Brazil City Brazil Province Namibia City Peru City Peru Province Thailand City Thailand Province
Figure 1:

10%

20%

30%

40%

50%

Percentage of Women Who Reported Physical Abuse by a Partner Who had Not Talked With Anyone About the Violence. WHO Multi-Country Study on Womens Health and Domestic Violence.

In the WHO VAW Study, across the seven sites in the four countries previously mentioned, between 21% and 46% of women who reported physical partner violence reported that they had not previously spoken with anyone about their situation. This suggests that for between one fifth and almost one half of the women reporting violence, the interviewer was the first person to whom she had disclosed violence (see Figure 1). The reasons given by respondents for not seeking help, as well as the verbatim comments of the respondents at the end of the interview, suggest that this was largely due to the stigma associated with domestic violence. In some cases, women only disclosed violence once they were absolutely sure that they could trust the interviewer, as illustrated by the following comment by an interviewer in Brazil:
I finished the questionnaire and the respondent had not mentioned a single occurrence of violence in response to the questions. Upon asking the question about how she felt about the interview, the woman told me, I see now that you are my sister, and she told about a situation of severe chronic domestic violence including death threats. (The information reported was subsequently entered on the questionnaire)

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

845

WOMENS SATISFACTION WITH THE INTERVIEW

An indication of the benefits of careful selection and training of interviewers can be obtained by assessing respondents satisfaction with the interview. At the end of the interview, respondents were asked the following question: I have asked about many difficult things. How has talking about these things made you feel? The answers were written down verbatim and coded by the interviewer into the following three categories: good/better, bad/ worse, and same/no difference. The majority (between 60% and 95% in seven sites) of women who had experienced physical or sexual partner violence reported that they felt good/better at the end of the interview. The range is quite similar between women who have or who have not experienced partner violence, except in Thailand where women who had experienced partner violence were more likely to report feeling better after the interview (60%) than those who did not report violence (43%). Very low percentages of respondents reported feeling worse after being interviewed: between 0.5% and 8% of women reporting partner violence ever (highest in Peru) and between 0% and 3% of women with no history of partner violence. As illustrated by the following quote from a woman interviewed in Japan, this sentiment was usually because it was difficult to revisit or to talk about painful events:
I was reminded of my experience of being sexually mistreated in the past, which I had forgotten about.

In all countries, the overwhelming impression from the study was that women were not only willing to talk about their experiences of violence but also were often deeply grateful for the opportunity to tell their stories to a nonjudgmental, empathetic person. The fact that so many women who had never discussed their experiences previously chose to do so in this context underscores how the quality of interpersonal communication between interviewers and respondents may enhance or inhibit disclosure. As a woman interviewed in Japan explained,
I have experienced violence, too. I did not know where I could go for help. [After receiving the list with services], I now know where I can go. I was looking for such places. It is good to address this type of issue in a survey. I am happy now. (Woman interviewed in Japan)

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

846

VIOLENCE AGAINST WOMEN / July 2004

THE VALUE OF TRAINING

Additional evidence of the value of the WHO VAW Study training approach comes from Serbia and Montenegro, the most recent country where the survey was conducted, between March and May 2003. With technical support provided by WHO, 13 previously inexperienced interviewers were fully trained over 2.5 weeks. However, because of pressure to finish the fieldwork, 6 weeks after the start of the fieldwork, an additional group of 21 professional interviewers from a survey company were recruited to assist with the interviews. This new batch of experienced interviewers received less than a days training, which included orientation on gender and violence issues and a brief review of the questionnaire and field procedures. In total, 1,445 complete interviews were conducted in Serbia and Montenegro, with 47% of the households being visited by the study-trained interviewers and 53% by the professional interviewers. This provided a unique opportunity to see how levels of participation and disclosure differed between the two groups. The previously inexperienced but carefully selected and trained interviewers obtained a significantly higher response rate (93% vs. 86%; p < .0001) and a significantly higher disclosure rate (26% vs. 21%; p < .05) for physical and/or sexual partner violence than the professional interviewers. The interviews done by study-trained interviewers were shorter than those done by untrained interviewers: The median duration was between 5 and 7 minutes less for every type of partner violence. Respondent satisfaction at the end of the interview was significantly higher for women, both with (46% vs. 29%; p < .01) and without violence (46% vs. 38%; p < .05) interviewed by the trained interviewers. These findings highlight the degree to which interviewer selection and training can affect levels of participation, disclosure, and satisfaction with the interview and illustrate that it is not advisable to assume less training is needed when using professional interviewers. DISCUSSION This article has described the critical importance placed on the careful selection and extensive training of interviewers in the

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

847

WHO Multi-Country Study on Womens Health and Domestic Violence (WHO VAW Study). In our view, these measures contributed substantially to the validity of our findings by enhancing disclosure as well as minimizing risks to respondents and interviewers. The experiences of the WHO VAW Study suggest that failure to provide special training and support to interviewers is a concern not only from the perspective of the safety of interviewers and respondents but also in terms of data quality. This conclusion is confirmed by studies in countries where prevalence estimates have been obtained using different interviewer training methods and by the analysis of data from Serbia and Montenegro, where the WHO VAW Study was replicated by simultaneously using interviewers who had received different levels of training. The experience from the WHO VAW Study suggests that a priority in interviewer selection is to identify women who are empathetic, have strong interpersonal skills, and are interested or concerned about these issues. In practice, the issues of high levels of education or prior survey research experience appear to be less important. Providing intensive and extended training for interviewers involves a considerable commitment of time, money, and other resources, and it is often considered to be impractical, particularly when resources are limited. In our experience, however, the overall benefits to the study of such an investment, in terms of enhanced data quality as well as participant and interviewer satisfaction, greatly outweigh the costs. Moreover, despite the numerous challenges associated with such research, the WHO ethical and safety guidelines underscore that the use of methods to optimize disclosure and to ensure the safety of respondents and staff is an ethical obligation for researchers. From our experience, four essential components of the interviewer training are to provide interviewers with (a) a basic grounding in gender and domestic violence, (b) opportunities to talk to and interview survivors of violence, (c) the opportunity to reflect on their own experiences of violence, and (d) the basic skills to respond to and deal with emotional distress resulting from the interviews in both respondents and themselves. During data collection, interviewers need to be supported with ongoing emotional debriefing sessions and referrals for individual counseling as well as the more technical supervision.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

848

VIOLENCE AGAINST WOMEN / July 2004

It has also been evident from our research that participation in the training and the survey has served as an important catalyst for the greater involvement of many interviewers in the antiviolence movement. The process of strengthening the links between researchers and activists also strengthens the capacity of local organizations to carry out evidence-based advocacy on violence against women and improves the research communitys ability to conduct action-oriented research on domestic violence. These partnerships are essential, as with a problem as pressing as violence against women, research efforts need to promote social change as well as collect high-quality information. REFERENCES
Beijing Declaration and Platform for Action, adopted by the 4th World Conference on Women, Beijing 4-15 Sept. 1995, UN Doc. ACONF.177/20 Rev. 1 (96.IV.13), para. 129a. Ellsberg, M., & Heise L. (2002). Bearing witness: Ethics in domestic violence research. Lancet, 359, 1599-1604. Ellsberg, M., Heise, L., Pea, R., Agurto, S., & Winkvist, A. (2001). Researching violence against women: Methodological and ethical considerations. Studies in Family Planning, 32, 1-16. Garcia-Moreno, C., Watts, C., Jansen, H., Ellsberg, M., & Heise, L. (2003). Responding to violence against women: WHOs Multi-Country Study on Womens Health and Domestic Violence. Health and Human Rights, 6, 113-127. Jewkes, R., Watts, C., Abrahams, N., Penn-Kekana, L., & Garcia-Moreno, C. (2000). Ethical and methodological issues in conducting research on gender-based violence in Southern Africa. Reproductive Health Matters, 8, 93-103. Koss, M. (1993). Detecting the scope of rape: A review of prevalence research methods. Journal of Interpersonal Violence, 8, 198-222. Liss, M., & Solomon, S. D. (1996). Ethical considerations in violence related research. Unpublished document. Parker, B., & Ulrich, Y. (1990). A protocol of safety: Research on abuse of women. Nursing Research, 38, 248-250. World Health Organization. (1996). Violence against women, WHO consultation (FRH/ WHD/96.27). Geneva, Switzerland: Author. 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. World Health Organization. (2003). WHO Multi-Country Study on Womens Health and Life Experiences. Core Questionnaire, Version 10. Geneva, Switzerland: Author

Henrica A. F. M. (Henriette) Jansen is an epidemiologist for the World Health Organization (WHO) Multi-Country Study on Womens Health and Domestic Violence in the Department of Gender and Womens Health of the WHO. She provides on-site technical support to all countries in the study and manages data

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011

Jansen et al. / INTERVIEWER TRAINING

849

collection and analysis. She has more than 20 years of experience in the areas of primary health care, child health, and epidemiology in South America, Asia, and Africa. Charlotte Watts is a senior lecturer in epidemiology and health policy in the Health Policy Unit, Department of Public Health and Policy, London School of Hygiene and Tropical Medicine. She leads programs of research on violence against women and on HIV/AIDS with a focus on low- and middle-income countries. She is currently senior technical advisor to the WHO V AW Study. Mary Ellsberg is the director of the Gender, Violence, and Human Rights Program of the Program for Appropriate Technology in Health (PATH), based in Washington, DC. She is also an adjunct researcher in the Department of Public Health and Clinical Medicine, Ume University, Sweden. In addition to the WHO V AW Study, she has participated in research on violence against women in Nicaragua, Indonesia, and Ethiopia. Lori Heise is director of the Global Campaign for Microbicides at the Program for Appropriate Technology in Health (PATH) and a research fellow in health policy at the London School of Hygiene and Tropical Medicine (LSHTM). She is a member of the core research team of the WHO V AW study and has worked for more than two decades on the intersecting issues of gender, power, sexuality, and violence. She is coauthor with Mary Ellsberg of Researching Violence Against Women: Practical Guidelines for Researchers and Advocates (WHO & PATH, in press). Claudia Garcia-Moreno is coordinator in the Department of Gender and Womens Health of the WHO. She has extensive international experience in public health, primary health care, and womens health. She is the coordinator of the WHO Multi-Country Study on Womens Health and Domestic Violence and a coauthor in the World Report on Violence and Health. She also chairs the Coordinating Group of the Sexual Violence Research Initiative, recently launched under auspices of the WHO and the Global Forum for Health Research.

Downloaded from vaw.sagepub.com by RAVI BABU BUNGA on October 23, 2011