Pursuing Authenticity From Process to Outcome in a Community-Based Participatory Research Study of Intimate Partner Violence and HIV Vulnerability in North Karnataka,India

Research Topic Selection

During meetings in June 2012 with UM and KHPT, Chaitanya leaders expressed an interest in being more involved in research and identified IPV as the most pressing issue to study. They determined the following as important research questions:

Development of Ongoing Researcher–Community Partnerships

This study was embedded within the wider organizational commitment of long-term partnerships between Chaitanya, KHPT, and academic organizations. Initial meetings were conducted in early 2013 with 15 leaders from Chaitanya to determine study objectives, criteria for selecting members of a community Research Committee, a timeline, and selection criteria for both potential participants and community research investigators (CRIs). A community Research Committee was then formed by women in sex work who volunteered to be involved with good verbal skills, knowledge on the issues, and some previous involvement in research. Seven women joined the Research Committee as a representative body for the CBO. After some attrition due to migration and time constraints in the first year, the remaining members of the Research Committee decided that consistent attendance would solidify membership, while new members would be selected democratically through votes by the CBO Board of Members.

Roles and Responsibilities

Important in the initial stages was the determination of the roles and responsibilities of all partners. Each of the partners expressed their expectations from the others, and agreed on what they could contribute so as to build on each other’s strengths, as Israel et al. have recommended (Israel et al., 1998). For example, the community Research Committee felt that they had knowledge in the topical areas, ethical and logistical considerations, and were experienced with recruiting participants for other studies. In return, the Research Committee members viewed training and guidance by the other partners on methods for data collection, analysis, and writing of reports as a renewable resource that would strengthen their involvement in this and future studies. Sometimes the diverse partnerships caused confusion and language barriers, but these were addressed through close involvement of program field staff from KHPT who could ensure the training processes were conducted in local dialects.

Interview Question and Tool Development

An important component of CBPR to which our team committed much time was training and knowledge exchange (Israel et al., 1998). The NGO and academic researchers provided the community Research Committee training on evaluation and CBPR research processes and methods at multiple points in the study. To begin, KHPT program directors and AKB held training workshops for the Research Committee on research principles and processes in July 2013 in Bagalkot, followed by workshops to develop the questions, tools, and ethical protocol.

Interview tool development involved multiple stages: We discussed a range of issues involved in understanding IPV; the Research Committee decided key areas that could form themes for the interview guide, and specific questions were constructed for each theme. Simultaneously, the researchers worked to streamline the tool with an eye toward the evaluation outcomes. The final interview guide included the following theme areas: (a) important relationships; (b) personal qualities, self-esteem, hopes, and dreams; (c) life problems and coping mechanisms; (d) intimate relationship history, feelings, expectations, relationship with family; (e) IPV experiences, norms/acceptance, and methods to address; (f) condom use experience and norms; and (g) exposure to the intervention and knowledge of laws on domestic violence. As part of the CRIs’ training, it was important to identify the precise questions in the local dialect through role-plays. For example, questions around women’s self-esteem were framed in terms of what they liked or disliked about themselves, rather than about good or bad personal “qualities” as in English. Through pilot-testing with two women in sex work, the CRIs and KHPT staff further refined the interview questions into a pictorial tool with brief prompts to allow for more natural dialogue.

Ethical Protocol and Approval

Numerous ethical issues are important in a CBPR study, such as cultural appropriateness of methods, feasibility of timelines, training for community partners or interviewers, and the joint use of data for program and research outputs (Buchanan et al., 2007; Campbell-Page & Shaw-Ridley, 2013; Jamshidi et al., 2014; Schaffer, 2009). We worked as partners to determine the ethical considerations for each stage, which was formed into a protocol for all to follow. Important elements that the Research Committee identified were (a) safe means of approaching participants and concealing their identities; (b) a sensitive and relevant interview guide; (c) participant counseling if requested, and debriefing with CRIs; (d) awareness by the CRIs of the neighborhood context; (e) interviews with men by an external interviewer, after consent by and separately from the FSW participant; and (f) participants’ honoraria equivalent to but not beyond travel costs, to ensure fair and voluntary participation. Although concerns were raised by some academic partners regarding couple interviews, recent experience has suggested that this could be conducted ethically if precautions were observed (Dandona et al., 2006; Enosh, Eisikovits, & Gross, 2013; Karandikar & Gezinski, 2012; Kostick, Schensul, Singh, Pelto, & Saggurti, 2011; Luke, Schuler, Mai, Vu Thien, & Minh, 2007; Panchanadeswaran et al., 2012; Sikweyiya & Jewkes, 2011). Research Committee members stated that interviewing both members of a couple separately was feasible because Devadasi women usually lived independently with their own incomes and many of their partners did not belong to the same village as them.

To obtain informed consent, the CRIs used a “vague first contact” approach toward potential female participants (Sikweyiya & Jewkes, 2011) in which the study was introduced generally as a health study to potential participants. If they showed interest, this was followed by a full informed consent process both for interviews and for contact with their intimate partners. A concern of the community Research Committee was that participants would benefit from doing the interviews, which is a key ethical consideration. As part of consent, the CRIs told the women that through the interview they may gain more courage or peace of mind, but it would ultimately help the CBO to move ahead in addressing the wider community issues related to IPV and HIV/AIDS.

To ensure confidentiality, all the interviews were one-on-one and the locations were determined based on the interviewees’ preferences (e.g., home, a neighbor’s home, a public place, or CBO office), with a code to stop talking if others approached. During analysis, interviews were de-identified and discussed individually between external researchers and the interviewers first, then aggregately with the Research Committee, and KHPT as a whole. Ethics approval for the study was obtained by all partners, including the St. John’s Medical College to KHPT in Bangalore, the LSHTM’s Research Ethics Board, and UM’s Human Research Ethics Board (H2014:108).

Training With CRIs

Chaitanya selected a pool of potential CRIs from within the FSW community. In December 2013, the CRI candidates were trained on qualitative research and interviewing skills. The workshop was followed by numerous role-plays and “fish-bowl” exercises to practice skills, where pairs of women practiced interviewing each other with a set of questions in front of the group. Observers provided feedback on the importance of open body language, facial expressions, clear questioning, probing, and the need to give time for longer answers. Through role-plays, the phrasing of the interview questions in Kannada was determined, and allowed increased familiarity with the questions by the CRIs. The final selection and hiring of two CRIs was made jointly by the Research Committee and academic researchers.

Sampling Strategy

The Research Committee suggested that important and ethical criteria for sampling included: participants from villages that the CRIs are not from; those who reported experience of chronic events of IPV in last 12 months; those having a single intimate partner (for the couples, to ensure clarity of their accounts and safety for participants); participants from different age groups(those both below age 25 and 25 or above); representation of home-based FSWs as well as all other typologies of sex work; and women both with and without children. The sample was selected by KHPT researchers in Excel from their “line-listing” of women registered in the Samvedana Plus intervention in 2012 to maximize variation on the abovementioned characteristics. In total, the CRIs and male RI interviewed 38 participants, including 13 individual FSWs, five individual intimate partners, and 10 couples (FSWs and their intimate partners), which will likely preserve a sufficient sample for further rounds.

Data Collection

The CRIs made initial contact and obtained informed consent. Though all partners agreed on the importance of maintaining confidentiality by ensuring the CRIs went to villages they were not from, this caused initial challenges for logistics and discomfort for the CRIs that they had to overcome through further training. The community Research Committee and KHPT staff were central in supporting the CRIs throughout, especially as it was a completely new process, involving extensive travel and perseverance to contact the interviewees. The main difficulties in data collection for the CRIs occurred in recruitment, mostly due to migration or deaths of the participants in the initial sampling frame, wariness on the part of the participant of being recorded, and fear of identifying as Devadasi if they were a part of the Government’s Devadasi Rehabilitation Programme. ¹ Male participants were more difficult to recruit due to transient work, suspicion around recording, and the topic of violence itself. After informed consent, interviewers completed one-on-one, semi-structured in-depth interviews using digital recorders, as well as taking written field notes before, during, and after each interview. Field notes included the process of consent, the manner in which the questions were answered, and any occurrences outside the recorded interview that provided insight (e.g., when an intimate partner interrupted, or body language in the interview). The most difficult aspect of the interviews was in broaching sensitive topics, particularly around condom use. The CRIs stated that if it had been an outside interviewer, the topic would have been difficult to broach. However they were able to assess through the way the participant answered whether they were being open or not. The field notes and interviews were transcribed verbatim by KHPT RIs and the CRIs, and translated by KHPT’s trained local translator who was familiar with the interview guide and data collection process. SGN read through each transcript for accuracy and completeness in English before analysis, consulting interviewers where needed.

Collaborative Data Analysis

Because few CBPR studies have reported on the involvement of community partners in interpreting results, we undertook a process of “meaning-making” that involved a novel combination of analytic methods in multiple steps (Cashman et al., 2008; Ellsberg & Heise, 2005; Lorway et al., 2014; Miles & Huberman, 1994). First, AKB coded the interviews with a set of “nodes” developed with SGN in NVivo 10.0, in light of the interview questions and STRIVE’s evaluation outcomes. Shortly thereafter, AKB and SGN met in Bagalkot to discuss with the CRIs and male RI what they found to be the most important findings from each interview, and then to identify main points for each of the interview guide sections across participants.

In a subsequent workshop, these main descriptive points were translated into Kannada and presented in PowerPoint by the CRIs to the community Research Committee and KHPT intervention team. This was the basis for a collaborative “meaning-making” exercise that adapted card-sorting and causal flow diagram methods used elsewhere in qualitative violence research (Ellsberg & Heise, 2005). Specifically, we formed two groups composed of one CRI and KHPT or UM researcher, with half of the Research Committee members in each. Each of the main points from our descriptive analysis was written on small cards of paper in Kannada, which were enumerated to correspond with a list of points in English. A copy of each was given to the two separate groups. The groups arranged the sheets on a poster board to show which points related to one another, with violence and condom use as the central, though not sole, outcomes of interest. The groups decided which points were most closely related to IPV and condom use outcomes, and which were less important and placed further to the periphery. There was much dialogue to prioritize, debate, and rearrange the cards before the final placements. A representative from each group shared why they arranged the diagram that way, leading to a general discussion between the groups. The cards in Kannada were turned over to be photographed and decoded to produce an English version (as shown in Figures 1 and 2, respectively). We used these diagrams to consider the prioritization of and relationships between the NVivo coding of interviews, which helped form themes to answer the research questions.

OPEN IN VIEWER

Figure 1.

Example of a collaborative “meaning-making” analysis diagram (Kannada language).

OPEN IN VIEWER

Figure 2.

Example of a collaborative “meaning-making” analysis diagram (English language).

Action and Application of Results

Our interpretations of the findings on IPV have fed into a cyclical process of research, implementation, and evaluation, with both inherent and instrumental benefits. The CRIs and Research Committee members shared that being involved in the study was beneficial. For example, one member stated that the analytical process encouraged her own personal reflection:

After joining the Research Committee, because our women express how there is violence with them, I felt it is like that for me also. Then I started understanding . . . I am personally experiencing this. Now I know what is right and what is wrong. Earlier if he beat me, scolded me also I didn’t say anything to him, I wouldn’t answer back. Now I feel, why does he talk like that?

Many female participants also said that they found the process helpful, as one CRI reported back: “At the end [of the interviews] we asked, ‘by talking to us did you find any problem?’ They said, ‘[my] mind became lighter’.” Similar reflections were shared by a respondent at the end of the interview: Interviewer:

Thank you very much, see you.

This approach continues to have instrumental benefits by informing programs and facilitating future CBPR research, as one Research Committee member expressed,

Because this is important research, people from other countries are also involved. Such studies are very few . . . What we have learnt here may be useful to others . . . Still, more research should take place about other topics. It should involve our community members . . . We can prepare them [researchers]; on who should be there, what questions should be there and all . . . We have taken training, we can ask questions. We know how to go stepwise [in research processes] . . . If we do research about a new topic then it will be more learning for us. We will come to know about other methods.

The Research Committee has continued to meet regularly with KHPT in the development of the tools for the second round of the evaluation study. The tools for the next round will delve deeper into complex and relevant issues that were raised in the baseline study, such as how differing degrees of intimacy and dependency in intimate relationships, and the associated gender role expectations, together shape levels of violence. This is based on the common finding that it was acceptable for male partners who were consistently providing financial support to use violence as a form of discipline, as one woman shared, Interviewer:

Do you mean that men beat the women when they make a mistake?

A male partner affirmed this by saying that he was not supporting her and therefore would not fight with her if she went outside without permission: Interviewer:

Don’t you ask her why she went to other places?

Another area that was identified for further exploration in the second round was the social meaning of being a Devadasi, and how this is reconciled with the social expectations for monogamous male–female relationships. The acceptance of violence was seen as a sign of love as in a marriage, but the aspiration to be like a marriage was heightened by the changing legal and social environment in which the Devadasi tradition has been criminalized and blamed for the spread of HIV/AIDS:

[T]hey [intimate partners] have to be like married couples if they want to continue their relationships . . . they [government] told us not to continue this observance of tying the righteous pearls to women [dedicating women as Devadasi sex workers] since dangerous infections are spreading. Awareness is being created among the families of our community to re-establish their lives. If there are any unmarried lovers, they are getting married. (Female sex worker, aged 26)

The knowledge gained on reasons for IPV and condom use from the study has also continued to inform KHPT and Chaitanya’s implementation of the Samvedana Plus program. Specifically, a major finding in the study was around the level of suspicion and infidelity in the FSW-intimate partner relationship as a trigger of violence. The analysis showed that the relationship was based on gender role expectations to be like a marriage, leading women to stop or at least conceal their involvement in sex work:

I forgot that [sex work] profession. We are living like married couples. He and I are like wife and husband and the whole village knows it . . . (Female sex worker, aged 33)

Furthermore, violence was justified when expectations for fidelity were contravened either through sex work or other forms, as one man stated,

She [female partner] should not have any contact with other men, because I have taken the responsibility of her life. So I will beat her if she commits a mistake. I will be like a husband with [lover’s name] as how you [interviewer] are with your wife. (Male intimate partner, aged 31)

The Samvedana Plus curricula for counseling female and male intimate partners have therefore been revised to be more sensitive to addressing triggers of violence like jealousy and trust in the context of gender role expectations to be like a marriage. The information on complex social dynamics in intimate relationships that the baseline study provided will help to strengthen the work of Chaitanya Mahila Sangha and KHPT on Samvedana Plus both methodologically and programmatically.