By Gillian Elam

What are the challenges to developing qualitative research with black and minority ethnic populations and how to meet them were some of the questions debated at the last African HIV Research Forum (AHRF) seminar day Research Methods session.

Participants in the seminar gave their views on what they thought the challenges were and I offered some answers, drawing on my experiences of using qualitative methods to research general and sexual health-related topics among Africans in the UK. I suggested that well designed and executed qualitative research conducted by trained interviewers can meet many of the challenges identified by participants. But there are also areas where qualitative research needs to be strengthened to ensure that methods are appropriate, relevant and inclusive. However, some of these methods can raise further challenges of their own and often the key to successful qualitative research with black and minority ethnic populations is the development of research strategies in collaboration with study communities.

Many of the challenges for research identified by participants related to those arising when researching communities for whom English is a second language or comprise a range of dialects. The beauty of qualitative research is that the interview provides an ideal opportunity to explore language and the ways in which people construct their accounts of the social world. The words that research participants use to describe a feeling or experience can provide great insight into understanding phenomena. A good interviewer does not impose their own jargon, but is able to negotiate with an interviewee the appropriate words to use in an interview. Nor would they let words or cultural references slip by without exploring what they meant to the respondent and why the respondent has applied them to their behaviour or circumstances. Qualitative interviews and text- based approaches to analysis provide opportunities for exploring perceptions and attitudes through the language used by respondents to tell their stories. But the ability for interviewers to do this well also depends on community collaboration to agree how research questions should be framed and to inform the design of tools used to collect data. Such collaboration is key in any study, including those led by community researchers, to ensure that population diversity is represented.

For language-based techniques to succeed, it is imperative to work with translators or conduct the interviews in the relevant community languages. Here the challenge can be cajoling funders to resource studies so that the range of languages used in a population can be included, and for researchers to feel confident enough in pursuing this. Often the scope of such studies ends up being focused on a smaller selection of linguistic population groups. This can be adequate, as long as any sample limitations are openly acknowledged, but there is a danger that more vulnerable groups are omitted because of the limited availability of researchers that can speak minority languages. Such groups may be those who are in greater need, for example the more vulnerable members of the population that have recently migrated, or are isolated in care roles at home that limit opportunities to learn new languages, or are dependent on others to mediate their everyday interactions.

One way of ensuring that interviews are conducted in appropriate languages and that vulnerable ethnic groups are included, is the involvement of ethnically matched interviewers. This is a widely used strategy that has been successfully employed in many studies. However, participants identified further challenges, for example a concern that sometimes confidentiality could be compromised if the interviewer is too close to home with the African study population. This can be addressed by: ensuring that interviewers do not work in their own neighbourhoods; including the names of interviewers in recruitment documents so that potential respondents are aware who is conducting the study; collecting data in anonymised forms; storing data in tamper proof holders until it reaches the research centre; or anonymising interview data. The latter can be time-consuming and requires not just the removal of names, but also any references that may lead to deductive disclosure, for example details of employment or voluntary work. An alternative is to explain to respondents what will happen to their interview data and who will see it within the research team, and then storing data with access limited to those named researchers.

For the interviewer, matched interviewing can be a challenge too. Matched interviewers often have to draw on a broader skill set than others because in addition to conducting interviews, they may be expected to recruit respondents and translate interviews. During the interview they will have to work hard to adapt their interviewing techniques to remind the respondent to explain any cultural references or practices that respondents may not elaborate on due to assumptions about shared cultural knowledge.

Whilst workshop participants agreed that community researchers were essential, one of the problems faced by researchers engaging with sexual health research among African communities is that the commissioning and funding of such studies tends to be sporadic, providing specialist community researchers with patchy employment opportunities. More does need to be done to invest in the research capacity of African community interviewers and researchers to enable them to develop sustainable research careers and develop their specialist skills in other areas. One approach to this would be to develop a Research Directory so that community researchers could advertise their research and linguistic skills to a broader research and commissioning community.

Moving on from language, other challenges raised included achieving viable samples from such a diverse population as the African UK population. General population studies have used focussed enumeration to improve the representation of black and ethnic minority populations. This involves door-to-door recruitment in electoral wards identified from census data with higher than average black and minority ethnic populations. However, this can be a resource intensive strategy. Often research is with special populations, for example patients, young people or clinic users. Here the challenges include addressing concerns about why Africans are being targeted for research and require dedicated teams of community recruiters who have been sufficiently trained and briefed to feel confident in addressing respondents’ concerns. Again, qualitative methods, in the form of purposive sampling, provide a means of rising to this challenge. Those features of diverse populations that are relevant to the research question are identified – another activity best carried out in conjunction with community collaborators - and these are used to guide the selection and recruitment of study participants. A combination of setting sample quotas and ongoing monitoring can ensure that sufficient diversity is achieved to answer the research question.

Participants commented on the difficulties of remaining objective during the analysis of qualitative data and ensuring the reported findings reflected the needs and experiences of the study population. Strategies for analysing qualitative data are well documented and a range of manual techniques and computer packages are available to manage the detail and volume of information generated by qualitative studies. These approaches share some basic tenets to ensure that data handing and interrogation are grounded in the verbatim interview data and that analytical questions are applied systematically throughout the entire dataset. Community review during this process can aid researchers in developing appropriate research questions and challenging any unfounded assumptions.

Finally, seminar participants agreed that it was essential that study communities saw the study outcomes and that promises to disseminate to the community were fulfilled. Putting adequate funding for dissemination in place from the beginning of a study and active participation of community collaborators can help ensure that study findings are accessible.

Throughout the discussion, the central role of community collaborators was clear, but in order for community representatives – who include volunteers and those from poorly resourced enterprises as well as established organisations - to collaborate fully and equally, then investment is also required to provide financial and technical support so that all sectors of the community are represented, not just those with the resources to do so.