ARVAC Bulletin 100

Getting the Green Light for Red Light Research

When conducting community research it is common to encounter groups that are marginalised, excluded from mainstream society and are therefore often difficult to locate or engage in traditional research. Researchers may find that conventional methods of obtaining a representative sample are defunct with certain disenfranchised groups rightly suspicious of ‘official’ questioning. For researchers, this presents a set of specific research dynamics that are important to consider in the research design and planning stage as well as the data collection or fieldwork phase.

Using my experience of researching the female sex industry in the UK over the past six years, this article will explore some of the common issues relating to accessing hard to reach populations for purposes of research. Many of the problems I faced in my research with sex workers are applicable to research with other difficult to access populations such as homeless people, drug users, travellers, migrant workers etc.

Gatekeepers
Marginalised groups are usually contactable through tried and tested routes of access such as social or health care services or the criminal justice system. A ‘captive audience’ can be found in drop-in centres, hostels, outreach projects, and other similar venues where individuals seek assistance and support. While this is the first port of call to make contact with hard to access groups, from my research I have noted that access is not a single transaction, but is ‘layered’.

This notion of ‘layered access’ can be understood through the example of how I gained access to female sex workers who worked in saunas and massage parlours. The barriers to reaching this group of women were related to the illegality of working in the sex industry in the UK, the stigma relating to selling sex, the general suspicion of anyone asking questions, and women’s understandable concerns regarding anonymity because many women are secretly involved in sex work. I gained access through a staged processes of presenting myself to different ‘gatekeeping audiences’. First, official access to the sexual health project that was to be my main ‘introducer’ was needed via the NHS. Second, I had to gain the trust and support of the health care professionals that could introduce me to the saunas. Third, once I was presented to the saunas as a trusted researcher who had no intention of exposing their identities, I had to gain trust from the sauna managers. Finally, after this access into the venues of work was secured, I had to approach individual sex workers and ask if they were willing to be interviewed.

It is in these complex relationships of access that the researcher has different relationships, obligations and responsibilities to each audience. This makes the initial task of gaining access a tricky process of negotiating and wearing different hats.

Sampling
The methods of achieving a group of informants who are not accessible through conventional means is often one that is queried for not reflecting scientific procedures. Using random or representative sampling strategies is often not possible because there is no database from which individuals can be selected. There are additional issues of no-fixed-abode, highly mobile groups and chaotic lifestyles to consider when selecting how to contact individuals. As a result, most samples are not representative, but that doesn’t mean individuals cannot be chosen based on certain criteria such as age, gender, ethnicity, geography, etc.

This method of quota sampling goes some way to finding a representative sample based on what is known of the wider population. For instance, I wanted to obtain a sample of women who were across the ages, had various lengths of involvement in sex work and experience of different sex markets. The other important access route is through the strategy of snowballing. This technique works as a chain referral system whereby individuals recommend other possible informants from their own networks. With snowballing, it is important to remember to check that recommendations do fit the criteria set out in the research design, and that the referral system is monitored.

Paying respondents
As in medical and health research, it is now customary in social science research to pay informants for their time and as a gesture of goodwill for taking part in the research. This is ‘normal’ and good practice that follows ethical codes of conduct with regards to responsibilities and obligations researchers have to informants. However, where the target group may be involved in drug use or exploitative relationships, the monetary exchange can pose problems. For instance, in my research where women were paid £20 to be interviewed, there was great competition between women to be involved in the research so they could receive the financial reward. This caused hostility and aggression between established street sex workers who objected to the new comers taking part in the research.

The concern was also raised that offering payment to be interviewed was indirectly ‘buying knowledge’. Researchers need to the check the balance between providing an incentive for excluded people to take an interest in research, the ethics of the payment and the processes for checking that individuals fit the sampling criteria.

Over-researched populations
Often, groups that are difficult to access become over researched, especially if the issue slides onto the policy agenda and the topic or group becomes momentarily fashionable as ‘subjects’ suitable for research. Prostitution, in particular women who work on the street, has been the focus of government attention for the past two decades, first in relation to HIV/AIDs and more recently because sex work on the street has been re-defined as a ‘problem’ to neighbourhoods. Flurries of research activity result in the same gatekeepers being asked to take part in research, leading to both individuals and topics replicated in research.

Some solutions for avoiding this are to make sure outreach projects are involved in the consultative stage of the research. This can include involving organisations and service users in decision making about the research questions, the process for access, what service-led involvement there will be and whether there is the possibility of training ‘indigenous interviewers’ so that there are elements of empowerment written into the methodology framework.

Effects on researchers
Researching hard to access groups is often a demanding enterprise, mainly because of the constant need to chase up referrals, maintain good relations with various gatekeepers at different levels, as well as deal with the realities of being in the field. When working with groups who are marginalised, excluded, or have been victims of violence and oppression, the content of their stories can be emotionally stressful. Planning for reflexive time and support as a researcher is an important aspect of gaining access to hard to reach populations and should be considered an integral part of the research process.

Accessing difficult to locate groups can be extremely rewarding because, when it works, there can be a real sense of ‘giving voice’ to marginalised groups. Community-based research can be pivotal in contributing to the policy agenda and service changes. Enabling people to tell their stories and give their opinions is an empowering process, yet equally it is complex, detailed and needs foresight and sensitive planning.

Teela Sanders is a senior lecturer in the sociology of crime and deviance in the School of Sociology & Social Policy, University of Leeds.

Sex Work A Risky Business (2005) is published by Willan Publishing at £18.99.

Paying for Pleasure: Men who Buy Sex (forthcoming) will be published by Willan Publishing in 2007.

Visit www.leeds.ac.uk/sociology/people/ts.htm.

Contact: t.l.m.sanders@leeds.ac.uk