Methodology

Establishing ethical processes, methodological vigour and robust methods in sex work research is vital in producing research that is effective in improving the lives and health outcomes of some of the most marginalised members of the community. Many sex workers have an antagonistic relationship with research and researchers (Wahab 2003). This relationship is complicated not only due to suspicion of authorities, but also because some research has provided ammunition to suppress the sex industry and victimize sex workers (Metzenrath 1998). There is a broadly held position in the sex industry across the world that ‘academic careers are made on our backs’ (Metzenrath 1998: no page) and that research has not been responsive to issues that are a priority to sex workers (Loff 2008). Sex workers themselves have repeatedly criticised researchers and called on them to involve sex workers in the research process, to generate and provide information useful to their lives and to engage with participants as subjects and agents (Pyett 1998; Agustín 2002). 

​

Collaboration is therefore a vital strategy for centering sex workers in generating the methods to be used in the study. Collaborative approaches borrow from participatory methodologies in that they emphasise the value of local knowledge and multiple knowledges and seek to include research subjects as participants with investment in research processes (Chambers 1994a, 1994b). Collaboration is founded on a number of premises, including the sharing of authority and the inclusion of differences and disagreements (Benson and Nagar 2006; Rouverol 2000). Therefore, having gatekeepers, identified by our partner organisations, to assist with access to sites was an added advantage not only for accessing sex workers in an appropriate way, but also for ensuring that the questions asked were relevant and useful to sex workers themselves. Moreover, centering sex worker led groups in the research design and consulting with them throughout the process meant that the important ‘local contextual and cultural factors that can impact the ethics of research… in many African communities’ (Koloi-Keaikitse et al. 2021) was adequately accounted for within the research project. A key part of this involved employing peer researchers who live and work as sex workers in the four countries to collect some of our data - in this way, the project was better able to respond flexibly to the cultural realities and context (Shaibu 2007) of sex work in southern Africa, as and when unforeseen issues inevitably arose from the research.  

​

In this research, we engaged in a process of ‘dialogic collaboration’ developed by PI Hardy in research with sex workers’ organisation in Argentina and used with organisations across Latin America over the last fifteen years, using a repeated cross-section longitudinal design to gather both quantitative and qualitative data (Hardy 2012; Hardy and Gillespie 2021). Dialogic collaboration is influenced by feminist standpoint theory, Marx’s social ontology (Ruddick et al. 2018) and participatory methods, drawing on Freire (1970) and developed further in Latin America (Winton 2007). Dialogic collaboration emphasises the importance of relationship building, responsibility and participation in research design and implementation in order to develop more mutual relations between researchers and participants across different geographic locations in the global ‘North’ and ‘South’. It is guided by three principles: the production of useful data which actively seeks to highlight and challenge processes of oppression; creating ways of working in solidarity with subjects and participants whose social positionality differs from that of the researcher; and answering the call by sex workers for methods which complement their lives and needs and develop research ‘with’, rather than ‘about’ them. Collaboration has also specifically been promoted as an appropriate approach for working with sex workers in sensitive and useful ways (Pyett 1998; Metzenrath 1998; Wolffers 2004) and inviting sex workers to engage in every element of the research from initial question formation to the analysis of data (Wahab 2003). 

​

In practical terms, this meant that all elements of the research, from design to implementation and dissemination will be developed through dialogue and collaboration with both Aidsfonds, our partner organisations, and sex workers themselves. In this research, we used both quantitative and qualitative data collection including a survey and semi-structured interviews with sex workers. Sex workers associated with the partner organisations were trained and paid as peer researchers to undertake the collection of the survey data, with the aim of defusing some of the power imbalances in the research process. Engaging peer researchers to collect sensitive data such as this was also helpful in that they were able to use the same language as their peers to understand the situations that respondents describe. Using peer researchers is now an established ‘best practice’ principle of sex work research, including by the WHO (GarciaMoreno et al. 2005).

​

By enabling direct dialogue between affected communities and researcher, we can respond to the specifics of the various contexts in southern Africa. Due to the scope, length and complexity of the study, the research design proposed was a multi-strategy, repeated cross-sectional, longitudinal design with data collection taking place in 2022 and 2025. Owing to the nature of this research and the often transitory nature of sex working populations, it is unlikely that the same people who partake in the research in 2022 will necessarily be accessible in 2025 As such, the sample of sex workers surveyed and interviewed in 2022 and 2025 will therefore differ.