Ethical considerations

This study has been approved by the University of Leicester’s Research Ethics Board (approval ID #35229). Prior to interview data collection all participants were given a detailed participant information sheet, outlining the research aims, their role, and how we would collect, store and use their data. The researchers read out the consent form at the beginning of the interview and recorded the verbal consent of the participants. Consent is not assumed to be continuous, therefore participants had the option to withdraw without giving a reason and their data would be deleted up until 31^st January 2023.

For the survey, the participant information sheet and consent form was embedded within the survey tool. Agreeing to complete the survey was taken as consent of participation. Withdrawal was possible, up until the participant submitted the survey.

All research participants were safeguarded from emotional distress and a document containing relevant support services made available to them if they required further support.

Data sources and tools

The core challenge to be addressed in this research was the fact that Adult Services Websites (ASWs) are an extensive and pervasive feature of the digital world across the globe that facilitate the advertising, negotiation and purchase of sexual services yet are also considered to be harbourers of sexual exploitation, modern slavery and human trafficking (MSHT) [3, 7, 18]. The policy and legal landscape under which regulating ASWs falls is significant, including the ongoing development of the UK Online Safety Bill [19]. To address gaps in the literature and practice knowledge this project had three core partners: National Crime Agency (NCA), National Police Chief’s Council (NPCC) and the leading NGO for the prevention of modern slavery in the UK, Unseen (who run the national helpline for victims).

Collaboratively with our partners this study first aimed to understand how ASWs facilitate exploitation and second how they can become part of the prevention mechanisms and reporters of crimes. The objectives centred on creating new knowledge and understanding to influence future law and policy reforms. Our objectives were fourfold:

1. To improve evidence on how Adult Services Website businesses can act as prevention stakeholders in MSHT by finding out what data they would need to keep to facilitate police investigations and good practice mechanisms around safeguarding;
2. Gather the views of ASWs operators and those who engage with buying sexual services regarding prospective laws to regulate ASWs in order to give insight into how regulatory reform should be designed to ensure compliance;
3. Identify the risks associated with policing around ASWs regarding victim identification and apprehending offenders to ensure any regulation does not displace MSHT or make spotting victims harder;
4. Establish a national network of stakeholders working with ASWs for long term action planning and consolidation of expertise to speak directly to law, policy, guidance and practice

Thus, our research questions centred around these objectives and informed the methods of study:

1. How can ASWs be used to disrupt offenders? What are the mechanisms adopted by ASWs for identifying victims and what do they do with the information?
2. How do operators of ASWs see their role in preventing abuses?
3. How do users of ASWs (providers and consumers) perceive their role in identifying crimes, barriers to reporting and prospective further criminalisation?
4. What are police expectations on the usefulness of tighter control of ASWs in victim interventions and apprehending offenders?

These research questions, alongside a multi-method approach offered a holistic view of the problems to be answered, and hopes to encourage collaborative working between ASWs, the police and policy makers tasked with preventing MSHT online.

Peer researchers and Action Learning Set

A growing body of research and associated policies advocates for meaningful involvement of stakeholder communities, particularly those historically marginalised, in the design, development and administration of the research that affects them [20–22]. This project relied on Unseen for ethical survivor involvement as a peer research project. We worked under Unseen’s Survivor Involvement protocol and strategy to inform the setup of the project and the ethical approach to involving adult survivors of exploitation. Ten peers from seven countries were involved through participation in the advisory group, instrument design, implementation, analysis and the dissemination phase. All ten peers engaged throughout the project, however a protocol was in place such that they were free to withdraw at any time, at which point a further round of peers would be hired if any withdrew from the programme.

A training and support needs analysis was completed at the beginning of the project to ascertain what skills people would bring, what they would like to develop for further personal and career development and if there were any additional requirements they needed to enable participation (e.g. interpreting services, childcare costs, travel costs and/or payment). Consequently, provided capacity building through an Action Learning Set (ASL) methodology, made up of five sessions over the lifetime of the project, which was given to Unseen as a tool for peer research methods in MSHT research. The training needs analysis and input from the peers guided these sessions, which included research skills (survey and interview), researcher reflexivity, data analysis, writing for different audiences, impact and dissemination including creating resources.

An Action Learning (AL) approach acknowledges the agency and expertise of researchers as active participants and co-researchers [23]. Academic knowledge is often assumed to be derived from academic researchers. Action learning and peer research methodologies acknowledge and encourage peers with lived experience to create knowledge through their experiences [24]. Action learning is inclusive and understands the value of non-academics as experts with knowledge of the research topic [23]. Thus, research does not necessarily strive to be objective, but inclusive of lived and diverse experiences. This not only recognises the agency of co-researchers but also critiques the ‘objective’ academic researcher who assumes they know right from wrong.

Practicing an inclusive AL method for the purposes of this project, meant considering and integrating the beliefs, experiences and skills of survivors of MSHT as co-researchers within the study. Action learning meant: asking fresh questions; learning from and with one another; working collaboratively on solving complex research problems; sharing experiences, ideas, feelings; and critically reflecting on what works and what does not, how and how not, and why or why not [24]. Bawden and Williams [25] argued that this approach challenges current world views and the status quo, a necessary prerequisite for social science research whose whole purpose is to enact change and challenge previous knowledge or practices.

Successful peer research methods require training where peers lack experience which can add complexities and extra time commitments to a project [26]. Without adequate training in research skills, the validity and reliability of research findings can be questioned and undermined, thus limiting its impact. Lushey and Munro [26] argue that participatory peer research methods must address ethical, practical and data quality issues from the outset by providing adequate resources and effective research management to deliver and upskill participating peers. Peer research methodologies that adopt this approach avoid being tokenistic, in which the researchers are listened to, supported in sharing their views and ideas, are involved in decision making processes, and have some power and responsibility [27]. It is also important, however, to acknowledge the ‘insider’ and lived experience knowledge peer researchers bring to a project that can offer rich insight into the topic [26].

Thus, each session with the peer researchers began with a start-up skills block wherein co-researchers got to know the background and aims of the study, the research skills to be implemented in that session and ask questions regarding what was to be expected [23]. This was facilitated by Unseen, who offered continual support, training and meetings to the peers. This therefore helped to foster a more relaxed environment for relationship building. Moreover, as per good practice guidelines [23] the academic research team provided participants with all materials needed before each session, with references to recommended reading, literature, and websites so everyone would be prepared to contribute and get the most out of each session. Also, a planned schedule for each workshop was made ahead of time, underpinned by a training needs analysis undertaken by the peer researchers, and matched to the various stages of the research.

At each session, an overarching theme was chosen as the focus that both met the desires and interests of the peer researchers, whilst also matching the stage of the research project. The first two sessions were on data analysis (both qualitative and quantitative) whilst the next sessions focused on reflexivity, research impact and dissemination. During each session, the group critically reflected on their knowledge, and the purpose of each session, ending with a mutual, group-built understanding of the session’s objectives. Important to the academic researchers was a cascade of learning and knowledge to the peers, such that they had the confidence and the skills to meaningfully engage and share their insights and reflections in the research process. One aim of peer research methods is to encourage peer researchers to become competent researchers in their own rights, thus delivering skills training based on a training needs analysis was helpful in supporting this goal.

Upon completion of the Action Learning Set sessions, the peer researchers were awarded with a certificate of recognition, endorsed by the University of Leicester. This was to recognise their contribution to the research process as meaningful, as well as offering them a tangible document to add to their CV for future employment. The peer researchers were also given the opportunity to share their learning and findings in some public dissemination settings as well as contributing to journal outputs. The community events had the benefit of being a more relaxed environment for dissemination as well as raising awareness in the community regarding research findings to lead to positive change.

Qualitative methodology → sample → procedures → analysis

We successfully conducted 51 interviews with the police (n = 30), practitioners (n = 13) and those working in the Adult Service Website industry (n = 6). The interview tools were jointly devised by the research team and peer researchers, with input from the partners who were used to assess the feasibility and focus of the research questions. Our engagement plans included utilising contacts through the NCA and police networks and snowball sampling as participants recommended others with the relevant subject matter experience. Relying heavily on the dynamics of organic and collaborate social networks it was important to design a strategic engagement plan prior to data collection to ensure maximum uptake. Our research partners proved instrumental in this process given their active involvement in this field.

In-depth interviews with ASW staff aimed to understand how ASWs viewed their role and procedures in identifying offenders of MSHT and their appetite for working alongside law enforcement in preventing sexual exploitation. The in-depth interviews with police and practitioners involved with ASWs engagement and enforcement aimed to understand their experiences of MSHT cases on ASWs and their expectations on the usefulness of tighter control of ASWs, including the feasibility and applicability of the Online Safety Bill. Together, these three interview groups offered a 360-degree insight into adult service website operations and the engagement of working parties in preventing MSHT. An omission of any one of these interview groups would overlook necessary and important information from groups that are and ought to be working together collaboratively to prevent exploitation online.

The interview questions were designed collaboratively with the peer research group, engaging both their expert knowledge of adult service websites, and specific areas of focus regarding police and platform responsibility in responding to MSHT in ways that centred the victims’ experiences free of judgement. An advantage of working with survivors in peer research is evident in achieving a holistic view of a research topic. Researchers, who whilst hold skills in emotive and sensitive research, lack the lived experiences offered by the peers, whose on the ground information about the realities of working online, experiencing exploitation and can attest to the needs of sex workers online can drive the research in a direction that respects sex workers’ rights to safety in ways that is meaningful for them. This research, as a peer research methodology, centred and prioritised sex workers’ experiences and desires, whilst also working collaboratively alongside law enforcement and policymakers to shape the laws to maximum effect.

The interviews were carried out and recorded via Microsoft Teams. Once the recording was completed the protocol, by default, meant the audio files were safely transferred to the designated Microsoft Teams research folder (a secure research file store used at the UoL), accessible only by the four interviewers, whilst the original recordings were deleted from the worker’s computer. This was to ensure maximum confidentiality and security. The audio files were then sent via a secure transfer to transcription service, Transcribe It, with a confidentiality agreement in place with the University of Leicester. This service provided verbatim Microsoft Word 2017 files which gave a transcription of the interviews. This was sufficient for discursive analysis of the interview data. The raw data was not shared across the team to protect the anonymity and confidentiality of research participants. The original recording will be destroyed five years after the project has ended, as is the University’s protocol.

All interview participants were informed of their rights and how their data would be collected, stored and used through a detailed participant information sheet shared prior to the interview. This has been approved by the University of Leicester’s rigorous ethics process. Consent was not assumed to be continuous, therefore participants had the option to withdraw without giving a reason and their data to be deleted up until 31^st January 2023.

The interview data was analysed using thematic analysis [28]. Thematic analysis was presented through thematic networks: ‘web-like illustrations (networks) that summarise the main themes constituting a piece of text’ [28: p.386], which enabled an emotionally sensitive, insightful and rich exploration of the data’s patterns. Thematic networks are an analysis tool which, similar to augmentation process, explores the connection between the explicit statements made by respondents and the implicit meanings, drawing on context [29]. They are a clear way of organising the data, through unearthing key themes using a representational tool that is accessible and easy to interpret [28].

The peer researchers assisted in identifying the key themes. Taking into consideration the inexperience of the peer researchers, training on thematic analysis was first delivered to offer a basic understanding of how to develop a codebook and identify key research themes from the findings. The peers’ lived experience proved useful in this regard, as they had an understanding of the types of themes likely to arise. The codebooks were then finalised by the academic research team to ensure academic robustness. It could be argued that this approach results in only partial participation in the analysis stage [26]. However, pragmatically, this was the correct decision, both due to the time limitations and commitments of the peer researchers on the project, and the experience of academics on data analysis tools. However, many of the themes adopted were both identified and named by the peers, rather than imposed by academic researchers during their analysis.

Quantitative methodology → sample → measures → analysis

Using JISC’s Online Survey (certified to ISO 27001 standard), we carried out a survey for consumers using ASWs, to understand how they perceive their role in identifying crimes, barriers to reporting and prospective further criminalisation in relation to MSHT on ASWs.

The survey was jointly devised by the research team and peer researchers, with input from the partners and was used to assess the feasibility and focus of the research questions. This both recognised the expertise and experience of the peers, whilst also empowering them as research experts. Moreover, in addition, the peer researchers were offered research training specifically on designing a survey, and consequently, were nurtured in their academic research skills.

We collected data from 150 survey participants to gather a breadth of experiences and opinions to draw informed conclusions. Our engagement plans included advertising on adult service website platforms, and utilising our research partner networks, many of whom (including Unseen, the NCA and NPCC) work closely alongside adult service websites to improve their reporting and support mechanisms regarding abuse, MSHT and other illegal or harmful activities. This research project recognises that those ASWs which engaged and advertised the research are those who are active in support and preventing MSHT on their sites. However, to reach those consumers of sex on less regulated ASWs we hoped to either engage through snowball sampling and word of mouth.

The survey collected basic socio-demographic data from participants, whilst preserving anonymity. This was to understand the diversity of consumers of sex on adult service websites. To assess participants awareness of sexual exploitation, a mix of closed and open-ended questions were asked. The questionnaire’s inclusion of some open-ended free text questions allowed participants the opportunity to share their experiences and opinions in an unbiased way, thus giving agency to respondents. The survey questions were also designed with the peer research group who drew on their experiences and knowledge of ASW workings to inform and direct the survey’s focus.

The survey data was held on the Online Survey confidential secure site until the survey collection concluded and then the data was moved on to the Microsoft Teams research folder. No identifying features were collected within the survey data, however if anyone disclosed any potentially identifiable information, all efforts were made to anonymise this data and the raw data was again not shared with the wider research team.

The data was analysed using Statistical Packaging for the Social Sciences (SPSS), using a range of descriptive and crosstab analysis tools, as well as thematic analysis for the open-ended descriptive questions. This allowed for an in-depth exploration of consumers’ of sex experiences on ASWs to understand how they perceived their role in identifying MSHT crimes, barriers to reporting and prospective further criminalisation in relation to MSHT on ASWs. Thus, we sought to understand both their experiences of MSHT online, the current role of ASWs in providing a safe environment, and their views regarding tighter regulation on the purchase and selling of consensual sex, as distinct from trafficking offences to be regulated. Consequently, this research sought to understand how the users of ASWs perceive their role in this field.

Through this multi-way approach to gathering data and subsequent analysis we mitigated against any risk of ASW operatives and professionals not co-operating given the on the ground insight and lived experiences of consumers of sexual services on a wide variety of ASWs. This multi-method approach provided a 360-degree perspective of ASW activity and enabled several strands of insight into the project, maximising our understanding of the research questions.

The peer researchers assisted in the basic quantitative data analysis only. This was due to the skill levels required to analyse data using SPSS. However, their role proved instrumental in designing and presenting the quantitative data, in particular for the dissemination of our findings to wider audiences, including ASWs, and the providers and consumers of sexual services. In this way, the findings of our research reach beyond academic or political interest, but impact communities with lived experience, who can benefit from the research, including developing an understanding of how the Online Safety Bill may impact on their experiences on ASWs. The peers lived experience was harnessed to present the data creatively, in a way that was meaningful for them. Throughout the process of action learning, they also engaged in academic data analysis and dissemination to provide them with the necessary skills to develop as academic researchers and to be equitably involved in all stages of the research project [30]. Together, we worked on complex issues creating knowledge that could both contribute to the field of academic research and community knowledge that may achieve sustainable solutions to safe working environments online. However, this must be complemented by carefully crafted policy developments, as highlighted and shaped by our data analysis.