3. Researching Digitalised Criminal Justice with Empirical Methods

Qualitative research in the social sciences is considered a rigorous means to collect data that can then inform findings, policies and practice. Bachman and Schutt (2011, p. 16) define qualitative research methods as those ‘designed to capture social life as participants experience it’ rather than in predetermined categories. Such methods are often interpretative, exploratory and inductive. Qualitative methods aim to make meaningful connections and sense of social phenomena, and are appropriate when research is motivated by seeking descriptions, explanations and evaluations (Denzin et al., 2023). Bachman and Schutt (2011) argue that, because the data frequently include spoken words and observations, qualitative methods have an air of authenticity, texture and richness that is lacking in quantitative methods. Furthermore, mixed qualitative methods, involving multiple approaches, data and views, provide triangulation. This means that ‘the weakness of one method will be compensated for by the strengths of another’ and build a full picture (Hutchinson, 2018, p. 160). That is, mixed methods can generate comprehensive perspectives that minimise the risks of subjectivity and optimise the goal of balanced and rigorous findings (Richards, 2009; Richards & Bartels, 2011).

The analysis of technological trends through the social sciences and humanities can challenge techno-solutionist narratives that often dominate research agendas (Pink, 2022). Qualitative research enables reflexive, socio-technological and relational perspectives to contest such hegemonic solutionist accounts and keep the focus on social issues, people, lived experience, processes and environments. Importantly, qualitative research is crucial in ensuring that technologies are ‘ethical, equitable, and responsible’ (Pink, 2022, p. 6). When new legal technologies are assumed to be a panacea for the many complex challenges of criminal justice, that is, ‘legal tech solutionism’, de Souza (2022, p. 374) suggests that assumption must be critiqued comprehensively prior to new technologies being introduced.

My study has sought to problematise and evaluate videoconferencing technologies with a research design that is typical of mixed qualitative methods in its collection of data through caselaw content analysis, interviews, survey and observations. I have previously made use of similar research methods in my doctoral study (McKay, 2018) and collaborative projects (Kashyap et al., 2018). In the current study, these different methods have assisted me in eliciting descriptive accounts that were later interpreted, coded and analysed to reveal patterns, meanings, deep understanding and new knowledge regarding videolinks in criminal justice (Bachman & Schutt, 2011; Maruna, 2010). Collectively, the mixed qualitative methods have enabled a comprehensive understanding of the real-world workings of digitalised criminal procedure and vulnerable individuals from the perspectives of criminal justice professionals.

Since 2020’s notable increase in videoconferencing, there have been several qualitative studies regarding remote criminal justice that have sought the opinions and experiences of criminal justice actors, including judicial officers, legal professionals and a range of staff from justice agencies (e.g. see Alkon, 2022; Benninger et al., 2021; Fair Trials, 2020; Law Society of NSW, 2022; The Bar Council, 2024; Turner, 2021). Such studies have mostly made use of online surveys and the results have been seen as critically informing evaluations of remote procedure, policy development and advocacy for the profession and practice. Other recent studies have used courtroom observations and interviews with relevant criminal justice actors (e.g. Hutchinson, 2021; Nir & Musial, 2022; Ross et al., 2025; Wells Robey, 2024), courtroom ethnography (e.g. Flower & Klosterkamp, 2023), experimental mock jury studies (e.g. Mulcahy et al., 2020a, 2020b; Rossner & Tait, 2023), caselaw content analysis (e.g. Turner, 2024) and surveys of public attitudes to remote courts (e.g. Lemire-Garlic & Dunbar, 2024). My research design harnesses the strengths of select qualitative methods.

Because this research has involved humans, it has required close ethical attention to how participants were recruited for interviews and surveys, and how data have been collected, analysed and stored. All processes were conducted in accordance with the University of Sydney Human Research Ethics Committee Protocol (2022/142). Interview and survey participants were provided upfront with a Participant Information Statement that outlined the study and how the confidentiality of participants and privacy of their data would be protected in the dissemination of research results. Participants were also requested to provide Informed Consent if they wished to volunteer for the study. Following the Australian Research Council (ARC) research data management requirements, direct and known indirect identifiers were removed to break any links to participants’ identities. In adhering to de-identification protocols, interview participants and their collected data were identifiable only by an allocated code depending on their role in the criminal justice system, as per Table 3.1. Survey participants were automatically de-identified by the online Qualtrics platform.

My original ethics was incrementally amended to update the project team including the research assistant, Sydney Informatics Hub and an Honours student. In addition, before commencing my fieldwork travel in 2022, I became aware that some potential interviewees may be impacted by COVID-19 isolation, so I modified my ethics application to include online interviews. This was fortuitous as I caught COVID-19 during fieldwork in Alice Springs and had to complete my scheduled Darwin interviews online while I was in isolation. I also became aware that criminal justice-affiliated professionals, including police prosecutors, youth justice workers, witness intermediaries, witness assistance support officers, victim/survivor and justice advocates, would be interested in my study and offer invaluable insights, so my ethics approval was extended accordingly.

The richest data were generated by the national interviews. These interviews were conducted in-person and online throughout 2022–2023 across Australian cities, regions and remote locations with participants from Adelaide, Alice Springs, Brisbane, Broken Hill, Cairns, Canberra, Darwin, Dubbo, Griffith, Hobart, Kununurra, Launceston, Melbourne, Newcastle, Perth and Sydney (Fig. 3.1).

Bild vergrößern

Interview participants were recruited through direct contact with all states’ and territories’ judicial and legal professional bodies, supplemented by some word-of-mouth referrals from legal professionals. This represented a purposive sampling method in that each participant was selected as being an expert, knowledgeable about videolinks in criminal justice, and willing to talk (Bachman & Schutt, 2011). I had originally planned to recruit 10 criminal justice professionals from each jurisdiction, providing a sample of 80 participants and anticipated saturation. With 85 high-quality interviews, I exceeded my goal and, moreover, reached completeness and saturation, that is, no new issues were arising by the end of the interview processes, and there was depth and robustness in the responses (Richards, 2009).

There were challenges in recruiting participants. Some professional bodies and justice agencies were more responsive than others and, for the interviews, I elicited the greatest response from various legal aid services and affiliated services, particularly, North Australian Aboriginal Justice Agency (NAAJA) and Aboriginal Legal Service (ALS), resulting in an over-representation of defence lawyers compared with prosecutors. However, as the next section details, the online survey was overwhelmingly completed by prosecutors and witness assistance service professionals typically affiliated with prosecution services. Accessing judicial officers for interviews was especially difficult and, disappointingly, I could not recruit any in South Australia, Tasmania or Western Australia. Access to elite, prominent or influential professionals is well recognised as challenging for qualitative researchers (Collett, 2024; Richards, 2011).

The interviews were conducted with pre-prepared prompt questions for semi-structured or ‘guided’ conversations (Richards, 2011, p. 70), with questions designed to elicit open, free-flowing, in-depth narrative responses. Free-flowing responses definitely resulted, and a great benefit of interviewing criminal justice professionals is that they are incredibly articulate and passionate about their work. Prior to the interview, some participants requested a copy of the ethics-approved prompt questions, but most interviews were conducted conversationally and flexibly so that they sometimes deviated or, serendipitously, revealed unexpected issues and chance insights. The interviews provided participants with the opportunity to reveal their authentic knowledge of digitalised criminal procedure and their interactions with vulnerable individuals. They could express their diverse experiences and views of the benefits, disadvantages and reform ideas.

Interview participants were asked questions along broad themes concerning their demographics, AVL experience, understandings of vulnerability, evaluations of AVL, and future visions (see project website). With permission, interviews were audio-recorded and later professionally transcribed by an external transcription service. Participants could nominate if they wished to review their own transcript. The interviews were typically 30–60 minutes in duration, and I was always appreciative of the fact that the participants were being exceedingly generous with their time, given their significant professional and court commitments. Most interviews were conducted face-to-face although, due to COVID-19 issues continuing 2022–2023, some interviews were conducted using Microsoft Teams or Zoom. Online interviews tended to be a little shorter and more transactional than the conversational face-to-face interviews. The 85 interview participants included a broad range of ages, from decades of experience in the criminal justice sector, through to several lawyers who had just been admitted to the profession and for whom the remote mode was normalised.

While the national interview process had its challenges with recruitment and scheduling, it was by far the most rewarding part of the project. When I originally lodged my research application with the ARC, I had planned to conduct many interviews overseas. However, when funding was awarded at the end of 2020, pandemic travel restrictions were still in place, requiring the complete re-design of my fieldwork into a national study. Despite causing delays in commencing the study, this was, in many respects, a gift, enabling me to travel to parts of Australia that I’d never visited, and to experience the digitalisation of criminal justice across urban, regional and remote areas. I had the opportunity to converse with extremely dedicated and generous professionals and visit their various work environments. I chose Alice Springs in Central Australia (NT) as my first destination as I’d never been there before and guessed that it might present unique criminal justice issues.

A limitation of my research design is that vulnerable individuals were not part of project for three reasons. Firstly, I have previously undertaken empirical research with people-in-prison regarding their experiences of using videolinks to access courts and representation from custody (McKay, 2018). In this current study, I wanted to focus on the courtroom endpoint and seek the perceptions of the judiciary, lawyers and criminal justice professionals regarding a broader range of vulnerable persons being vulnerable witnesses, complainants, victims and defendants. The overarching aim has been to develop conceptualisations of digital vulnerability by gathering criminal justice professionals’ everyday experiences of working with vulnerable individuals, that is, I wanted to understand perspectives from the courts, prosecution and defence, so the targeted professionals were the best placed informants and experts. Secondly, a tightly focused study ensures its viability and delivery in a timely manner. Given that I conducted all interviews and analysed most data myself, it has been important that these processes were feasible for one main researcher. Thirdly, possibly insurmountable ethical, identification and recruitment issues would arise in relation to interviewing or surveying vulnerable individuals.

To enable the participation of a broader range of criminal justice professionals who were not available for face-to-face interviews or not in locations I visited, I conducted a survey through the web-based Qualtrics Research Suite, suitable for qualitative social science research. The survey link was open for 5 months: 4 May 2023–3 October 2023 and disseminated through state and territory legal professional bodies as well as through the Sydney Law School’s e-news, Sydney Institute of Criminology’s CrimNet e-newsletter, and my own professional social media on LinkedIn and X. This online survey provided an efficient and systematic way to capture comprehensive and direct insights from target professionals.

Initially, it seemed that there were 144 online survey responses, but closer inspection by the Sydney Informatics Hub team, who assisted me with data analysis, revealed that only 90 survey responses were valid and complete. Another issue was that due to the survey question design, respondents’ specific roles in the criminal justice system were not always clear, for instance, they could just respond with ‘solicitor’ or ‘barrister’ without stipulating ‘defence’ or ‘prosecution’. Nevertheless, the majority were prosecutors, witness assistance service professionals or witness intermediaries and, clearly, various prosecution services had successfully disseminated the survey amongst their staff. Of the 90 responses, only 6 explicitly identified as defence lawyers. The increased engagement by prosecutors addressed the shortfall in their interviews.

The survey questions were very similar to the interview questions and most were designed to yield open-text responses. The benefit of this method was the ease of coding survey responses to specific issues. However, while most survey respondents engaged with the open field questions that required a free-form textual response, at least 46% of responses were limited to the multiple-choice style questions only. On reflection, the open-ended questions required too much effort and should have been used more sparingly and, overall, the survey was too long. Nevertheless, the majority of respondents did provide some textual responses with many being high quality, insightful and detailed. On the downside, and in comparison with the dynamic interviews, the survey responses were constrained by the questions, so that no new or unexpected themes arose.

Throughout the 2022–2023 fieldwork stage, I travelled across Australia to not only interview people but also observe videolink technologies in diverse courtrooms (Table 3.2). Observation, as a research technique, immerses the researcher within the social site under examination to generate a direct, embodied experience. This can lead to a more profound, reflexive and nuanced understanding of the site being investigated—the digitalised courtroom—and a greater appreciation of the types of technologies being used (McKay, 2018).

These destinations provided an overview of all levels of criminal justice operating in capital cities, regional and remote locations. Collectively, they presented a snapshot of vulnerable populations, their use of audiovisual link technologies in many contexts and how digital technologies assist or compound vulnerabilities. Along the way, I also visited many judicial chambers, barristers’ chambers and legal offices.

Identifying courts that had videolink appearances was often challenging. Some online daily court lists were clear, but I would usually arrive at a court and simply peer through the courtroom doors to see if a videolink was occurring. Quite often court staff were bemused by my request to watch videolinks, for instance, in the ACT Supreme Court, one court staff member dismissively told me words to the effect that ‘there’s not much to see; it’s just like watching TV; it’s great’.

Bild vergrößern

Where possible, I made notes while in court with the permission of court officials. By way of example, on my first fieldwork day in Alice Springs, I spent the morning sitting in the public gallery of the Local Court where the vast majority of defendants or offenders were Aboriginal and Torres Strait Islander peoples (Fig. 3.2). I observed remote men and women being brought into the prison AVL rooms and developed a sense of those remote spaces and the sonic ruptures (McKay, 2020b):

Being in the public gallery of the courts allowed me to observe the interactions between the judge and lawyers in court and the remote defendants:

The observations allowed me to compare in-court versus remote interactions. For instance, in the Alice Springs Local Court, a young woman wearing an oversized orange sweater appeared by AVL and did not respond at all to the proceedings. In another matter, the judge was delivering a sentencing judgement to a young Aboriginal man by AVL:

The judge was endeavouring to engage with the young man and imploring him to desist from future criminal behaviour. During this, the young man was playing with his orange top, holding it in his mouth, putting his arms inside the top, and he had very little response towards the judge.

By comparison, an Aboriginal man was brought up from the cells into the plexiglass enclosed courtroom dock for an unopposed bail application, and I observed a much more engaged interaction between the judge and accused.

To understand the problems being presented in courts and how digital technology may solve or exacerbate issues of vulnerability, this stage involved reviewing Australian legislation, and mapping judicial reasoning and observations in criminal cases that concern videolink technologies and vulnerable witnesses, victims or defendants. Throughout this project, I have been collating criminal caselaw relevant to videolinks by setting up alerts on the online BarNet JADE open-access legal platform that provides Australian and international legislation and caselaw. It is designed and much used by Australian barristers. Relevant cases have been identified using keywords such as audiovisual link, AVL or videolink, the preferred terms in Australia. For focusing on specific issues, I employed Natural Language and Boolean Operators to finesse search results. I have collated hundreds of cases that have been manually coded to a multitude of themes including access to justice, advocacy, behaviour, demeanour, fairness, intermediaries, legal conferencing, open justice, pre-recorded evidence, presence, courtroom protocols, psychological assessments, self-represented parties, technical problems, trial procedure, victims, vulnerable children and young people, vulnerable defendants and vulnerable witnesses (see project website).

As we will see in Chapter ‘What Is Vulnerability in Criminal Justice?’, caselaw was analysed to understand how the courts currently respond to vulnerability when AVL is in use. This stage particularly involved research assistant, Kristin Macintosh, and we ended up co-authoring publications (McKay & Macintosh, 2023a, 2023b, 2024a, 2024b). I have used content analysis of caselaw in other publications (McKay, 2020a, 2020b, 2022a, 2022c) and, of course, many other scholars have adopted similar approaches (e.g. Turner, 2024). The content analysis of caselaw is a recognised, systematic and ‘uniquely legal empirical methodology’ for determining patterns, meanings and insights on a particular subject (Hall & Wright, 2008, p. 64; Hutchinson, 2018). As I argue elsewhere, there is great utility in the content analysis of caselaw due to it being a publicly available resource, and the cases reveal details of participants’ vulnerabilities and any special measures or responses from the court (McKay, 2022a).

I began the iterative process of thematically coding my empirical data by reading legislation, caselaw, interview transcripts and aggregated survey responses to identify emerging and recurrent themes. On the whole, these themes aligned with the focus of interview/survey questions and participants’ conceptualisations of vulnerability, as well as key benefits and challenges. While it was particularly easy to manually code the survey responses given their structured format, more surprising themes emerged from the interview transcripts. For instance, in McKay and Macintosh (2024a, p. 15) we wrote how the issue of the expression of emotion and empathy during remote court matters arose ‘unexpectedly and organically’ out of the interviews without a direct interview question, ‘demonstrating a strength of qualitative research methods in potentially revealing unforeseen matters’. Following my initial manual coding review, I then liaised with the Sydney Informatics Hub, a Core Research Facility of the University of Sydney, regarding optimal ways to analyse my empirical dataset. They developed a secure and private semantic AI tool that I used to systematically and speedily search the interview transcripts and survey responses for a range of themes. The results of these thematic searches using the semantic tool were downloaded and have assisted considerably in identifying emergent themes (see project website), coding and structuring the empirical data in Chapters ‘Remote Vulnerable Witnesses, Complainants and Victims’, ‘Remote Vulnerable Defendants’ and ‘Future Visions of Digitalised Criminal Justice’. The broad themes include understandings of vulnerability in practice; the interplay between remote access technologies and vulnerabilities; the benefits and challenges of remote participation; remote interpreting; access and barriers to justice and forward-thinking visions for the increasingly digitalised criminal justice system (Fig. 3.3).

Bild vergrößern