Reflexivity,Power and Positionality in the Field: Methodological Reflections of Researching Uber Drivers in Johannesburg

Introduction: From Passenger to Researcher in the Study of Uber Drivers in Johannesburg

My journey into researching Uber drivers in Johannesburg was not solely sparked by academic curiosity but emerged from my everyday life – as a passenger, a curious observer, and later, a researcher committed to understanding the nuances of digital platform-based work. Conversations with Uber drivers during routine trips revealed glimpses of hardship, resilience, and contradiction, as Uber drivers would mention long hours, at times dealing with difficult clients, and fears about safety, all while navigating the promise of flexible work. These interactions remained with me and laid the groundwork for an academic inquiry, focusing on how Uber drivers experience the working conditions of Uber driving, and how they experience and manage work-life boundaries in a digital labour economy (Bayane, 2024).

Reflexivity is an essential part of conducting research and immensely contributes to the knowledge-production process. Patnaik (2013) suggests that researchers should practice an active reflexive process of constantly reflecting on their positionality and beliefs to assess and minimise their impact on a study. Reflecting on positionality is crucial in qualitative research as it enables researchers to recognise that they do not exist in isolation but are part of the social world they are researching (Holmes, 2020). Researchers must therefore constantly introspect their role of subjectivity throughout the research process to realise its impact, particularly on the representation of participants’ views and experiences of a studied phenomenon (Dowling, 2006; Parahoo, 2006; Pillow, 2003).

As I transitioned from a passenger to a researcher, I realised that my positionality significantly shaped how I was received in the field. Entering the study as a male academic and car owner inevitably influenced interactions with Uber drivers. Despite approaching the research with the desire to understand and learn from the lived realities of Uber drivers, my social position involuntarily created barriers. These moments of discomfort were instructive, as they forced me to reflect not just on my methodological tools, but on who I was in relation to the people I was trying to study. Reflexivity is thus a central tenet in qualitative research and ensures that scholars remain critically aware of their role in shaping research outcomes (Holmes, 2020; Patnaik, 2013).

This paper reflects on the complexities of conducting fieldwork in the gig economy and digital platform work context, where class, gender, language, and researcher identity intersect and shape the research process. Drawing from but not reporting the empirical findings of my PhD research, I reflect on my positionality during this study, examining how factors such as my academic status, gender, power, and prior experiences influenced data collection and interpretation. I critically reflect on how reflexivity shaped my data collection, the strategies I used to build trust, and the ethical tensions I navigated throughout. This reflective engagement is not only about transparency in research but also a methodological stance that values critical self-awareness and reflexivity in knowledge production.

A Glimpse of my PhD Study on Male and Female Uber Drivers

The primary aim of my PhD study was to explore the working experiences of male and female Uber drivers in Johannesburg, using precariat theory to understand the extent and nature of their precarious working conditions (Bayane, 2024). The study also examined how these drivers navigate work-life boundaries, drawing on spillover and border theories to analyse the intersection of personal and professional domains. The rise of Uber and other digital labour platforms has generated a growing body of research, with some studies commending Uber for its convenience and income-generating potential (Anwar & Graham, 2022; Geitung, 2017; Henama & Sifolo, 2017; Kute, 2017; Webster, 2020). However, others highlight the risks associated with Uber driving, such as ambiguous employment status, lack of benefits, and exposure to violence and insecurity (Anwar & Graham, 2020; Chinguno, 2019; De Stefano, 2016; Fairwork, 2020, 2021; Heeks, 2017; Van Belle et al., 2023). While these debates are well-documented globally, there is limited sociological research on the working conditions and work-life dynamics of Uber drivers in Johannesburg, especially from a gendered perspective. Hence, the study addressed this gap by using the precariat theory to examine male and female Uber drivers’ working conditions and paired with spillover and border theories to explore work-life balance in Uber driving.

The precariat, spillover, and border theories informed and guided the framing of the research objectives, questions, and analysis of my PhD study (Bayane, 2024), which underpins this reflexive paper. Although there are contestations regarding the precariat as an emerging class of workers, scholars agree that it consists of individuals working in unstable, insecure, and precarious conditions (Fobosi, 2021; Foti, 2017; Johnson, 2014; Standing, 2011; Wright, 2016). Standing (2011) identifies forms of labour-related security, i.e., labour market, employment, job, work, skill reproduction, income, and representation security, whose absence defines precarity and the precariat. These constructs guided the research objective and questions, focusing on Uber drivers’ working conditions and shaping the design of interview questions on contractual arrangements, income stability, algorithm control and management, protection, and representation, as well as analysis of data collected from male and female Uber drivers.

Meanwhile, spillover theory examines how experiences, attitudes, and behaviours transfer between work and personal life, distinguishes between positive and negative spillover, and recognises bidirectional flows (Chen et al., 2009; Staines, 1980; Zedeck, 1987). This theoretical lens informed the second research objective and guided the inclusion of interview questions exploring how long hours, safety concerns, and the chasing of earnings targets influenced participants’ family life and social relationships. Border theory, which focuses on the permeability, flexibility, and blending of work and non-work boundaries (Clark, 2000; Emslie & Hunt, 2008), shaped the objective and guided questions on how Uber drivers manage or separate work and personal roles. Spillover and border theories also informed the analysis of boundary management, role conflict, and gendered differences in balancing paid and unpaid work in digital platform work (Bayane, 2024). Therefore, these theoretical constructs provided an integrated framework that structured the qualitative design from data collection through to analysis.

A constructivist paradigm guided the study, acknowledging that reality is subjectively constructed through lived experiences (Adom et al., 2016; Vanderstoep & Johnston, 2009). This paradigm allowed for close engagement with participants to understand how they interpret their work and manage its impact on personal life. Honebein (1996) and Creswell (2009) suggest that accessing such meaning requires trust and rapport, which were critical during fieldwork. To achieve this, I also adopted a qualitative research design (Babbie, 2021), deemed appropriate for examining the complex social dynamics and subjective experiences of Uber drivers. Qualitative methods enabled me to generate rich and detailed narratives, and also uncover underlying meanings (Bless et al., 2013; Sarantakos, 2005).

The research was conducted in Johannesburg, a city marked by economic disparities and high levels of unemployment, especially among migrants (Ahmad, 2010; World Population Review, 2021). Since Uber’s introduction in 2013, many residents have turned to platform-based work to earn a living (Henama & Sifolo, 2017; Webster, 2020). Uber has played a crucial role in offering income-generating opportunities and supporting livelihoods, especially since launching its operations in Johannesburg before expanding to other cities (Fairwork, 2020; Henama & Sifolo, 2017; Kute, 2017). While existing research has examined the emergence and popularity of Uber in Johannesburg (e.g., Berndt et al., 2021; Giddy, 2020; Kute, 2017; Wilmans & Rashied, 2021), few studies have explored the specific working conditions and work-life dynamics of Uber drivers in the city. This study was set out to address this gap by offering a sociological investigation into these dimensions of digital platform work (Bayane, 2024). Johannesburg presented itself as a fitting research setting due to its widespread adoption of gig economy services and the visible integration of platforms like Uber into daily urban life.

Participant recruitment occurred at malls and shopping centres across Johannesburg, including Mall of Africa (Midrand), Fourways Mall (Sandton), Clearwater Mall (Roodepoort), and Tembisa Mall. These places attract a diverse customer base from within and beyond Johannesburg and provided practical points of access to both male and female Uber drivers. Using non-probability sampling, I employed both convenience and snowball sampling to recruit participants (Babbie, 2021; Etikan et al., 2015). Convenience sampling allowed for practical access, while snowball sampling was particularly effective in reaching female drivers, who are less visible in the field.

In total, 30 semi-structured interviews were conducted with Uber drivers operating in areas such as Midrand, Tembisa, Roodepoort, and Fourways. It is important to note that some drivers declined to participate, citing potential income loss from not accepting ride requests during the interview. Snowball sampling was particularly effective in reaching female drivers, who were less visible in the field. After finding one female Uber driver in Carlswald, Midrand, I asked her to refer me to others. She also directed me to the Mall of Africa, a known hotspot for female Uber drivers. Referrals from both male and female participants significantly streamlined the recruitment process and ensured that all participants met the study’s inclusion criteria.

To gather insights into the lived experiences of male and female Uber drivers in Johannesburg, semi-structured interviews were conducted with male and female Uber drivers. Semi-structured interviews were selected for their flexibility and suitability for exploratory research, as they facilitate open-ended dialogue while still allowing for comparability across interviews (Bless et al., 2013). An interview guide structured into four thematic sections, demographics, Uber driving experiences, working conditions, and work-life balance, served as the foundation for these conversations, although it was adapted fluidly depending on the participant’s responses. Prior to participation, individuals were provided with an information letter outlining the purpose of the study and asked to sign a consent form, in line with ethical standards. Conversations were conducted in a mix of vernacular languages such as isiZulu, Xitsonga, and Sesotho, alongside English, and were later verbatim transcribed and translated for analysis. The interviews, which averaged around 55 minutes with each Uber driver, were occasionally interrupted by Uber trip notifications, yet drivers remained engaged. The semi-structured interview process not only enabled the collection of rich, nuanced data but also highlighted the importance of ethical sensitivity, adaptability, and reflexivity in qualitative fieldwork (Bless et al., 2013).

To analyse the experiences of male and female Uber drivers in Johannesburg, I used thematic content analysis, a method well-suited for identifying patterns and interpreting meaning within qualitative data (Braun et al., 2016). This method allowed me to organise participants’ narratives around key themes linked to the study’s objectives. The process began with verbatim transcription of interviews, many conducted in vernacular languages and translated into English, an important step that required careful attention to preserve meaning (Bailey, 2008). I used my field notes and interview guide to clarify unclear audio and remain grounded in context. Coding involved reading and re-reading transcripts to identify recurring ideas and contrasts, using open coding and colour-coding techniques to develop initial themes such as employment status (Babbie, 2021). Finally, I interpreted the themes in relation to the study’s core questions, ensuring that the analysis reflected both individual voices and broader patterns across the dataset. This structured yet flexible approach enabled a grounded understanding of Uber drivers’ working conditions and their strategies for managing work-life balance.

Ethics in social science research refer to the standards and procedures that guide researchers to conduct studies responsibly and protect participants from harm (Babbie, 2021; Bless et al., 2013). For this study, ethical clearance was obtained from the University of Johannesburg’s Faculty of Humanities Research Ethics Committee (Ethical Clearance Reference Number: REC-01-034-2021). To uphold ethical principles, I followed Bless et al. (2013) guidelines on non-maleficence, autonomy, confidentiality, and keeping promises. Non-maleficence involved informing participants about the study’s purpose and any potential risks, particularly since sharing personal experiences as Uber drivers could cause vulnerability. Participants were given full details regarding the purpose of the study before voluntarily consenting to participate. Respecting autonomy, I provided an information letter and consent form, allowing participants to make an informed decision about their involvement (Kumar, 2011). I honoured all commitments made during interviews, including maintaining confidentiality by assigning pseudonyms and securely storing data on a password-protected cloud accessible only to me. Finally, participants were clearly informed that the data would be used for this PhD research.

Negotiating Access and Power Through Class and Institutional Identity

Researchers’ positionality, shaped by social identity, institutional affiliation, and prior experiences, inevitably influences access, rapport, and knowledge production in qualitative research (Foote & Bartell, 2011; Holmes, 2020). At the outset, my own assumptions about Uber drivers were informed by countless informal conversations during rides, in which drivers recounted hardships and dissatisfaction with Uber driving. While these exchanges were insightful, they generated subtle bias – as I entered the field with preconceived notions of digital platform labour, shaped more by passenger sympathies than by critical distance. Confronting this bias reflexively required ongoing awareness of how it influenced my questioning, listening, and interpretations (Berger, 2015). Although my understanding of Uber drivers’ working conditions and work-life balance was informed by both personal experience and academic literature, I approached the study with a commitment to learn from Uber drivers’ own experiences and perspectives.

My dual role as researcher and academic complicated early fieldwork encounters. I drove to various locations to approach drivers, beginning with an introduction such as, “Good day, I am doing research for my PhD and would like to interview Uber drivers to hear about their experiences” (Researcher, September 2022). While this framing reflected my intent, it elicited suspicion. By identifying myself as a PhD student, I unintentionally reinforced power hierarchies that Uber drivers regularly face in their interactions with institutions or individuals. Some drivers redirected me to others or declined to engage altogether. These responses highlighted how my language, posture, and even attire affected rapport-building. Kvale and Brinkmann (2014) emphasise the importance of researchers’ embodied presence, particularly in precarious or unequal settings, where social markers are carefully read and interpreted.

Moreover, the use of a personal vehicle and my disclosure that I was an academic further marked me as belonging to a different class position and not someone Uber drivers can relate to. On several occasions, Uber drivers asked where I worked or observed my car before responding. Upon learning that I was an academic, some expressed disinterest or redirected me to other drivers, implying a perceived lack of shared experience. These responses made it evident that my institutional and class positionality created social distance between myself and Uber drivers. This aligns with findings from Anwar and Graham (2022), which illustrate how platform workers in African cities typically perceive researchers and formal institutions as occupying privileged and extractive positions, contributing to the mistrust. Similarly, Wood et al. (2019) argue that power asymmetries are intensified in digital platform labour relations, where gig workers are aware of how information can be used for or against them.

Recognising the limitations of my approach, I reflected on how to foster a more collaborative and less hierarchical field presence. I therefore, revised my introduction to, “Good day, may I please work with you today?” (Researcher, September 2022). This subtle but meaningful change to a participatory framing opened up space for engagement. It signalled humility and a willingness to learn, which made many Uber drivers more receptive as I continued with the fieldwork. Hammersley and Atkinson (2019) highlight the importance of relational ethics and cultural humility in unfamiliar field contexts, where researchers must earn rather than assume access.

In addition, I began parking my car at a distance and walking to Uber drivers' parking zones. This was not an attempt to conceal my identity, but a deliberate effort to minimise class-based signifiers (Bourdieu, 1990) and present myself as more approachable. I also shifted toward using vernacular languages when possible, such as isiZulu, Xitsonga and Sesotho, which opened new pathways to rapport. Conversations became more relaxed, expressive, and even humorous – reminding me that language is not a neutral tool, but a social and emotional medium tied to identity and power (Manderson et al., 2006; Temple & Young, 2004).

These adaptations emphasised a core lesson: positionality is not fixed but fluid – continuously shaped and negotiated through everyday interactions (Sultana, 2007). Being reflexive required more than acknowledging privilege; it demanded attention to how that privilege manifested in the field, shaped participants’ responses, and influenced what knowledge could be co-produced. It became clear that reflexivity was not simply a methodological note but a central component of ethical and credible research practice in contexts marked by inequality and precarity.

Gendered Power Relations and Embodied Ethics in the Field

The data collection process is inevitably shaped by intersecting structural, social, and contextual factors such as gender, age, race, and perceived social status (Manderson et al., 2006). Researchers ought to be reflexive, which requires a critical awareness of how these factors influence not only access to participants but also the nature and depth of the data generated during interviews (Bourke, 2014). During fieldwork, I realised that my identity as a male both opened and constrained research relationships in gendered ways. Interviews with male Uber drivers were relatively easy to initiate and maintain. Several male participants expressed an assumed sense of companionship based on shared gender, using phrases like, “As a man, I’m sure you understand” (Uber driver, October 2022). These moments of assumed identification not only signalled comfort and openness, but they also highlighted how participants positioned me within their own frameworks of masculinity and solidarity.

However, my gender also posed challenges, as approaching female Uber drivers was marked by hesitancy and suspicion. This wariness was understandable in the context of widespread violence in the ride-hailing industry, where females face particular risks of harassment and assault (Anwar & Graham, 2022; Simpson, 2023).

To mitigate these concerns, I carried printed information letters, consent forms, and presented a copy of my ethical clearance certificate to participants, especially female Uber drivers. In addition, I also asked female Uber drivers where they would be comfortable doing the interviews, and all interviews were conducted in their cars with doors open. I also maintained a respectful physical distance and allowed participants to set the pace of conversation. These subtle, embodied strategies were as important as formal documentation in helping to establish legitimacy and transparency.

My reflexive engagement with safety also extended to my own positionality as a researcher in Johannesburg’s volatile ride-hailing environment (Fokazi, 2021; Simpson, 2023). The risks associated with Uber driving, such as hijackings, violent disputes between ride-hailing and meter taxi drivers, and general urban crime, meant that safety was not only an ethical obligation toward participants but also a necessity for me as well (Fokozi, 2021; Simpson, 2023). Hence, I adopted various fieldwork strategies such as conducting interviews only during the day and in public spaces such as mall parking areas with visible security presence and informing a trusted contact (my cousin) of my location, each day of conducting interviews until the end. I also avoided known high-crime zones also referred to as hotspots by Uber drivers. These precautions reflected some of the safety strategies shared by participants, highlighting that safety is a mutual concern connecting both researchers and participants. Yip (2024) also notes that positionality involves negotiating insider–outsider roles within specific cultural and situational contexts, and in this study, my safety planning was part of navigating these roles in a setting marked by both familiarity and risk. This awareness enhanced my understanding of the field as not just a place for generating knowledge, but also as an environment where embodied vulnerability influences both participation and the collection of data.

Despite these efforts, I could not fully neutralise the uneven power relations embedded in the gendered context of our interactions. Some female Uber drivers still declined to participate, while others maintained a guarded stance during our conversations by providing short or minimal responses. These silences and constraints were not simply methodological issues but reflections of the broader gendered power dynamics shaping fieldwork in precarious and gendered contexts. Feminist and intersectionality-informed theoretical scholarship reminds us that researcher identity is not a neutral lens but one that can constrain access to certain experiential domains (Crenshaw, 1991; Gunaratnam, 2003; Harding, 1991). For example, my positionality as a male researcher may have constrained the in-depth disclosure of certain experiences, particularly those relating to the double burden of paid and unpaid labour, including caregiving and parenting, which can intersect with the unpredictability of digital platform work. Moreover, the study’s binary gender framing meant that non-binary and transgender perspectives were absent, representing a methodological limitation that future research can address and contribute to how gender diversity shapes safety, precarity, and work–life boundaries.

Reflecting on these encounters, I argue that the field is not a neutral space; it is infused with power, history, and personal differences (Gunaratnam, 2003). My experiences resonate with scholarship that foregrounds the fluidity and complexity of researcher-participant dynamics, particularly in studies involving gendered or class-based interactions (Pillow, 2003). As a male researcher interviewing females in a precarious and sometimes dangerous industry, I had to constantly negotiate my positionality and remain sensitive to how my presence may be perceived – not just as a researcher, but as a male occupying a gendered space, especially from the perspective of female Uber drivers. These reflections also highlight the importance of ongoing reflexivity – not only to acknowledge how identity and power shape the research process, but also to navigate relationships marked by vulnerability, discomfort, and inequality. They also speak to the need for what I call intersectional reflexivity ¹ , recognising how multiple axes of difference – including gender, class, and institutional affiliation – intersect to shape the dynamics of fieldwork (Crenshaw, 1991).

Ethical Dilemmas and Time Sensitivity in Researching Precarious Workers

Conducting fieldwork with Uber drivers, individuals whose income is directly tied to ride availability and time spent on the platform, required careful attention to ethical, temporal, and representational dimensions of qualitative inquiry. While I consistently applied core ethical principles such as informed consent, voluntary participation, and confidentiality, several unexpected ethical challenges surfaced during the fieldwork process, requiring flexibility and critical reflection. The most significant of these dilemmas involved time. For gig workers navigating economic insecurity and extended working hours, participating in a research interview without financial compensation typically translates into a tangible loss of earnings. Dowling (2022) and Warnock et al. (2022) argue that individuals in precarious forms of labour often lack the economic freedom to engage in unpaid research, especially amid rising living costs and structural insecurity. This tension was evident early in my fieldwork, as several Uber drivers declined to participate, citing potential income loss. Others signalled conditional interest, proposing informal arrangements such as meeting over drinks, suggesting a form of compensation or reciprocal exchange for them to participate in the study.

These interactions prompted difficult but necessary ethical considerations. While compensating participants can enhance inclusion and acknowledge their time, it also introduces the risk of coercion, potentially influencing the authenticity of participants’ responses. In other words, incentives in research with economically vulnerable populations may lead to strategic or performative participation, especially when participants perceive the interaction as transactional rather than dialogic (Dowling, 2022). Mindful of these dynamics, I chose not to provide compensation, a decision clearly communicated in the participant information letter and reiterated during initial engagements with potential participants.

Although this approach may have reduced participation rates, it upheld the voluntary and non-coercive principles of the study. To mitigate the burden on participants, I made efforts to conduct interviews during off-peak hours, typically after 10:00 am, when many drivers were offline after the morning peak hour. This strategy minimised potential income loss and resulted in interviews conducted during natural pauses in drivers’ work schedules. In several cases, participants expressed greater comfort once they understood the study’s aim and were assured there was no pressure or cost to their involvement.

These moments of engagement were particularly significant, as they reflected a shared commitment to co-producing knowledge, even within the constraints of platform labour. They also highlight the structural tensions of conducting research with precarious workers, where time, money, and trust are all interlinked. Researchers must remain critically reflexive not only about power dynamics and positionality, but also about the material implications of research participation for those already operating on the margins of labour systems (Warnock et al., 2022). This reflexive encounter with the ethics of time and compensation highlights the need to view research ethics not as static principles but as negotiated practices shaped by context.

Reflexivity in Interpretation: Subjectivity and Situated Meaning-Making

Reflexivity remained a critical component beyond fieldwork and extended into the processes of data transcription and interpretation. Before transcribing the interviews, I had to consciously reintegrate myself into the role of researcher, intentionally creating some distance from my prior fieldwork interactions to engage with the data with analytical clarity. This deliberate repositioning helped reduce the likelihood of personal bias shaping the narratives. Attride-Stirling (2001) highlights that researchers must reflect cautiously on their position throughout data analysis to avoid contaminating the research findings.

During transcription, I listened attentively to the recordings while engaging with field notes to ensure that my presentation of participants’ experiences – particularly those of male and female Uber drivers – was accurate and grounded in their own voices. However, I noticed that even this careful process was influenced by my knowledge and interpretations of flexibility, precarity, and autonomy – concepts widely debated in the literature on platform work (Keller, 2023; Standing, 2011). For example, while transcribing narratives about setting earnings targets and choosing working hours, I had to question whether I was framing these accounts within the dominant discourse of autonomy or critically interrogating them as strategies for coping with precarity. I initially accepted the dominant discourse that presents flexibility as an objective benefit of Uber driving, where drivers choose when and where to work. Yet, through critical reflexivity, I revisited these excerpts, re-coded and critically analysed them to capture participants’ uncertainty, highlighting how both male and female drivers experienced precarious flexibility – a term I coined in my PhD study, arguing that flexibility is constrained by precarious conditions such as operating in ambiguity, algorithmic control and management, and safety concerns (Bayane, 2024). This reflexive process actively reshaped my analytic framework, moving it beyond prior assumptions and allowing the development of themes and analysis that authentically represented participants’ lived experience of Uber driving and digital platform work.

To further support this reflexive engagement during analysis, I maintained an audit trail, which Payne and Payne (2014) describe as a structured way of documenting decisions, adaptations, and reflections throughout the study. I kept all my field notes and interview guides, which were not only vital for transparency but also assisted in tracing the evolution of my analytical thinking. For instance, based on participants’ feedback, I modified the language in the interview guide and during interviews – replacing customers with clients – to better reflect the terminology used by Uber drivers in their daily lives. Similarly, I adapted interview questions depending on participants’ circumstances, such as omitting parenting-related questions for drivers without children. These decisions were recorded in fieldnotes, highlighting how reflexivity informed both the analysis and my ethical interactions with participants.

A notable shift in my thinking also occurred when I rephrased the research objective. Initially, I sought to “explore the rise of flexible working conditions of platform work in South Africa,” but engagement with the literature revealed that flexibility is common in Uber driving (Anwar & Graham, 2022). Moreover, the data showed that flexibility, though present, was constrained by algorithmic control, customer demand, and safety concerns – especially for female Uber drivers. I thus came to view flexibility not as a universal attribute of Uber driving, but as a context-dependent and sometimes contradictory feature of platform labour, as my PhD study also argues that it is ambiguous and hides precarity and insecurities in digital platform work (Bayane, 2024).

Gendered dynamics also emerged strongly during analysis, particularly around safety and the double burden of paid and unpaid labour. Pre-existing literature has often centred male experiences in gig work (Woodcock & Graham, 2020), but my PhD study included female drivers, whose accounts revealed distinct challenges, such as safety concerns and the double burden of paid and unpaid labour. In analysing these narratives, I had to remain reflexive about my positionality as a male and how it may shape whose stories I centre and how I interpret them. Zanoni (2019) argues that researchers must embrace situated reflexivity, acknowledging that knowledge is co-produced through the research encounter and shaped by our social positioning. For example, I initially interpreted female drivers’ accounts of avoiding certain areas during the day and night under broad safety concerns in digital platform work. However, upon critical reflection, I realised that my position as a man limited my initial understanding of the distinct, gendered, and structural nature of these concerns. Reflexive engagement made me realise that this broad framing omitted the fundamentally gendered different experiences. For instance, for female Uber drivers, these safety choices were shaped by embodied vulnerability and gendered harassment, whereas for male Uber drivers, safety concerns were more related to general crime, such as car hijackings. This reflexive exercise prompted me to critically analyse participants’ perceptions and experiences of safety, focusing on the gendered nature of safety in Uber driving, which situated the realities of male and female Uber drivers.

Reflexivity in data analysis was thus not a one-time event, but a reiterative and evolving process. It involved returning to the data with a critical lens, being open to discomfort, and recognising how my own beliefs may influence what I noticed or interpreted. While striving to centre participants’ voices, I remained aware that their narratives were being filtered through my analytical lens. I believe this awareness enhanced the rigour, credibility, and ethical integrity of the study.

Methodological Contributions and Implications for Qualitative Research

This study contributes to qualitative methodology by demonstrating how dynamic, embodied reflexivity can be practised when researching precarious digital platform workers. It illustrates practical strategies for navigating researcher positionality, including class and gender dynamics, institutional mistrust, and ethical concerns in time-sensitive, insecure labour contexts (Berger, 2015; Holmes, 2020; Sultana, 2007). By adapting language, modifying presence, and consciously managing asymmetrical power relations, the paper provides an applied model of reflexive practice that extends beyond confessional approaches toward a more situated, critically reflexive methodology (Pillow, 2003; Rose, 1997). These insights align with recent calls for developmental reflexivity in qualitative inquiry and are broadly applicable to scholars working in urban, unequal, and digitalised labour environments (Karcher et al., 2024). Moreover, the emotional demands of conducting sensitive qualitative research necessitate structured self-reflexive practices to safeguard the researcher’s well-being. Karcher et al. (2024) also highlight the importance of such practices, including reflexive journaling and peer support, in managing the emotional impact of sensitive research. Therefore, the paper reaffirms the need to treat reflexivity not as an isolated step, but as a continuous epistemological and ethical stance in qualitative research.

Conclusion

This reflexive paper has illustrated how methodological choices, ethical challenges, and power asymmetries are closely entwined in qualitative research, particularly when engaging with precarious workers in the gig economy. By reflecting on my fieldwork with male and female Uber drivers in Johannesburg, I have shown how positionality – shaped by academic status, class, power and gender – both facilitates and constrains access, trust, and knowledge co-production. Rather than viewing these positional tensions as obstacles to be overcome, I argued that they offer critical entry points for in-depth understanding and more ethical research practice. Reflexivity, as demonstrated in this study, is not an academic exercise but a necessary epistemological and ethical orientation that calls for humility, flexibility, and self-awareness throughout the research process, from access and interaction to analysis and interpretation. In contributing to wider discussions on qualitative methodology and digital platform labour, this paper emphasises that engaging with precarity requires not only intellectual commitment but also emotional and ethical labour. This reflexive account affirms the value of reflexivity in making visible the hidden power dynamics of research and in striving toward more just, inclusive, and accountable forms of sociological inquiry.