Fostering habits of care: Reframing qualitative data sharing policies and practices

Abstract

This article argues that navigating the formalized, regulated and institutionalized data sharing landscape is challenging for researchers tasked with making qualitative data available for re-use. Archiving empirical material has progressively become a requirement enshrined in the policies of UK research institutions. Yet, how qualitative researchers feature – as data (co)creators and curators – within a process governed largely by quantitative data management strategies remains undocumented. Using examples from the ESRC Timescapes initiative, this article argues that to advance ethical practice in qualitative secondary analysis (QSA), data sharing policies and practices need to be re-framed to respect, value and care for the particularities of qualitative data and the emotional, intellectual and temporal investments made by qualitative researchers working in an increasingly pressurized Higher Education (HE) environment. Accordingly, ideas from the ethics of care literature are employed to propose areas where ‘habits of care’, attuned to the needs of qualitative researchers and data, can be fostered.

Introduction

Increasing volumes of research data, along with the development of infrastructure to support data archiving and sharing, has created new possibilities for repurposing all manner of data. Whilst there has been much focus on the re-use of quantitative datasets, this evolving landscape is equally impactful for qualitative research material (Antonio et al., 2020; Bishop & Kuula-Luumi, 2017; Broom et al., 2009; Hughes & Tarrant, 2020; OECD, 2019). Access to research data and the possibilities for using existing material has evolved in alternative ways in different contexts, with the infrastructure enabling the re-use of qualitative material developing more slowly than that of quantitative data (Heaton, 2004). Within the United Kingdom, it has progressively become a requirement enshrined in the policies of many research funding bodies, universities and publishers (Carusi & Jirotka, 2009; Mauthner, 2012; Neale, 2021) with guidance on key expectations outlined in the Concordat on Open Research Data (UKRI, 2016). Sharing research data is seen as a means of maximizing the potential of datasets, using public funds assiduously, ensuring that research endeavours are not duplicated, and that the potential of existing datasets is maximized (Carusi & Jirotka, 2009). Indeed, making research data available for re-use is considered vital to accountability and transparency and for enhancing public trust and support for research (Bishop, 2009; Hughes & Tarrant, 2020; Mauthner, 2012; Mauthner & Parry, 2013; Slavnic, 2017; UKRI, 2016). Interest in re-purposing qualitative datasets has been fuelled further by restrictions placed on data generation during COVID-19 times (Jamieson et al., 2020; Nind et al., 2021).

The sharing of qualitative data and, in turn, qualitative secondary analysis (QSA) has been facilitated by technological advances and a growth in appropriate infrastructure and resources, such as large-scale national archives and smaller specialist repositories (Bishop and Kuula-Luumi, 2017; Edwards and Holland, 2020). The UK Data Archive (www.data-archive.ac.uk/), for instance, houses over 1000 qualitative and mixed methods social science and humanities datasets, offering qualitative researchers many opportunities to ask new questions of historic and contemporary material, and to work across multiple datasets (Boddy and Morrow, 2021; Davidson et al., 2019). The development of infrastructure has been accompanied by the establishment of procedures and practices for data sharing and re-use, as well as much debate about epistemology and the ethics of so doing (e.g. Bishop, 2009; Boddy & Morrow, 2021; Carusi & Jirotka, 2009; Coltart et al., 2013; Irwin, 2013; Mason, 2007; Mauthner et al., 1998; Morrow et al., 2014; Neale, 2013; Tarrant, 2016). How qualitative researchers feature – as data (co)creators and curators – within the formalized, regulated, and institutionalized landscape of data sharing has, however, gone largely undocumented (for exceptions, see Mauthner, 2012; Coltart et al., 2013). Indeed, regulatory frameworks such as the General Data Protection Regulation [GDPR] have focussed on procedural matters at the expense of considering the emotional and intellectual investments necessary for knowledge creation, curation and re-use (Mauthner, 2012).

A focus on the place of qualitative researchers needs to be considered within the context of wider neo-liberal reforms in Higher Education (HE), which, as Edwards (2020) argues shapes ‘the content and practice of academia, and the identities of academics within it’ (p. 9). These include increasing emphases on: auditing, accountability, and surveillance; marketization and massification; managerialism and the casualization of work (Archer, 2008; Holmwood, 2017; Leathwood & Read, 2020; Mauthner & Edwards, 2007; Radice, 2013; Slavnic, 2017). For HE Institutions (HEIs) and researchers, this often means greater scrutiny and regulation over the storage, sharing and reuse of data. Whilst there are benefits to both researchers and research participants, in terms of encouraging effective data management practices and the protection of data, increased bureaucracy and regulation may be experienced as constraining or stifling. For qualitative researchers, the neo-liberal backdrop has seen higher value placed on quantitative inquiry, methods and data management practices (Edwards, 2020; Edwards & Holland, 2020; Slavnic, 2017). Moreover, as Slavnic (2017) argues ‘Neoliberalism also creates a new relationship between power and knowledge that results in new strategies of power use and new technologies of governance’ (p. 250). Little attention has been afforded to how qualitative researchers are impacted by these new technologies of governance.

Changes in the HE landscape such as the casualization of employment and the propensity for many researchers to work on, often multiple fixed term and/or part-time contracts (short-termism hereafter) affect researchers’ capacities to accommodate expectations regarding the making available of data for re-use through archiving (Leathwood & Read, 2020). In 2019/20, 25% of full-time and 49% of part-time academic staff in UK HEIs were employed on fixed-term contracts (HESA, 2020). Indeed, pressures on time and concerns about the precarity of research contracts have become even more evident during COVID-19 times (McKie, 2020). With increased emphasis on accountability and transparency comes the possibility that researchers spend more time ‘reporting on what we do rather than doing it’ (Ball, 2012: 19). Moreover, there is a danger that the requirement to prepare and deposit datasets, often seen as ‘low-skilled relatively invisible administrative work’ (Hadfield, 2010: 71, see also Coltart et al., 2013) is likely to fall on those in more junior positions and may reduce their capacity to pursue analyses of interest and/or career-enhancing opportunities or outputs. Beyond guidance for data depositors on the practicalities of preparing material (e.g. Corti et al., 2019; Neale, 2021), discussions regarding the investments made by, and implications for, researchers are markedly absent.

A key change implicated in the positioning of qualitative researchers to any data made available for re-use is the shift in power and the detachment of intellectual knowledge and resources from producers of that knowledge. Unlike other research outputs, in most cases the Intellectual Property (IP) of research data rests with the research institution where the researcher(s) was employed at the time of production (Summers & Woollard, 2020). Whilst there are some variations to this, such as the sharing of IP between an HEI and a funder in the case of some externally funded work, such moves generally signal the shift from regarding data as a private entity owned morally and legally by an individual/group to a public good, asset or commodity (Mauthner, 2012; Slavnic, 2017). This is particularly evident in regimes of accountability (Archer, 2008) such as the UK Research Excellence Framework (REF, https://www.ref.ac.uk/), designed to measure research performance, and the impact and/or products of research of all HEIs. The REF is highly influential in terms of shaping what is valued as a research output (Edwards, 2020; Kandiko Howson et al., 2018). The forthcoming REF 2021 upholds notions of non-portability placing publications and outputs with institutions, rather than individuals, which can submit outputs whether the creators/authors have moved on, been made redundant, or died (REF, 2019). As this article demonstrates, positioning data as a commodity in the knowledge economy is particularly problematic for qualitative work and, as Slavnic (2017) argues, ‘… directly opposes the central epistemological and methodological principles of qualitative inquiry’ (p. 252).

Framed within the context of wider neoliberal expectations, trends and pressures in UK HE, this article reflects on some of the key impacts of the institutionalization, regulation and standardization, (intended in this context to refer to the expected and prescriptive ways of preparing data for ingestion) of data archiving and sharing on qualitative researchers. It is not my intention to challenge or discredit the project that is data curation and re-use for the merits, and ethical obligations many researchers feel, are clear and have been outlined elsewhere (e.g. Antonio et al., 2020; Corti et al., 2005; Davidson et al., 2019; Hughes and Tarrant, 2020; Irwin et al., 2012; Mason, 2007; Neale, 2021). Rather, I seek to highlight the implications and power dynamics for qualitative researchers, particularly those at earlier stages of their careers, and to enhance debate about what constitutes good ethical practice in data sharing in qualitative research.

This article takes as an example experiences from working on the ESRC Timescapes initiative (https://timescapes-archive.leeds.ac.uk/timescapes/); a collaborative project that developed the specialist infrastructure for the scaling up and sharing of qualitative longitudinal research [QLR] (Neale, 2021; Neale & Bishop, 2012; Neale et al., 2012). To highlight the challenges for qualitative researchers, the article focuses on three inter-related sets of personal and professional investments: time and hidden labour; emotional labour; and interpersonal collaborative relationships between original researchers/teams and secondary analysts. Employing ideas from the ethics of care literature, the article then proposes a means of reframing procedures for sharing and re-using qualitative data(sets) that forge more engrained habits of care that are attentive to and value the investments, often early-career researchers make in generating rich data, and how vital these are to QSA, good ethical practice and wider knowledge production.

An example of data archiving and sharing in practice

The ESRC Timescapes initiative was a 5-year programme of work dedicated to driving forward QLR, including the archiving and re-use of QLR data. Conducted across five UK HEIs, the initiative comprised seven empirical projects each concerned with documenting change and continuity in key relationships and identities across the life-course. Three overarching strands relating to archiving, secondary analysis and knowledge transfer united the seven projects. Time and resources were invested in creating the Timescapes Qualitative Longitudinal Data Archive (Timescapes archive, hereafter; https://timescapes-archive.leeds.ac.uk/timescapes/); discussing the practicalities of preparing qualitative data for deposit; and shaping procedures sensitive to qualitative research. Not only did the initiative support the ethos of qualitative data archiving but it also brought attention to the possibilities and value of reusing data and the ability to work across or bring datasets into conversation (Edwards & Irwin, 2010; Irwin & Winterton, 2014; Irwin et al., 2012).

Each empirical project explored a different facet of the life-course from children’s relationships with their siblings and friends through to the lives of the oldest generation. Repeat interviewing and a variety of creative tools were used across the studies, with data generated over a period of 2–12 years (with the latter in cases where the studies preceded and/or superseded the programme). Each study comprised a small team of researchers – led by one or two Professors with early to mid-career researchers carrying out the bulk of the data generation, analysis and curation. Most of the early-mid career researchers were on fixed term contracts, some of which ended before the close of Timescapes.

My role evolved from early career to senior researcher, albeit on a series of fixed term contracts. My relationship with the data also shifted, from working on one of the precursor studies, to then collaborating with Rosalind Edwards on a core Timescapes project - Siblings and Friends: The changing nature of children's lateral relationships (SAF, hereafter, https://timescapes-archive.leeds.ac.uk/timescapes/research/siblings-and-friends/) and subsequently obtaining additional funding as Principal Investigator to develop a fourth wave of data generation (https://www.ncrm.ac.uk/research/MIP/videotelephony.php, Weller, 2015, 2017). SAF followed the lives of 50 young people from mid-childhood to early adulthood and, through three separate grants, lasted for 12 years. I conducted much of the data generation travelling the length and breadth of the UK and invested time in establishing and maintaining long-term research relationships. I was heavily involved in negotiations around the development of the archive and in preparing SAF material for deposit and re-use (Edwards and Weller, 2011).

More recently, I co-led the day-to-day management of a methodological study - Working across qualitative longitudinal studies: A feasibility study looking at care and intimacy (BigQual hereafter, http://bigqlr.ncrm.ac.uk/) - that formed one of the National Centre for Research Methods’ key work packages. Working across six Timescapes studies, including SAF, this project was concerned with exploring the feasibility of conducting QSA across multiple datasets and resulted in the development of a new breadth-and-depth method for large-scale qualitative analysis (Davidson et al., 2019; Edwards et al., 2020). As part of this work, we consulted former Timescapes colleagues over our plans to conduct QSA across the datasets (see Weller, 2019 for further details). My own position shifted from data generator and curator to that of secondary analyst, offering a different lens through which to consider the institutionalization, regulation and standardization of data archiving and sharing.

Personal and professional investments

To explore the key impacts of data sharing policies and practices on qualitative researchers and qualitative work, this section focuses on three inter-related sets of personal and professional investments: time and hidden labour; emotional labour; and interpersonal collaborative relationships between original researchers/teams and secondary analysts.

Investments of time and hidden labour

The first set of investments concerns the time and hidden labour involved in data curation. Despite the growing compulsion to prepare data for archiving and re-use, the interconnected, often complex, and time-consuming processes involved, are afforded little recognition (Coltart et al., 2013; Thomson & Berriman, 2021). This lack of appreciation of the value of such work often sees labour-intensive data curation being allocated to early-career researchers. Whilst more senior academics may have the final say with respect to decision-making, some may feel that managerial and other responsibilities limit their opportunities to contribute actively to the miniate of the process (see also Mauthner, 2012; Mauthner & Edwards, 2007). Ostensibly, tasks associated with data curation overlap with good data management practices, for instance, diligent file organization, attention to compliance with ethical guidelines and national or cross-state legislation such as GDPR (2016, EU). In addition, care needs to be taken over how material is represented (Neale, 2021). The SAF study involved preparing three waves of data, from nearly 50 young people for deposit in the Timescapes archive. The resultant 145 transcripts and audio files of in-depth interviews, 500 activity sheets and contextual notes, required over 600 hours of detailed pseudonymization, removal of personal direct and indirect identifiers, digitization, checking and editing, and comprehensive metadata documentation. This time did not include decision-making around pseudonymization, balancing data integrity with concerns about confidentiality, controls on access to and/or the safeguarding of specific cases or data files, and the development of project guides and resources to promote and support QSA (Weller et al., 2011). For more senior researchers/project leads, there are often hidden investments of time in terms of mentoring and supporting colleagues. All these facets of the process, including the preparation and ingestion of data, need to be adequately priced and timetabled in grant proposals (Thomson & Berriman, 2021).

The Concordat on Open Research Data (UKRI, 2016) does recognize the demands on researchers’ time, along with the possibility that such work may extend beyond the lifetime of a project. Yet providing open access data is positioned as the researcher’s responsibility and is expected ‘within a short and well-defined period’ (UKRI, 2016: 7). Indeed, it is also possible that an archive may choose not to accept a dataset or may not have the resources required to ingest it. Furthermore, well managed, compliant qualitative datasets are not necessarily ready for sharing without further investments of time. The labour of preparing qualitative data for re-use by others involves a range of complex ethical and epistemological decisions not provisioned in standardized data management guidance. Rather than the relatively simple administrative task it is often purported to be, much of the work revolves around the ethics of balancing data integrity with concerns about key issues such as anonymity and confidentiality (Hadfield, 2010; Neale, 2021). As Moore (2012) argues, ‘the technical is never only technical’ (p. 335) and that processes such as anonymisation have ‘…a history – and a politics –as well as an ethics’ (p. 332). Generating and preparing data for a study is, therefore, a very different prospect to organizing it for others to use (Carusi & Jirotka, 2009). Providing enough contextual information to render the data useful for QSA whilst ensuring that sufficient pseudonymization has been undertaken requires deliberation regarding the granularity of the stories narrated. The removal of direct and indirect identifiers from individual files may not be enough, for example, where the material contains multiple family members or an organization, is longitudinal, comprises visual data or includes highly sensitive material that may render a participant(s) visible. Decision-making around whether data can be archived or what level of access might be appropriate requires complex ethical deliberations and the development of appropriate protocols (Neale, 2021). The size of this endeavour is often under-estimated, presumably in part, because procedures are, as Parry and Mauthner (2004) argue ‘…derived from a positivist, quantitative model where the researcher tends to be rendered invisible in the research process’ (p. 145; see also Antonio et al., 2020; Mauthner & Parry, 2013; Moore, 2007; Slavnic, 2017).

With an increase in short-termism, greater competition for funding, the demands of formalized research assessment processes, and a culture of working long hours (Kinman & Wray, 2013; Leathwood & Read, 2020; Slavnic, 2017), added pressure has been placed on researchers to prioritize the production of quality publications in high impact journals, gain prestigious research grants, and ensure their work has impact outside academia (Edwards, 2020). This echoes Kandiko Howson et al.’s (2018) arguments regarding the prestige economy, which highlight the valorization of metrics such as publication rates, author order, invited international presentations, and editorial positions. Preparing and making datasets available for re-use, whilst regarded as a key responsibility for researchers (UKRI, 2016) and vital to knowledge production, is not yet held in the same esteem.

Pressures to make data available within a reasonable timeframe, coupled with competing demands from other, often more valued activities, render the benefits for individual data depositors limited. For researchers, the prevalence of short-term research contracts may mean that many simply do not have time to have ‘reasonable first use’ of data, as outlined in the Concordat on Open Research Data (UKRI, 2016), before a project ends and/or sharing is expected. Equally, after a study is complete, many may not have the time and/or resources to prepare a dataset for re-use (Neale, 2021). Alternatively, Timescapes was conceived as a ‘living archive’ with data deposited throughout, rather than at the end of each study. This strategy is suited particularly to QLR, which is often an on-going and evolving endeavour. In theory then, material could be used by others before the original researcher(s) had completed their own analyses and/or published key findings. For the Timescapes team, this was a key point of deliberation, with discussions around emergent good practice founded on a stakeholder approach designed to be attentive to the needs and interests of all those involved, from the original researchers through to universities and funding bodies (Neale, 2013).

In the UK, data re-users must acknowledge the scholarship of the original researcher/team in generating a dataset by including the appropriate citation in outputs resulting from re-use. Furthermore, datasets will feature, for the first time, as an output for the REF 2021. Whilst this move may help raise the status of datasets as outputs, it may not result in all members of research teams receiving equal recognition. If, for instance, a researcher relocates to a new institution, credit remains with the HEI where the study was undertaken. This is likely to affect disproportionately those in more junior positions and/or on fixed-term employment contracts who often must move between institutions to take up new opportunities. Moreover, it is probable that early-career researchers would not be eligible for the REF, with datasets then accredited to more senior academics. Indubitably, whilst the REF is not the only means of understanding academic outputs, it is highly influential in shaping the metrics governing what is valued as an output and is, therefore, of importance to career progression (Edwards, 2020; Kandiko Howson et al., 2018).

This example serves as an illustration of the temporal resources, expertise and skills required and the challenges faced by competing demands that disproportionately affect those in more junior positions and/or on fixed-term contracts, who are likely to complete this hidden labour potentially at the expense of other career-enhancing tasks (Coltart et al., 2013).

Emotional investments and labour

The relational nature of qualitative knowledge production means that researchers’ emotional investments and labour are manifest both in the data archived and in the process of curating material for re-use. Yet beyond writings on the emotional effects of qualitative work, particularly fieldwork, on researchers (e.g. Hochschild, 1983; Holland, 2009; McGarrol, 2017), little attention has been afforded to this in data sharing policies and guidance (Morrow et al., 2014). Drawing on feminist new materialist theorizing, Corple and Linabary (2020) illustrate how researchers become ‘disembodied and placeless’ from online big data and similar arguments could be made regarding the archiving of qualitative material. Data offered for re-use is shaped by researchers’ experiences and the intellectual investments of the wider team. Those ‘in the field’ often share aspects of their own lives as part of the processes of building trust, rapport and developing research relationships (Carusi & Jirotka, 2009; Parry & Mauthner, 2004). I spent hours with SAF participants over a 12-year period, visiting their homes, meeting family members and engaging in lengthy discussions that also captured aspects of my own personal and professional trajectories from early career to experienced researcher, with motherhood thrown into the mix. Whilst acknowledging the epistemological challenges in so doing (see Fielding, 2004; Moore, 2007), I invested a great deal emotionally in creating, what I perceived to be, important contextual material through field notes and reports. I wrote detailed reflexive accounts documenting observations and feelings about fieldwork journeys, participant’s homes, greeting/leaving interactions, and the interview dynamic until my midwife declared I had to stop work. It felt like I gave my all to ensuring that an ‘unknown someone’ might, at some unspecified point in the future, better understand the decisions I (and the team) had made and make meaning from the material. As Mauthner (2012) argues, ‘Data producers, and their methods and practices, help produce, and are part of, the data and phenomena they constitute’ (p. 12). This view of data as co-constructed situated narratives complicates notions of data as being owned by researchers, participants, or institutions (Antonio et al., 2020; Broom et al., 2009; Parry & Mauthner, 2004).

Regulatory frameworks, fuelled by a concern for data protection/mitigation against breaches, prescribe the need to remove personal direct and indirect identifiers to ensure individuals and collectives cannot be recognized (Antonio et al., 2020; Mannheimer et al., 2018). Researchers are, however, not subject to the same consideration and are not anonymous (Carusi & Jirotka, 2009); rather they are susceptible to scrutiny. Indeed, procedures for gaining consent for re-use position researchers as responsible for ensuring participants understand to what they are agreeing. Given then the emotional investments made by both researchers and research managers (Mauthner & Edwards, 2007), some may feel vulnerable or exposed personally and/or professionally. In contrast to quantitative data, qualitative material is often regarded as co-constructed; the product of the interaction between the participant(s) and researcher(s) (Boddy and Morrow, 2021; Edwards and Weller, 2012; Holstein and Gubrium, 2003; Kvale and Brinkmann, 2009; Ruggiano and Perry, 2019; Thomson, 2010). The identities of researchers and what they reflexively reveal of themselves, how they interact with participants, their techniques and approaches and the messiness of qualitative work are laid bare within the artefacts of qualitative data. Even with a commitment to data sharing, willingness is often not without reservation. Key concerns comprise challenges to individual and group credibility, standing or reputation, and a lack of duty of care in the way data may be handled and/or represented by others (Boddy & Morrow, 2021; Hadfield, 2010; Morrow et al., 2014; Neale, 2013).

Inter-personal relationships are crucial to qualitative work. Yet, the ways in which caring practices infuse the process of data preparation have been neglected. Emotional investments arguably transcend regulatory frameworks and procedural discussions surrounding consent or data management, and it is often those with direct relationships to participants, who feel a sense of moral responsibility towards their rights and wellbeing, and any data relating to their lives, particularly how it might be re-purposed (Carusi & Jirotka, 2009; Hadfield, 2010). In SAF, for instance, decisions about data curation were guided by a relational and contextual feminist ethic of care (Edwards & Mauthner, 2012). Accordingly, the inclusion of material was not only determined by formalized consent processes but through continuous dialogue with participants, reflections regarding the wider context of participant’s lives, and our own sense of moral responsibility. Many of the young people were keen advocates of data re-use, and participant support for such endeavours places a further ethical duty on researchers and institutions to make data available. Nevertheless, over time some participants divulged aspects of their lives that could severely compromise key relationships or cause moral harm (see also Mauthner, 2012). I felt protective of these aspects of their stories and, in consultation with the team, decisions were made on a case-by-case basis to embargo segments of transcripts (Weller et al., 2011). A further poignant example concerned the tragic death of a participant, which raised many issues surrounding our feelings of care and moral obligation towards his family and to the positioning of the data generated (please see Edwards and Weller, 2013, 2015). For researchers then, regulatory frameworks and data management policies gloss over the challenges that relationships and practices of care bring to the process of preparing datasets for re-use. In the final section, of this article I focus on how ideas from the ethics of care literature might help reframe qualitative data sharing policies and practices to respect, value, and care for researcher’s investments.

Investments in collaboration and collegiality

The third set of investments concern collaboration and collegiality as a marginalized aspect of data sharing practices. The stakeholder approach to archiving adopted by the Timescapes team sought to not only forge close working relationships between researchers, but also between researchers, archivists/data managers and data re-users ensuring that procedural matters were sensitive to, and appropriate for, complex qualitative data (Neale, 2013; Neale & Bishop, 2012). That said, collaboration is generally not a requirement or expectation of many data sharing policies. The emphasis on open access via centralized repositories and processes imposes a formalized and perhaps more distanced means of data sharing, rather than through personal and professional collaborations of the researcher’s and/or teams’ choosing (Mauthner & Parry, 2013). The general assumption is that re-users will work with archived material independently of the original researchers/teams. New users may not regard the explicit involvement of primary researchers as necessary or desirable instead viewing data as embodying new knowledge or alternative insights. Moreover, it may simply not be possible to work collaboratively especially in the context of more historic datasets or where researchers have retired, moved on to new projects, or embarked on careers outside academia.

In practice, the distinction between the primary and secondary use of qualitative material, and associated ideas about a researcher’s proximity to, or distance from data, are far from clear-cut (Coltart et al., 2013; Hughes et al., 2021). There is a small, but growing movement towards more collective endeavours (see Bornat et al., 2008; Hughes & Tarrant, 2020; Neale, 2021). As Tarrant and Hughes (2020) argue, such approaches disrupt the boundaries between ‘primary’ researchers/teams and new users of the data. In our own work, members of two of the original Timescapes project teams collaborated to conduct large-scale QSA across the datasets as part of BigQual (Weller, 2019). Developing ideas around good secondary analytic practice featured as a key part of this work. Guided by our perceived duty of care and shaped by our own understandings of the temporal and emotional investments involved in qualitative work, we consulted former Timescapes colleagues over plans to re-purpose the datasets (Weller, 2019). In the early stages, we liaised by email with individuals, asking questions oriented to data management or context. Whilst inclusivity was our intention, in practice we spoke with only one or two members of each team; those with whom we had strong professional relationships. Later, we took a more formalized approach using an anonymous online consultation to garner insights into changing connections to the data, feelings and concerns about data sharing and re-use, and the forms of collaboration deemed appropriate or valuable. Sustained collaboration was not uniformly desired, expected or possible to accommodate, with some constrained by time, funds, workload or longevity of employment contract. Indeed, early-career/field researchers are often best placed to enlighten secondary analysts on the minutiae of a project but may be hindered by time pressures, competing demands and short-termism. Some colleagues had left academia for new ventures or were not available. We also became concerned that our reliance on goodwill could result in exploitation, leaving those in precarious employment situations feeling obliged to make further hidden investments of time.

Some were still using Timescapes data, whilst others explicitly stated that they felt more detached from the material over time. After spending 4 years working closely with multiple datasets, we conversely felt a growing sense of connection to participants and their accounts, and to the commitments of the original researchers. We felt vulnerable in making public our approaches and analyses, but also had positive experiences of others re-purposing and publishing new interpretations of the SAF data (e.g. Wilson, 2014). The temporal and emotional labour we invested in QSA meant that we became attached to it as our production, thereby shifting our perception of ownership. Indeed, we re-constructed the material in different ways including producing and archiving a new dataset (Weller et al., 2019). Making research data available for re-use implies that knowledge production has not ended. QSA disrupts usual understandings of collaboration introducing it as emergent, iterative and unexpected.

Given the issues outlined above, what then are the necessary conditions for the ethical, responsible, and respectful sharing of qualitative data, and how can we try to ensure the investments made by qualitative researchers are not hidden or lost amidst regularity frameworks and obligations?

Ethics of care in data sharing

In this final section, I employ ideas from the ethics of care literature (Edwards and Weller, 2013; Tronto, 1993, 2013), notably the thinking of Joan Tronto (1993, 2013), to consider some of the ways in which data sharing policies and practices might be re-framed to respect, value, and care not only for the investments of research participants but also those of researchers. In so doing, the aim is to ameliorate some of the constraints and challenges afforded by the wider neoliberal context in which we, as researchers, are working. Although not originally applied to data sharing, Tronto’s focus on ‘habits of care’ (1993: 127) – or the disposition to care – is particularly valuable in thinking with and adopting caring practices in relation to the processes that enable qualitative data to come into fruition, to be made available for others, and to be re-used. Similarly, but with reference to their work with those in creative industries, Alacovska and Bissonnette (2019) argue the case for ‘adapting an ethics of care approach that recognizes deep relationality, human interdependencies, the needs of others, and hands-on caregiving as the basis for action and work’ (p. 3). This, I argue, could help develop ethical practices in data sharing that are more attuned to the needs of qualitative researchers, data, and research participants.

Tronto and Fisher (1990) outline four aspects of care: caring about, caring for, caregiving and care-receiving to which there are four related ethical values. The first of these, attentiveness, concerns the disposition to identify a care need and an appreciation that the need must be met. The second is the willingness to assume responsibility for and attend to that care need(s). The third, competence, relates to ‘caregiving’ or responding pragmatically to the need(s). Finally, responsiveness is concerned with ‘care receiving’ or the way in which care is received. Within the context of data sharing and re-use, there are multiple levels at which different care needs could be addressed. Integrating these ethical values into data sharing policies and practices could enable more meaningful ‘habits of care’ to be fostered. Based on the investments outlined in this article, I offer five examples.

Caring for and about investments in qualitative data production

An attentiveness to the intellectual, emotional and temporal investments made by both researchers and participants is a key starting point in terms of forging more engrained habits of care. Caring for and about investments in qualitative data production and curation requires a willingness, by the institutions driving data sharing agendas (e.g. universities, funding bodies, publishers), to take responsibility for the way such work is represented and valued. This involves not only regarding qualitative endeavours in higher esteem but also looking beyond the procedural data management elements of the task(s) to acknowledge that well managed, compliant qualitative datasets, are not necessarily ready for sharing without further investments of time. To achieve this, requires competence in caring for qualitative researchers and data by elevating, for instance, in funding application procedures and training opportunities, the work associated, to recognize adequately the time and resources necessary and that it is a complex interpretative process with ethical and epistemological implications (Neale, 2021). In terms of responsiveness, researchers, at all career stages would then judge if, when and how this care need has been met and could contribute to raising awareness by (continuing to) share, through social media, (in)formal networks, special interest groups, dedicated sessions at professional conferences, and publications, detailed discussions about the investments made.

Thinking with care about researchers’ differing positions in the wider HE landscape

Given that it is often researchers in more junior positions who are likely to be investing heavily in the preparation of qualitative data for archiving and sharing, we need also to be attentive to the ways in which wider trends in the HE landscape, notably short-termism impact on academics at different career stages. Obligations to make data available within an acceptable timeframe often impact on researchers’ capacity to complete tasks regarded as vital to career progression. The pressures more senior researchers/project leads face also need to be considered to value the hidden labour involved in mentoring and supporting colleagues, especially on an informal basis. Taking responsibility for this requires structural changes to address the precarious employment situations of many. In terms of competence, the care-giving element concerns the provision of adequate terms of employment, or more stable positions, that allow (more than) sufficient time and resources to prepare qualitative datasets, as well as have the opportunity to conduct analysis, publish and apply for funding. Ensuring greater transferability of credit for the creation of datasets when researchers move to new institutions is also necessary. For all career-stages, training in data archiving, sharing and QSA must transcend procedural matters to focus on the intellectual, ethical, and emotional demands. Again, the effectiveness of such shifts would be evaluated by those inhabiting precarious positions, but responsiveness could also include collective advocacy amongst researchers.

Being careful about the place of researchers in the production of qualitative data

We need to think carefully about the co-constructed nature of qualitative data and how data management policies, protocols and practices might be reframed to be more assiduous to the rights of all those involved. Attentiveness in this respect means caring for the nature of qualitative work, for research participants, and for the position of researchers, whose identities, professional standing, and personal lives are likely to feature as an integral part of the data and potentially subject to exposure or scrutiny (Moore, 2012). Assuming responsibility requires a shift in regulatory frameworks and data management policies away from subjecting qualitative material to the same procedures as quantitative datasets in which researchers are often invisible (Parry & Mauthner, 2004). This view complicates both notions of data ownership (Antonio et al., 2020; Broom et al., 2009) and brings into question whether consent to archiving procedures not only address participant’s rights but also those of researchers. In terms of competence, providing successful care necessitates that institutions driving the open access agenda, along with data repositories develop, in collaboration with key stakeholders, policies, protocols and practices that reflect an understanding of data as constituted in different ways. This would also help ensure that the presence of qualitative researchers, their professional standing and aspects of their personal lives within the data are respected. Beyond assessing whether structural changes would meet such care needs, responsiveness could also include research teams developing their own policies on data re-use.

A relational view of care and responsibility in research teams

We need to be attentive to power dynamics within research teams and the ways in which, often early-career researchers, complete many of the tasks associated with preparing datasets for archiving and sharing. Recognition also needs to be given to the pressures placed on more senior researchers/project leads who may wish to be involved in the process but face a range of other competing demands. In thinking about responsibility, it is useful to draw on McLeod’s (2017) reframing of responsibility as a ‘relational disposition’ (p. 43). McLeod argues that attention needs to shift away from neo-liberal ideas of individual responsibility and all the inequalities of power imbued in such notions, to a more relational view. In these terms, competence includes fostering more collegial practices of data archiving and sharing in which interdependence, shared research tasks and decision-making thrive. Authors such as Mauthner and Edwards (2007) offer the possibility of teams, developing their own ethical guidelines for working together at the outset. This could be scaffolded by the development of codes of practice for data preparation and sharing by relevant professional bodies. Furthermore, access to re-use must not be at the expense of more marginalized members of a research team. In essence then, fostering more engrained habits of care requires team members ‘acting “other-wise” instead of “self-wise”’ (Alacovska & Bissonnette, 2019: 4). Indeed, responsiveness might also include the collegial sharing of examples of caring practices.

Thinking carefully about where and how qualitative data is archived

Whilst there are considerable benefits to making data available through centralized repositories, we need to be attentive to the implications of the potentially restrictive nature of their structure(s). The organization of such repositories, and the standardization of procedures may not be fitting for, and flexible enough to accommodate the complex, evolving and emergent nature of qualitative data (Antonio et al., 2020). Taking responsibility for this necessitates a willingness by, for example, funding bodies to consider other means of sharing and repurposing data as both valid and valuable, and to respect all stakeholders by democratizing the process. This is likely to involve thinking carefully, critically, and creatively about what constitutes an archive, what it means to co-construct and curate qualitative data. This would certainly require some degree of support for researchers to guarantee data protection compliance, and care amongst both original researchers and new users of the material to ensure good ethical practice. Competence concerns enhancing researchers’ skills in developing researcher and participant-led archiving practices and opportunities, and encouraging alternative means, for example, in smaller, specialist repositories such as the Timescapes archive or through co-constructed community archives (see Moore et al., 2017). Thomson and Berriman (2021) propose one example. In developing the ‘Everyday Childhoods’ (https://sussex.figshare.com/Everyday_Childhoods) open access collection, archiving was positioned as the starting point of the project and a collaborative and prospective approach ensued. In terms of responsiveness, this might include greater emphasis on whether researchers, and indeed participants, feel they have any autonomy in determining and shaping how and where qualitative data are archived.

Conclusions

The research data landscape is evolving rapidly. Global digitisation, increasing volumes of, and accessibility to data, and the development of infrastructure to support data sharing has created new possibilities for repurposing qualitative data. Within the UK, and well as the USA and in Europe, making datasets available for re-use has progressively become a requirement enshrined in the policies of many research funding bodies, universities and publishers. Indeed, as Thomson and Berriman (2021) argue researchers in the UK increasingly must include long-term provision for material in data management plans ‘with a view to archiving by default’ (p. 11). This article has examined how navigating the formalized, regulated, and institutionalized landscape of data sharing is particularly challenging for qualitative researchers and qualitative work. Designed largely with quantitative data management strategies in mind, the policies that govern data sharing are not attentive to the co-constructed nature of qualitative data.

Moreover, regulatory frameworks often focus on procedural matters at the expense of considering the emotional, intellectual, and temporal investments necessary for knowledge creation and curation in qualitative work. Well managed, compliant qualitative datasets are not necessarily ready for sharing without further investments of time. The labour of preparing qualitative data for re-use by others is not a simple, menial task. Rather, it involves a range of complex decisions not provisioned in standardized data management guidance. Considering these investments in the context of neo-liberal reforms illuminates the implications, particularly for those in more junior positions and the increasing number of HEI staff on fixed-term contracts. Opportunities for career progression are tied to the production of particular types of research output (Edwards, 2020). At present, the production of an archived dataset is not held in the same esteem as other endeavours. Moreover, such policies can also perpetuate power relations within teams leaving data co-producers most vulnerable to scrutiny.

Integrating and adapting ideas from the ethic of care literature, particularly the work of Joan Tronto, reframes the discussion, diverting attention from purely procedural matters to encourage us to think with care about the particularities of qualitative data and the emotional, intellectual and temporal investments made by qualitative researchers working in an increasingly pressurized HE environment. It is also helpful in emphasizing the contributions made by research participants both in terms of the co-production of data but also in any negotiations and decision-making about what to archive. Integrating ethical values such as attentiveness, responsibility, competence and responsiveness into data sharing policies and practices has the potential to enable more meaningful ‘habits of care’ to be fostered and to challenge some of the constraints afforded by the neoliberal HE context. It is essential to recognize that the responsibility for infusing habits of care into the process lies not only with individuals or research teams but, importantly, in wider structural change. Such moves are vital to driving forward ethical practice in QSA and wider knowledge production.