Abstract
Fundamental to action research in healthcare is the co-production of knowledge by researchers and participants, including professionals, patients and policymakers. Despite high expectations of co-produced knowledge, few scholars have reflected critically on how researchers navigate situations in which different kinds of knowledge are articulated and potentially clash both at the front- and backstage of research projects. Drawing on our own experiences in various healthcare action research projects, we analyze the ways in which action researchers can relate to the ‘epistemological politics’ of action research, i.e., the negotiations and tactics that include and exclude certain types of knowledge and knowers in the production of scientific and societal insights. To understand this ‘relating of researchers’ and its implications for knowledge production, we draw on the concept ‘sorting attachments’, which we conceptualize as the practical work of figuring out how to relate to particular individuals, agenda’s, institutions and organizations, while at the same time acknowledging the wider political, economic and cultural relations that shape the normativities of stakeholders in action research projects. We distinguish different dimensions of sorting attachments; latching attachments, extending attachments, balancing attachments and foregrounding and backgrounding attachments. By teasing out these forms of sorting attachments, as well as their implications for knowledge productions, we provide researchers with heuristics to navigate contested knowledge claims in action research projects.
Keywords
Introduction
In the study of changing healthcare systems, action research is increasingly recognized for providing valuable embedded insights into healthcare practices while also co-producing situated interventions with multiple stakeholders that reconfigure those very practices and empower marginalized actors (Abma et al., 2019; Cordeiro & Soares, 2018; Heerings et al., 2022; Hughes, 2008; Kjellström & Mitchell, 2019; Morrison & Lilford, 2001; Woelders & Abma, 2019). In the past two decades, action research has been applied in many different healthcare settings and domains, varying from co-design in healthcare quality improvements trajectories (Abma et al., 2019; Donetto et al., 2015; Heerings et al., 2022; Robert, 2013; Woelders & Abma, 2019); organizational change and innovation processes (Bridges & Meyer, 2007; Day-Duro et al., 2020; Waterman et al., 2007) to public health and community-based health initiatives (Barrington et al., 2016; Gibbon, 2002). These studies demonstrate that action research can help instigate organizational and social change, increase the uptake of health interventions, and emancipate marginalized groups or individuals by providing communicative spaces and involving them in the co-design of healthcare (Kjellström & Mitchell, 2019).
Due to the growing popularity of action research in social care and healthcare settings, scholars have recently called for the need for more reflection on the role and practices of the action researcher (Kjellström & Mitchell, 2019); roles and practices that are partly out in the open and partly hidden from view, taking place at the backstage of academic knowledge production and the creation of societal impact (Abma et al., 2025). Especially in healthcare settings where co-production of knowledge is viewed as an important aim, power differences between key stakeholders (i.e. patients, family members, doctors, nurses, policymakers) often play out (Ärleskog et al., 2021) and stakeholders may clash over the legitimacy of co-produced knowledge and knowledge informed interventions (Kjellström & Mitchell, 2019; Van de Bovenkamp & Zuiderent-Jerak, 2015; Vennik et al., 2014; Verhage et al., 2024; Woelders & Abma, 2019). This is what Tanenbaum (2005) understands as ‘epistemological politics’, i.e. the negotiations and tactics that include and exclude certain types of knowledge and knowers. A clear example are the conflicts that emerge in action research projects, when evidence-based knowledge of doctors clashes with more experientially based forms of knowledge of patients (Kuijper et al., 2024). While many studies have paid attention to the methodological and ethical challenges of action researchers that need to work in settings with power imbalances (Gustafson & Brunger, 2014; Khanlou & Peter, 2005; Wilson et al., 2018; Øye et al., 2019), scholars have paid less attention to the contested nature of knowledge co-production as a process that evolves over time and in which the action researcher needs to (re)position him/herself/themselves. In this paper, we therefore draw on our own experiences of action research in (health)care and use the concept ‘sorting attachments’ to analyze the various types of practical work researchers do to figure out how to relate to particular individuals, agenda’s, institutions and organizations’ as well as the cultural, political and economic relations in which normativities of stakeholders in action research projects are embedded (Bruun Jensen, 2007). By excavating various practical ways of relating to epistemological politics in action research, as well as teasing out its consequences for knowledge production, we provide researchers with heuristics to navigate the at times intricate terrain of working with divers actors and agendas in action research.
Theory
Action research comprises a ‘family of approaches’ that are participative, grounded in experience and action-oriented, their aim being to intervene meaningfully in practice (Reason & Bradbury, 2008). Numerous approaches can be considered action research, including appreciative inquiry (Cooperrider & Srivastva, 1987); participatory action research (Fals Borda & Rahman, 1991) and co-operative inquiry (Heron, 1996). They share a number of characteristics that set them apart from conventional qualitative research, although some of the features we describe in this paper also matter for qualitative approaches more generally. Action research has a distinct ‘purpose’ (Reason & Bradbury, 2008); it investigates practical concerns of and with interlocutors and aims to produce knowledge ‘useful to organizations and communities’ (Bjørn & Boulus, 2011). Action researchers, as embedded scholars, derive insights from and intervene in practices to make transformations possible (Hölscher et al., 2021). As others have argued, this purpose does not necessarily compromise conventional scientific knowledge production because action research can contribute to social scientific theory and concepts (Heron, 1996; Zuiderent-Jerak, 2015). Action research involves groups with varying interests in a ‘collaborative,’ ‘participatory’ and ‘inclusive’ endeavor in which knowledge is produced and interventions are undertaken with practitioners (Bjørn & Boulus, 2011). The extent to which stakeholders participate in designing and conducting the research varies from full participation in the problem definition, questions, data collection and data analysis to partial participation in certain phases.
Ontologically and epistemologically, action research defies the positivistic idea that there is a reality out there waiting to be discovered using sophisticated research methods (Heron, 1996). Action research starts from the premise that ‘a single truth’ is a deeply problematic notion (Coleman, 2015, p. 3). Social reality is an ongoing social construct; the researcher is an inherent part of that construct, and all research by definition involves interpretation. Action research does not fully endorse a constructivist position into relativistic extremes, however, since it is committed to universal ideals, such as social and environmental justice and aims to improve things in the world (Coleman, 2015; Heron, 1996). The epistemology of action research recognizes that ideals need to be co-constituted and contextualized in co-production with stakeholders and social truth is always partial. This is particularly relevant for action researchers committed to collaborative work with interlocutors, for instance in problematizing problem definitions and in processes of sensemaking and practical theorizing. As Coleman (2015) argues, ‘we must, therefore, be open to multiplicity, to emergence, to partial and sometimes conflicting truths’ (2015, p. 4).
In the research process, this multiplicity often creates tension between participants, with groups struggling over who gets to foreground their ‘definition of the situation’ (Goffman, 1959). As action researchers, we are often implicated in struggles over meaning, since endorsing research as a collaborative practice means acting with different groups in our projects. Action researchers are thus often involved in what Tanenbaum (2005) calls ‘epistemological politics,’ i.e. the negotiations and tactics through ‘which knowledge and [certain] knowers are to be privileged’ (2005, p. 163). Epistemological politics requires researchers to maneuver and position themselves carefully in relation to struggles over meaning. This careful positioning is what Bruun Jensen (2007) calls ‘sorting attachments,’ where sorting refers to the ‘practical activity’ of researchers as they figure out how to relate to individuals, ‘organizations, institutions, agendas’ (Bruun Jensen, 2007, p. 239), and where attachments implies that ‘no such engagement is innocent’; individuals are always implicated in a web of ‘cultural, political and economic relations and institutionally sanctioned commitments’ that shape the normativities researchers encounter and affect the epistemologies they produce (Bruun Jensen, 2007). How sorting is done in practice, and which different types of sorting attachments there are at play in knowledge production, is the key topic of this paper.
Methods
Our inquiry into the epistemological politics of action research is based on our personal experiences as engaged healthcare action researchers (Benoot & Bilsen, 2016). Working in the same research department, and as coordinators in different action-research projects, we often discussed the challenges of action research in informal, ‘backstage’ settings (Goffman, 1959; Waring & Bishop, 2010). These informal exchanges piqued our interest in – and concern about – the contested nature of co-produced knowledge and the consequences of epistemological politics for our role as action researchers. In particular, we discussed the different expectations that various stakeholders (research participants, academic colleagues, funders) have about action research, our decisions about which topics to address and the consequences of foregrounding some objects of analysis over others (McAllum, 2018). We also wondered how to make our engagements visible and how to account for them within the academic environment in which we operated and built our careers.
Summary of the Three Projects
Prompted by our informal conversations backstage, we decided to collect and thematize our experiences and confront the epistemological politics we recognized in them, as well as how we related to it (Benoot & Bilsen, 2016). The ‘we’ in this sense, refers to the authors of this paper and our colleague academic researchers working on the aforementioned project that shared stories about epistemological politics during the research process with us. We explored and structured these experiences in three steps: I) sharing experiences during organized sessions on epistemological politics as well as the ways we related to this as researchers; II) writing them down in narratives; III) ordering them thematically. Our aim was to develop a coherent and theoretically informed reflection on our experiences with epistemological politics in our action research projects (Chang, 2008), iteratively moving back and forth between the narratives, the themes, and the literature (Tavory & Timmermans, 2014). It was during this last stage that we became aware of the concept of sorting attachments; as the way in which researchers relate to the interests and agendas of fellow stakeholders in the action research projects. Informed by this sensitizing concept, we identified distinct types of sorting attachments in our narratives and analyzed their consequences for knowledge production. These types are respectively latching attachments, extending attachments, balancing attachments, foregrounding/backgrounding attachments.
Results
Below, we analyze different ways in which we, as researchers, relate to and navigate the epistemological politics in action research. We use the concept ‘sorting attachments’ to explicate the practical work of researchers of relating to individuals, organizations, institutions and agenda’s and tease out its implications for knowledge production (Bruun Jensen, 2007).
Latching Attachments
Latching attachments refers to accepting and working with pre-determined problem definitions and solutions. As action researchers, we often have to work with pre-established problem definitions and solutions, and this often results in knowledge that affirms an institutional status quo. In textbook action research, participants are expected to co-construct and reflect together on the problem definition at the start of a project. In reality, however, the problem definition is often formulated prior to the action researchers’ involvement, especially in projects co-financed by healthcare organizations (or third-party funders; see e.g. Wehrens, 2016) and the attachments of participants to these pre-established problem definitions are ossified, offering scant opportunities for epistemological politics. In our project on regional networks, the preformulated problem definition applied by many healthcare organizations reproduced a national government’s dominant policy framing. The framing of the key problem was the shortage of medical trained personnel in older person care. In relation to this problem, healthcare organizations had already defined possible actionable solutions, amongst others, increased regional collaboration as this was supposed to offer more robust services (less dependent on the commitment of a few physicians) and enhance efficiencies (for instance by organizing pan-organizational medical services). Other predefined solutions were task reallocation initiatives and a larger number of and more interesting advanced training opportunities for medical professionals. As critical researchers, we questioned this preformulated construction of the problem and solution by asking questions: e.g. is the shortage of medically trained personnel not a question of over-medicalizing older person care? Such a drastic reassembling of the problem definition called for other solutions in our view as action researchers. In practice, however, there was little room for such epistemological politics as the attachments of actors with the established problem definition were entangled in a complex web of institutional arrangements. For instance, professional guidelines, quality criteria and laws and regulations in older person care specify the required availability and tasks of physicians in older person care facilities, such as post-mortem examinations. In practice, this means that efforts to demedicalize older person care are extremely complicated, requiring adjustments in a large number of interconnected institutional frameworks. These latching attachments have consequences for knowledge production as the insights we gained in the actionable solutions ultimately reaffirmed an institutional status quo in which older person care is provided within facilities with a dominant medical perspective on care (Schuurmans et al., 2023).
Extending Attachments
In other cases, there are opportunities for reassembling problem definitions. Extending attachments refers to the work of including new actors, new agenda’s, new problem definitions and new solutions in action research projects. This requires epistemological politics of researchers and often results into novel situated insights contesting the institutional status quo. An example of these extending attachments occurred in the project on integrating support for people with multiple problems in the Dutch social care domain. At the start of the project, the research team adopted the problem definition underpinning the policy reports of the funder (i.e. the City of Rotterdam), which framed the problem as an organizational issue of service provision fragmented by differing regulations, financing and work methods. The ‘organizational’ solution was to better align organizational interests, routines and methods in different departments to provide more integrated support. After observing many multidisciplinary meetings, however, we concluded that the main problem was not necessarily only organizational in nature (a lack of integration/coordination between different services) but more fundamental: citizens perceived social security as increasingly untethered from macro-developments. A national housing crisis that manifested itself at city level and the precarization of work and rising debt had plunged many people into crisis. The initial solution, to better coordinate services, was inadequate simply because the services were unavailable in the first place, due to scarcity (social housing) or strict regulations about access (only specific groups). As action researchers, we used these insights by reframing the problem from one of fragmented care to one of social security/precarization, foregrounding other solutions/alternatives. We engaged in epistemological politics by bringing new actors to the table. For example, actors involved in countering precarization into the multidisciplinary case discussions, including the local ombudsman, representatives of housing associations and a national program director responsible for new social housing and job opportunities for vulnerable citizens. By bringing in these actors, discussions shifted, and participants increasingly questioned the strict eligibility criteria for public services and the current policy focus on self-reliance and self-responsibility. As a result of these extending attachments, the project went beyond mere technical coordination of fragmented services and opened up a discussion about the underlying normative principles of local policy. This discussion was not easy, as we were sometimes viewed by stakeholders as academics who ‘problematized’ hands-on ideas rather than as action researchers who contributed to quick wins in the existing organizational framing, such as facilitating smooth organizational referrals between different professionals. Therefore, the strategy of extending attachments requires continuous efforts of action researchers to engage with critiques of organization actors that prefer pre-established problem definitions. As action researchers, we organized workshops in which we combined reflection on underlying assumptions of existing and alternative problem definitions, while also ensuring quick wins for organizational stakeholders that aligned with the pre-defined problem definition of fragmented service provision. This ensured rapport and commitment in the short run, while also opening op ways for more systematic reflection in the long run.
Balancing Attachments
Balancing attachments refers to efforts to do justice to different perspectives of participants in action research projects, perspectives that might foreground different problem definitions, challenges, solutions and interests. Problem diagnosis and planned interventions can be formulated on an abstract level, allowing a plethora of participants to relate and commit to them. At a certain stage, however, participants in action research must experiment with interventions in practice and make visible their effects. This can generate a new round of epistemological politics between stakeholders involved in action research; especially when interventions have broader organizational implications and when outcomes can be made visible, interpreted and valued in different ways. As Zuiderent-Jerak and Bruun Jensen (2007) argue, any research site has a surfeit of normativities and there is continuous interplay between competing ‘normative standards that are enacted, marginalized or challenged’ while research projects are unfolding (2007, p. 326). This specifically occurred in our research on nurse differentiation.
In the Netherlands, hospitals are struggling with a high turnover amongst bachelor-trained nurses. One reason behind this turnover is considered to be the fact that vocationally and bachelor-trained nurses are trained in different ways - with more focus on either technical or managerial aspects of nursing work. Yet, once graduated, vocationally and bachelor-trained nurses end up doing the same tasks. Many bachelor-trained nurses subsequently leave the profession to take on managerial roles elsewhere.
In an attempt to hold on to their bachelor-trained nurses, hospitals started to experiment with the (re)organization of nursing work. In one such experiment, a distinction was made between the roles and responsibilities of vocationally trained ‘nurses’ and bachelor-trained ‘supervising nurses’ (see also Table 1). The latter were provided with time and space to start enacting more organizationally oriented nursing roles next to their conventional nursing tasks. To learn whether, where and how to draw distinctions between the tasks of nurses and the additional tasks for supervising nurses, hospital managers had initiated in-vivo experiments on specific nursing wards. They subsequently asked us to study how nurse differentiation was taking shape and to reflect with both vocationally and bachelor-trained nurses on the implications of nurse differentiation for nursing work, professional identities and provided care.
We soon became aware that role differentiation was a contested issue among nurses on the participating wards. Some bachelor-trained nurses saw it as an important opportunity to reshape and revalue their work beyond the patient’s bed (a metaphorical way to refer to nursing tasks associated with direct patient care) and to give more weight to what they had learned during their bachelor training (e.g. in terms of leadership, project coordination, research and evidence-based practice). These nurses actively cared about the experiment and the associated objectives (implementation of role differentiation on their wards). Many vocationally trained nurses, however, thought role differentiation would devalue their position and their contribution to the nursing team. They also questioned some of the articulated differences between vocationally and bachelor-trained nurses; for instance in terms of organizational competencies. Few vocationally trained nurses therefore wanted the experiment, let alone participate in it. Yet many of them worried about its implications.
Within the broader problem definition of whether and how to organize role differentiation – and driven by a commitment to learn what role differentiation would mean for the broader nursing profession – we wanted to keep both groups of nurses on board. This was difficult because vocationally trained nurses mistrusted us from the start as they associated us with an agenda to introduce and normalize a distinction between vocationally and bachelor-trained nurses. Their implicit critique was that in our sorting of attachments, we placed too much emphasis on what role differentiation could mean for bachelor-trained nurses. Bachelor-trained nurses that cared for the experiment, in turn, worried that giving too much voice to vocationally trained nurses would endanger the experimental introduction of role differentiation in the first place. In a context in which participating stakeholders were roughly divided into two camps (those for and against the introduction of role differentiation), we struggled to balance our attachments.
To keep the different stakeholder groups involved, we organized meetings in which both vocationally and bachelor-trained nurses participated (importantly, the ward coordinators ensured that the vocationally trained nurses did). As action researchers, we were fearful of using the ‘wrong’ words, such as ‘role development’, in this conflict that could be associated with a particular sorting of attachments. Our first priority therefore was to get participants to express their thoughts and feelings. Meanwhile, we carefully weighed our words. After these first moments of articulating grievances, calming nerves, sharing perspectives and building trust, we emphasized that we were not there to legitimize preconceived notions about whether and how to differentiate between tasks, or to focus only on advantages for bachelor-trained nurses. Next, and together with all the participating nurses, we carefully reworked the experiment into an opportunity to (re)consider the ways in which nursing work was organized, valued and positioned within the broader context of hospital organizations and to make role differentiation amongst nurses part of that broader agenda.
Based on the input of both vocationally and bachelor-trained nurses, emphasis in the experiments started to shift from approaching role differentiation as a goal in itself, towards approaching it as one of the strategies through which to address challenges faced by nurses in terms of voice, workload, professional development and valuation. This framing mirrored broader post-covid societal discussions about the position of the nursing profession in healthcare decision-making. In this context, nurse differentiation became reified as a strategy to foreground tasks that were not directly associated with (bedside) nursing work, but nevertheless considered of key importance to reposition nurses in healthcare decision-making at different organizational levels (e.g. quality management, strategic decision-making etc.). The idea that more attention should be provided to these aspects of nursing work became generally accepted among both vocational and bachelor-trained nurses. In this balancing of attachments, the question how to stabilize such aspects in everyday nursing became the main object of experimentation and evaluation.
Fore- and Backgrounding Attachments
Stakeholders in action research projects are often committed to particular narratives and agendas; and this gives rise to negotiations about what can and cannot be said publicly. Researchers can relate to this epistemological politics by foregrounding/backgrounding attachments, which pinpoints to the work of (de-)emphasizing relations between actors, narratives and agendas.
The fore- and backgrounding of attachments often occurs because internal stakeholders of research projects are committed to particular narratives that align with their agenda. An example likely familiar to most action researchers in organizations is the commitment of influential stakeholders to uphold the reputation of an organization and to sanction compromising stories. At times, research findings may jeopardize this reputation; prompting negotiations about what stories become available for a public internal and/or external to the research project. In the project addressing capacity problems in older person care, we, for instance, were asked to investigate signs of heavy workloads among physicians during regionally organized evening, night and weekend shifts. We found that doctors in training mainly worked as frontliners and experienced elderly care physicians as backup. Work pressure varied greatly between the two shifts. The backup was relatively quiet, while the workload for the frontliners was heavy. Doctors in training told us stories of working 24-h shifts non-stop. There were instances of nurses being unable to reach a physician for an hour or more because the line was busy. Doctors in training seldom asked the backup for assistance because they wanted to show senior staff that they could manage. Our report on workload during evening, night and weekend shifts drew the attention of a national platform for knowledge-sharing in older person care. The report offered valuable lessons for older person care organizations wanting to experiment with regionally organized evening, night and weekend shifts. Since we were commissioned by older person care organizations collaborating in a region, however, we needed their approval to publish these practical recommendations. The managers of these older person care organizations refused permission because they felt the report might jeopardize their organizations’ reputation. Much to our frustration, we could not share our report with a larger audience. However, insights gained in this internal evaluation did eventually end up in a publication for a scientific audience. Although such epistemological politics of research participants occurs when publishing research findings in conventional social scientific inquiry, the contractual obligations in action research might mean that some knowledge productions cannot be shared publicly in the frontstage.
The fore- and background of attachments also occurs in interactions with external stakeholders. Action researchers at times need to relate to the epistemological politics of actors from the academic community, for example editors and peer reviewers. Epistemological politics occurs in discussions about what research design produces valid conceptual knowledge. Some of us have had articles rejected because of an action research design. Editors and reviewers, often from the more established and conventional disciplinary outlets, at times construe action research as practical, ‘health services’ type of research that by definition cannot produce interesting conceptual knowledge. This is extremely frustrating, because we believe that action research is rigorous and can produce cutting-edge conceptual insights. One way to sort these attachments is by backgrounding the connections to our research design, for instance by backgrounding the action elements and foregrounding the ethnographic research approach consisting of observations and interviews and a plethora of instances in which interpretations are member-checked. Although this backgrounding of attachments is understandable in individual cases where researchers want to publish their findings in well-established and highly regarded scientific outlets, it does little for the emancipation of the approach, while at the same time reproducing a vested hierarchy of epistemologies and research designs.
Discussion
Action research is becoming more common in healthcare settings, yet few studies touch upon epistemological politics in action research and how researchers can relate to this (Bromley et al., 2015; Holgersson & Melin, 2015; Mesman, 2007; Nugus et al., 2012; Øye et al., 2019). By analyzing our own experiences as action researchers in various Dutch healthcare and social care projects by utilizing the concept ‘sorting attachments’, we showed (a) various types of practical relating of researchers to individuals, organizations, institutions and agendas, while at the same time paying attention to the political, economic and cultural relations in which normativities of stakeholders are embedded, and (b) the consequences thereof for knowledge production. We described various forms of sorting attachments: latching attachments, i.e. working with pre-established problem definitions and political agendas of influential stakeholders in action research projects; extending attachments, as the bringing in of new actors and new perspectives; balancing attachments, as acting on and doing justice to contesting agendas of stakeholders; and fore- and background of attachments as the work of (de-)emphasizing connections with individuals, organizations, institutions and agendas. These different practices can help researchers in reflecting on decisions on how to navigate the often contested terrain of knowledge production in action research, as well the consequences of these informed choices for knowledge production.
Our overview of different ways of sorting attachments is not exhaustive but represents an attempt to launch a discussion on the ways in which researchers can relate to the inherent epistemological politics in action research. We, for instance, described practices of latching on and extending attachments as different ways in which researchers can relate to problem definitions in action research projects. Extending attachments resonates with what other authors have suggested as good practices in the initial stages of an action research project. Zuiderent-Jerak (2015), for instance, mentions ‘contexting’ as being attentive to the specific historical, political and economic circumstances in which particular problem definitions arise (Zuiderent-Jerak, 2015). Such critical reflection makes researchers and other stakeholder in action research projects aware that the formulated problem definition is just one of many possible readings of what the actual problem is that should be addressed in action research projects. Such critical reflection might thus open up a space for what Zuiderent-Jerak (2015) denotes as articulating ‘ontological multiplicity,’ in other words the practice of excavating alternative readings of the actual problem at hand and at the same time communicating these to stakeholders in the action research project. As our example of extending attachments shows, this contexting and articulating of ontological multiplicity is at times necessary to have real impact and address the root causes of an actual problem, instead of alleviating symptoms.
Our analyses of the balancing of attachments in an action research project on reorganizing nursing work showed, that even in situations where internal stakeholders agree on the problem definition (e.g. the issue of workforce retention among nurses) there can be different perspectives of what actionable solutions should be taken. We suggest that keeping the project structure flexible (e.g. providing space to work on various alternative solutions simultaneously), while at the same time trying to incorporate and do justice to dissonant voices is essential to uphold the emancipatory ambitions of action research projects. The latter resonates with an axiom of conventional social scientific inquiry. Researchers should be committed to ‘impartiality’ and make sure that dissonant voices are heard in research projects, even when other (often dominant) stakeholders oppose these views (Becker, 1967). While Becker (1967) suggests that researchers in these situations should ‘avoid sentimentality’ and not be afraid to upset groups by telling them stories they dislike, action research with its distinct temporality, its focus on collaboration, co-production and continuity, and the frequent points of contact between different groups, might be better aided with the idea of balancing attachments, as the delicate act researchers perform in working with and doing justice to multiple perspectives and keeping groups on board and in dialogue with one another.
We discussed the epistemological politics around formal knowledge productions through the lens of background and foreground attachments. We showed that influential stakeholders at times try to avoid unflattering research outcomes from getting out in the open. This, of course, is a well-known form of epistemological politics in action research (Bromley et al., 2015; Øye et al., 2019). In our experience, the success of such epistemological politics depends largely on statutory arrangements and contractual obligations to funders. Things get complicated when organizations that are the subject of the research also fund the projects and demand legal ownership of data. Researchers should be wary about such arrangements upfront, for example by drafting model agreement before the start of the research project that stipulate how data are collected, analyzed and shared. One can also question whether direct funding by the subjects of action research is desirable as it jeopardizes the autonomy and impartiality of the university as research institute. This taps into a wider concern about an academic institutional ecology unconducive to action research. Numerous action researchers have reported how epistemological politics at universities marginalizes action research and forces it to operate in the fringes of academia (Greenwood & Levin, 2000; Wittmayer et al., 2021). All too often, action research is not taken seriously as a research strategy producing valid and interesting scientific knowledge. As a result, action research and conventional research are for a large part institutionally separate, with distinct communities and ecologies (e.g. publishing outlets and conferences). Sadly, our practice of framing the findings of action research in terms of more accepted methodologies, for example ethnography, reinforces current distinctions between action and conventional research.
It is important to reflect on policy implications that may create an academic environment more conducive to a diversity of research methods, including action research. Various policy initiatives meant to diversify academic career paths and performance criteria at Dutch universities (see Vsnu et al., 2019) appear to create opportunities for action researchers to show colleagues from different disciplinary backgrounds and working in different research paradigms the value of their research. Rather than focusing solely on Q1 articles, H-indexes or significant results, the new performance criteria include societal engagement and the impact of conducting embedded research in the field. Societal engagement and impact do not necessarily need to be quantified but can also be shown in qualitative user narratives or demonstrated by enduring collaborative relationships between researchers and field parties. By taking these qualitative aspects into account, it becomes possible to appreciate the value of action research on its own terms. This is an important step towards building a more inclusive epistemological culture in academia that not only benefits action researchers but also other types of scholars who are engaged in participatory research practices and mode 3 knowledge production. Such steps to build a more inclusive epistemological environment not only benefits action researchers, but also society at large as action research helps finding actionable solutions to real world problems.
Footnotes
Acknowledgements
We would like to thank our colleagues at the healthcare governance department for their constructive feedback on an earlier version of this draft. We are especially grateful to Rik Wehrens for his insightful comments, and to the two anonymous reviewers of our submitted article. We also thank Cecilia Willems for her English language editing.
Ethical Considerations
The data for this paper was gathered in three different action research projects. The project on reorganizing Dutch nursing home work was approved by the ‘Centrale Commissie Mensgebonden Onderzoek’ (CCMO) (CMO, 2019-5992). The project on establishing regional collaboration networks in elderly care was approved by the ‘Medical Ethics Review Committee of the Erasmus Medical Center (METC), MEC-2019-0139. The project on improving support for vulnerable people in Dutch social care was approved by the Research Ethics Review Committee of Erasmus School of Health Policy and Management (21-020).
Consent to Participate
All patients included in these projects provided written informed consent prior to enrollment.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by funding provided by the healthcare administration offices of healthcare insurers Menzis, VGZ, De Friesland, Zilverenkruis and CZ, the City of Rotterdam, and the Dutch Ministry of Health Welfare and Sport.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The datasets generated during and/or analyzed during the current study are not publicly available due to privacy restrictions and the sensitive nature of the data, but are available from the corresponding author on reasonable request, and subject to appropriate ethical and legal considerations.