The importance of collaboration for knowledge co‐construction in ‘close‐to‐practice’ research

This commentary is offered in response to the British Educational Research Association (BERA)’s commissioned report on close-to-practice research. In conducting a rapid evidence assessment coupled with a small number of qualitative interviews, the report represents an overly dichotomised and partial approach to understanding the relationships between research and practice, and the nature of knowledge generated within such relationships. Specifically, the report fails to adequately address the central importance of collaboration to the generation of knowledge and assumes that knowledge is either academic research or practitioner enquiry, without considering a more integrated, co-constructed ‘third space’. I argue that practice-focused research should be fundamentally concerned with making an impact on practice and, therefore, effective collaboration between research and practice necessarily entails grappling with issues of power and democratisation. These are values that underpin and shape research in important ways that must be considered in conceptualisations of methodological quality. I also raise questions about the transparency and quality of decision-making in the close-to-practice BERA report, including whether the six papers identified as ‘high quality’ by the authors would meet their own definition. Their report is not definitive, but rather a catalyst for further discussion. I offer suggestions for some practical steps for how BERA could work to provide a more holistic framing for this vital field of inquiry.


Introduction
Being 'close to' something literally means 'nearly or almost', or 'from a short distance, close up' (Merriam-Webster.com Dictionary, 2020). 'Close to' presupposes a relationship that is alongside something but not actually that something. It encapsulates a gap between persons, things or ideas and implies that if we can get close to something then we can look at it, see it, study it as an object of scrutiny; like the Large Hadron Collider, we can get as close to reality as we can, but we would not be it or embody it. In short, being close to something excludes the possibility of being that thing or integrating, synthesising with it. There is a separation between that which is observed and those doing the observing.
The epistemological equivalent in research lies in the conceptualisations of knowledge transfer and exchange. The former assumes that academic knowledge is separate from practice and should be disseminated to practice in order to influence it; the latter assumes there is dialogue, a sharing between research and practice, but ultimately it is research that influences practice rather than the other way around (ESRC, n.d.). Such conceptualisations are supported or critiqued (depending on your stance) by the long-running and well-rehearsed debates around evidence-based practice in education more generally and how evidence is generated, studied and used (e.g. Hammersley, 1997Hammersley, , 2002Biesta, 2007;Hargreaves, 2007;Thomas, 2020). Nevertheless, the assumption that, at best, there is a mostly unidirectional relationship from research to practice is implied in the choice of terminology 'close-to-practice' (CtP) and exemplified in the way this term is operationalised within the British Educational Research Association (BERA)'s commissioned report on CtP research (Wyse et al., 2018(Wyse et al., , 2020. For example, the authors draw explicit parallels with health research to frame their own orientation to the topic: 'Hence, the notion of CtP research is relatively well developed in the health sector where, like education, research informs professional practice' (Wyse et al., 2020, p. 5;my emphasis).
This positioning reveals the bias of the authors of the CtP report and underpins a major omission from it: the role and nature of collaboration between research and practice. There are important critiques of borrowing assumptions, methods and approaches from other disciplines in educational research (e.g. Biesta, 2007;Thomas, 2012) and yet there is little or no critical reflection on these central arguments.
In agreement with Hordern's (2020) critique of Wyse et al. (2018Wyse et al. ( , 2020, there is an overly dichotomised view of research and practice that characterises the work, which undermines and misrepresents the breadth and depth of research in this field of inquiry.

Declaring bias in epistemological stance
We all have bias of course. That's not the issue. The issue lies in not declaring this and then presenting a definition of high-quality CtP research from a specific epistemological position that is intended to be, well, definitive for the field: Close-to-practice research is research that focusses on aspects defined by practitioners as relevant to their practice, and often involves collaborative work between practitioners and researchers. High quality in close-to-practice research requires the robust use of research design, theory and methods to address clearly defined research questions, through an iterative process of research and application that includes reflections on practice, research, and context. (Wyse et al., 2018, pp. 1-2) specific intention of addressing the persistent and much acknowledged gap between research and practice in education. My own stance is that addressing the researchpractice gap cannot be achieved by simply making research findings more easily discoverable and accessible to practice, or by maintaining an unhelpful dichotomy between research knowledge (implied to be superior) and practice-based knowledge (implied to be inferior) (Cochran-Smith and Lytle, 1999). It is not surprising that practice may feel disinclined to engage with research that others and patronises its knowledge and expertise, and that may feel like it is conducted in a different universe (cf. BERA-RSA, 2014). In my experience, we can only narrow or close the researchpractice gap if our research questions and accompanying epistemological orientation start from a different, more democratic place than much education research currently does (Parsons et al., 2020c)-that is, with practice rather than on or about it.
For me, this is about collaboration between research and practice such that knowledge (evidence) is co-constructed, by which I mean: 'new knowledge creation (the what) through the shared endeavours of research and practice working together equally (the how)' (Parsons et al., 2020b, p. 3; emphasis in original). Knowledge coconstruction is an explicitly and deliberately contrastive stance with knowledge transfer or exchange, since it recognises that there is a shared and more 'synergistic' (Leibowitz et al., 2014(Leibowitz et al., , p. 1258) space between research and practice which can offer new insights and theories for both research and practice . Such a 'third space' (Ostinelli, 2016, p. 542) enables the combination of practical (exemplary) knowledge of practitioners and families (Thomas, 2012) and the embodied knowledge of children (Parsons et al., 2020b) with more formalised, research-based knowledge, without positioning one type of knowledge as more or less valuable or important than the other. Rather, they are different but equally valuable forms of knowledge, some of which may be more tacit and all of which may be differently represented by the people working in collaboration .
At the core of this stance is an appreciation that such approaches are values-driven through explicitly grappling with the sharing of power in the design and conduct of research and, consequently, the generation and sharing of knowledge (evidence). This is by no means a new idea and I make no claims to originality or innovation in this respect (e.g. see Biesta, 2010). For example, Freire's (1970, p. 51; my emphasis) seminal work was clear that for education to be truly transformative, power needed to be shared more equally between those who have power and those who do not: Teachers and students (leadership and people), co-intent on reality, are both Subjects, not only in the task of unveiling that reality, and thereby coming to know it critically, but in the task of re-creating that knowledge. As they attain this knowledge of reality through common reflection and action, they discover themselves as its permanent re-creators.
This idea is at the very heart of participatory and inclusive approaches to research, which seek to be transformative, to make a difference to practice and authentically include the knowledge of those who are traditionally marginalised in research (Pascal & Bertram, 2012;Nind, 2017). In other words, the purpose of the educational endeavour really matters (Biesta, 2010(Biesta, , 2015Hordern, 2020), and a key purpose of CtP research is to actually make an impact on practice (Piggot-Irvine et al., 2015). Moreover, as Thomas (2012) argues, it is in the sharing of knowledge and experiences with Knowledge co-construction in 'close-to-practice' research 3 each other, and the reflection on these, that learning takes place and theories are generated for informing and transforming practice (see also Cochran-Smith & Lytle, 1999). The voices and experiences of practitioners are vital, and yet their expertise and knowledge are regularly undervalued and undermined in formalised research discourses (Guldberg, 2017). Likewise, the views and experiences of children and young people can (and should) be transformative for practice (Messiou & Ainscow, 2020), but are also regularly overlooked and underestimated Parsons et al., 2020a). Thus, the strength in doing research in co-construction with practice is more than simply a dialogue or knowledge exchange. Rather, it is about developing new insights at the intersection of research and practice that encompasses diversity in experiences, and different ways of knowing about the world (LeFrancois & Coppock, 2014), such that what is learned is more than the sum of its parts and makes an impact on practice.

Wherein the collaboration?
My own lack of originality in drawing attention to these matters is simply to make the point that this is not a marginal or fringe observation that may have easily slipped under the scholarly radar; the value and importance of collaboration between research and practice comes from a longstanding and influential pedigree (e.g. Freire, 1970;Biesta, 2007). For example, Cochran-Smith and Lytle (1999, p. 275) noted that '. . . knowledge emerges from the conjoined understandings of teachers and others committed to long-term, highly systematic observation and documentation of learners and their sense making'. Indeed, Wyse et al. (2018Wyse et al. ( , 2020 acknowledge the importance of collaboration between research and practice and include 'collaborative work' in their definition of CtP research, but then leave this important aspect underexplored. This is despite the authors citing feedback from the Research Excellence Framework (REF) 2014 Education sub-panel, which indicated that some papers from the REF '. . . were world-leading, particularly those featuring co-production or close collaboration between learners, teachers and researchers' (Wyse et al., 2018, p. 7). Given this, it would have been helpful to at least consider the nature of any collaboration, and what co-production may mean, in the evidence included in the review and to make this explicit. Indeed, it would have made sense to specifically ensure the inclusion of 'collaboration/collaborative' and 'co-production' in the search terms of the literature review, but they were absent (though 'participatory action research' did make it in). This feels like a missed opportunity to understand more fully what close collaboration could look like in high-quality CtP research.
Moreover, there were strong reasons for thinking that collaboration should have been on the BERA steering group's radar, given the BERA-RSA (2014) inquiry into the role of research in teacher education (which was cited in Wyse et al., 2018, though not considered further). The inquiry argued for the importance of collaboration between research and practice for developing a 'research-rich, self-improving' educational system: This. . . Inquiry. . . makes the case for the development, across the UK, of self-improving education systems in which teachers are research literate and have opportunities for engagement in research and enquiry. This requires that schools and colleges become research-rich environments in which to work. It also requires that teacher researchers and the wider research community work in partnership, rather than in separate and sometimes competing universes. (Executive Summary, p. 5; my emphasis) A similar point is made by Ostinelli (2016), which was also one of the six papers that Wyse et al. (2018) identified as 'high-quality' CtP research. Specifically, Ostinelli (2016, p. 534) argues that there is a '. . . need to displace the centre of gravity of ER [educational research] from academia to schools and classrooms, through an extended use of practitioner and action research' and that 'deep changes' are needed in the interactions between teachers and researchers through '. . . the expansion and creation of extended partnerships and Networks' (p. 545). The idea of 'deep change' is powerful since it implies that such partnerships are not the norm in educational research but that they should be, and is reminiscent of Freire's (1970, p. 31) call for a 'radical posture' to be taken-that is, we need to act differently in order to make meaningful changes. This point appeals to me because it chimes with my own bias of course (Parsons & Kovshoff, 2019;Parsons et al., 2020c). At least one of the interviewees in Wyse et al. (2020, p. 19) also observed that research and practice are '. . . both practice communities. . . they're different practices which intersect'. Overall, the clues were there, in plain sight, to signal the importance of collaboration, and specifically co-production, in understanding what at least some of the high-quality research in this area could look like, and yet those clues were not pursued. This is a major omission of the work from my perspective.

Wherein the quality?
The second major issue with the work relates to quality in terms of the transparency of reporting, the selectivity of sampling and the overall coherence of judgements made in the report (Wyse et al., 2018). The epistemological leanings of the report team have already been noted above and so it should not be a surprise that their definition of CtP requires 'robustness' and that this is largely interpreted in quantitative, positivist terms: Much less common was CtP research with larger sample sizes, including quantitative analyses based on statistical probability. In the context of the more general weakness of quantitative methods in social sciences, education as a discipline needs to continue to attend to this, including supporting greater use of quantitative methods in CtP research. (Wyse et al., 2020, p. 21) The problem with the uncritical conflation of robustness with positivist methodologies is well rehearsed elsewhere in the literature (e.g. Hammersley, 1997;Thomas, 2012), and Hordern (2020) also includes further discussion of this point in this special issue. Consequently, I will not re-rehearse those arguments here but merely note this limitation of the CtP report.
For brevity, I will focus on three main points relating to the quality of the report itself. First, there is no mention in the report (Wyse et al., 2018) or the paper (Wyse et al., 2020) about whether any of the interviewees were known to the research team and how specifically they were identified, beyond belonging to one of three 'types' of interviewees 'agreed to be important' by the BERA steering group for the project Knowledge co-construction in 'close-to-practice' research 5 (Wyse et al., 2020, p. 8). Slightly more information about the interviewees is provided in Wyse et al. (2020), but it is still not possible to discern their specific research backgrounds and approaches (i.e. their own biases). As I discuss elsewhere (Parsons et al., 2020b), none of the interviewees seemed to be from early years education and so this was a major gap in the sampling. Special educational needs and inclusion are also not mentioned; this is another bias of mine, and so is something I am always going to look for, but it is, nevertheless, entirely missing. More interviewees were clearly needed from across different educational sectors and research traditions to develop a more holistic understanding of CtP research.
Second, for pragmatic reasons ('time constraints'; Wyse et al., 2018, p. 2), the authors focused only on action research in their evaluation and synthesis of the evidence: partly because these approaches were well represented in the literature and also because the interviewees suggested this is how CtP research is 'generally perceived in the UK' (Summary, p. 2). This was despite noting the 'striking' finding that there is a 'wide range of methodological traditions of CtP research' (Summary, p. 2), most of which were then excluded from further consideration in the review. Given the already noted opacity and lack of breadth in the sampling of the seven interviewees, their perceptions about CtP may not provide the most authoritative view of what CtP is. Given also that the literature search missed out the terms 'collaboration' and 'coproduction', the methodology was skewed from the start. Consequently, the report represents at best a very narrow set of perspectives and approaches within the CtP research field. There is a much wider and more comprehensive search and evaluation that needs to be done.
Third, I have questions as to whether all of the six 'high-quality' CtP papers identified in the review would meet Wyse et al.'s (2018Wyse et al.'s ( , 2020 own definition of high-quality CtP research and, therefore, how internally coherent the evidence assessment was. As a reminder, Wyse et al.'s (2018, p. 2) definition of high-quality CtP research states, inter alia, that research should '. . . address clearly defined research questions'. Three of the six papers reported on original CtP research (Lamberg & Middleton, 2009;McDonnell & Curtis, 2014;Hourigan & O'Donaghue, 2015) and three were literature syntheses or position pieces about CtP research (Slavin, 2008;Piggot-Irvine et al., 2015;Ostinelli, 2016). However, while the purpose, aims or objectives of these six papers were articulated, only Hourigan and O'Donaghue (2015) included the specific research questions that they sought to address. Wyse et al. (2018Wyse et al. ( , 2020 applied the REF2014 quality criteria (reproduced in Appendix 4 of the report, 2018) to evaluate the papers identified in the search, which includes the following statement: Account will be taken of such qualities as the integrity, clarity, coherence and consistency of arguments and analysis, such as the due consideration of ethical issues.
However, despite ethics being explicitly mentioned as an important part of this criterion, only Hourigan and O'Donaghue (2015) said anything about it, and then only as two short sentences on general procedural aspects regarding consent and anonymity. There was no mention of the research being formally reviewed by an ethics committee, and no discussion of any wider ethical issues. There was also no mention of ethics review, procedures or issues in Lamberg andMiddleton (2009) or McDonnell andCurtis (2014), despite both reporting original CtP research. While formal ethics review will be less relevant for literature reviews or position pieces, we might reasonably expect there to be some consideration of broader ethical issues in action research, but these were not mentioned in Ostinelli (2016) and Slavin (2008) (indeed, Slavin does not mention action research either). There is very brief mention of ethics in Piggot-Irvine et al. (2015, p. 551), but only procedurally in terms of the need to complete an ethics application 'if required'.
Furthermore, Wyse et al. (2018, p. 2) note that: . . . the best CtP research gave a full and rigorous account of whichever methodology had been selected. . . In weaker CtP research, methodology was not explained in sufficient depth. The lack of a sufficiently rigorous account of data analysis was the methodological aspect that was frequently neglected, particularly in qualitative research.
Again, I draw attention to some inconsistencies here. For example, Ostinelli's (2016) literature review lacks many of the explicit steps that would be expected of a replicable literature search methodology including, as noted above, research questions. Moreover, very limited information is provided about how any evaluative judgements (analysis) were made. Specifically, the criteria used to judge 'methodological soundness' (p. 538) are not reported or discussed further in relation to the literature identified in the search. Piggot-Irvine et al. (2015) provided a meta-level evaluation of action research, but it is unclear how published studies were selected for inclusion in this overview. Slavin (2008, p. 5) presents an uncritically positivist account of 'synthesizing educational program evaluations' but includes no criteria for inclusion of programmes or studies and no justification for the categories identified as the 'most important issues' (p. 7). Overall, there was a lack of transparency in core methodological aspects of some of these 'high-quality' CtP papers that we might typically expect to see.

Next steps?
To move forward in this debate, the gaps in the research highlighted above clearly need to be addressed: (1) a greater consideration of the range of methodologies that were excluded from the review, including a specific targeting of research involving collaboration and co-production; and (2) hearing the perspectives of a wider range of interviewees across educational phases, and ensuring their methodological proclivities and practice-based backgrounds are made explicit. Nomenclature is also important, and I am not sure this issue is settled yet. I've raised my own concerns at the start of this commentary about the limitations of the term 'close-to-practice', and certainly others have used different terms, for example Ostinelli (2016, p. 543) discusses 'practitioner centred research', while the REF2014 panel used 'classroom enquiry' and 'practice-focused' (REF, 2015, p. 109). Pascal and Bertram (2012, p. 480) refer to 'praxeological research', which I have also applied in thinking about my own work (Parsons et al., 2020b), not least because it includes the values that underpin the research. Terminology matters, and it is something that should be revisited in any future discussions about this topic.
Relatedly, the purpose of the research also really matters (see also Hordern, 2020). From my perspective, the specific intent to demonstrate an impact on practice should Knowledge co-construction in 'close-to-practice' research 7 be central to the endeavour (Parsons et al., 2020b), and I am not alone in suggesting this: Piggot-Irvine et al. (2015) and Ostinelli (2016) both make the same point in their review papers, cited by Wyse et al. (2018Wyse et al. ( , 2020. This means more than only reflecting on practice and the research, as the CtP definition states (see above), and moving beyond this to demonstrate what has changed as a result of doing the research. How and what knowledge is generated, with whom and with what effects are key to really understanding quality in CtP research and, in my view, these are elements that should be included within any definition of research in this area.

Conclusions
The linked publications on CtP research from Wyse et al. (2018Wyse et al. ( , 2020 are, I suggest, pieces of a much larger jigsaw for understanding and defining what CtP research is, how it can be conducted and with what outcomes. There are many other pieces to add in order to populate the whole picture. This special issue is a welcome part of the process for developing other pieces that provide pause for thought through raising valid questions about bias, epistemological assumptions and quality. The Wyse et al. (2018) report and Wyse et al. (2020) paper are useful and interesting contributions to the debate, but they are not definitive. They used criteria that excluded a range of well-established approaches to doing CtP research, and so the results are narrowly defined and come from a particular epistemological perspective. BERA needs to look more widely to provide a more informed position on this field, and some of its own previous work has important contributions to make in this regard (e.g. BERA-SRA, 2014). It is an important topic that deserves more than a rapid and partial review.

Ethical Guidelines
Ethics approval was not required.

Conflict of Interest
There is no conflict of interest.

Data Availability Statement
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.