Introduction
Chapter 4 presented case-by-case accounts of each of the embedded initiatives studied in depth in workstream 3. The four case study sites (Bridgetown, Coxheath, Porter and Evansville) were introduced, as were their embedded researchers (see Table 10). This chapter presents a more thorough cross-case comparison of these four initiatives, informed by the analytic discoveries of workstreams 1 and 2. Specifically, the analysis is guided by the idea that embedded initiatives can vary across three main dimensions: their intent, their structures and their processes (see Chapter 3).
Methods and analysis
The methods of data-gathering in the four case studies were set out in Chapter 4 (and are elaborated on further in Appendix 3). The data analysis for the cross-case examination outlined here took an interpretative and abductive approach to explore the themes and dynamics in the cases, using the structure developed from the literature review and scoping analysis laid out in Chapter 3. An abductive approach seeks to combine inductive (grounded) data analysis with elements of more deductive (proposition-driven) analysis, with the goal of fostering close empirical–theoretical dialogue and, ultimately, proposition (theory) refinement.174 Following Corbin and Strauss176 and Clarke et al.,177 standard coding processes are used to generate empirically induced codes and concepts, which are then systematically related back to the concepts and theories derived from the literature, to confirm, clarify and challenge these thematic concepts as a basis for empirically informed theoretical elaboration.
Presentation of findings
In presenting the cross-case analysis, we use the 10 subthemes identified in Chapter 3 that encapsulate the domains of attention required in designing and describing embedded research initiatives.
These subthemes are arranged under the three headings of ‘intent’, ‘structures’ and ‘processes’, and these overarching categories provide the ordering for the findings presented here, with subthemes as follows:
- intent – intended outcomes, power dynamics
- structures – scale, involvement, proximity, belonging
- processes – functional activities, skills and expertise, relational roles, learning mechanisms.
The names and basic contours of the four cases were laid out in Chapter 4 (see Table 7). The individual embedded researchers are also described in Chapter 4, and cross-referencing to Table 10 and the full list of interviewees, including other stakeholders, in Appendix 2 will be necessary to make sense of the findings presented here. Fuller case accounts, with extensive data, are presented in Report Supplementary Material 3.
Findings 1: cross-case variations in intent
Intended outcomes
Analysing and comparing the case studies in terms of their intended outcomes demonstrated considerable variation between initiatives, and showed that the intended purpose and outcomes of a given initiative are likely to change over time. In all cases, there were important, but often subtle, changes in emphasis between producing knowledge, building capacity and enhancing reputation, which reflected broader transitions in the strategic priorities of NHS organisations and the evolving expertise of the embedded researchers.
Different partner organisations (NHS and university) and key actors (the embedded researcher, line managers, front-line staff) could hold divergent expectations about the purpose and intent of the initiatives. This was a potential source of tension as the researchers tried to meet the expected and preferred outcomes of various parties.
A strong example of this was when embedded researchers faced competing intentions from academic and clinical practice, in terms of which type of outputs were valued: research publications for university assessment or service-facing development outputs. In cases where the researcher had two sets of competing expectations and outputs, they sometimes found creative means to meet multiple requirements at the same event, for example raising the clinical trust profile while presenting to a wholly academic audience:
For senior NHS managers, the over-riding intention, in many cases, was capacity-building, specifically capacity-building that influenced patient care directly and delivered ‘real cultural change’. The researchers themselves shared similar intentions and were motivated by the potential to grow a research culture, change perceptions of research and ultimately improve staff and patient experience.
In this sense, there was sometimes a sense of superiority associated with the intentions of the embedded researcher, who had different expectations of their work to that of more ‘traditional’ academics:
I don’t want to be a full-time ivory tower academic [who] does research that’s philosophical or abstract or metaphysical or any of those kind of things. I wanted to do research that had some applied meaning, that would mean something to children and families.
The evolving and sometimes competing intentions and motivations of initiatives could put pressure on the embedded researcher, with practical implications for workload and their personal capacity. Anna, the embedded researcher at Bridgetown, was aware of this from the outset, particularly in relation to the work associated with research capacity-building (a strong intention underpinning her NHS support network). She had learned from her predecessor, Jill, that, to influence large-scale change, her approach would need to adapt:
One of the things I learned from Jill’s experience of having done a similar role before, running grant-raising groups . . . [you develop] one or two people who go on to have great careers, but they have great careers from an academic standpoint so they, they move over, they cross that boundary . . . So it doesn’t break down the boundary, it just helps one or two people cross over, as opposed to changing the system and how it works and dismantling the boundary or creating different boundaries.
Addressing and balancing multiple intended outcomes was central to these embedded research initiatives. Moreover, what was intended was not always explicit, fixed or even realistic, and the embedded researchers needed to pay close attention to the evolving discourses among partners and to be prepared to flex as appropriate.
Power dynamics
We observed how embedded researchers occupied an arguably unique space between clinical practice and academia. This often meant that the researcher must work across, and sometimes against, existing ‘silos’.
One of many implications of this was the distribution, and sometimes redistribution, of power and control, both for themselves and others. In some examples, such as those noted in the previous section, this redistribution of power and control was intentional; one aim of capacity-building was seen to be for front-line staff to gain more power and control to bring about change in services. There was a general sense that this change in dynamic was needed, although its far-reaching implications were possibly not always fully understood by those endorsing it, and, despite this widespread drive to empower front-line staff, researchers felt that organisational structures were not always ready for this change.
Jane in Coxheath, an experienced embedded researcher, offered her insights into ‘speaking truth to power’ – in this instance, that meaningful culture change (and power shifts) were unobtainable, being restricted by leadership styles and values:
‘Speaking truth to power’, in some form, was an established part of the embedded researchers’ role, and the respondents all shared their varied experiences of being a ‘critical friend’. This required a strong interpersonal skill set, contextual understanding and sometimes diplomacy (see Skills and expertise). It also required the embedded researcher to maintain professional objectivity and avoid ‘capture’ by the settings in which they were embedded.
For example, projects undertaken by the Evansville team regularly involved feeding back findings to senior NHS stakeholders. The tensions involved here were intensified for the researchers when findings were perceived as unexpected and sometimes uncomfortable. In these instances, the researchers shared how crucial it was to have senior support from outside the clinical setting, namely a senior academic to whom they could ‘diffuse’ responsibility when situations became difficult to manage:
I think that’s one of the big things around being an embedded researcher . . . all those relationships, but you need to constantly remind yourself that, you know, you are there to do a particular job. You are not like any other staff member within that organisation, you’re there as a researcher, and there is that kind of academic layer to the work that you’re doing, where you need to maintain that critical point really.
Of course, power was not always addressed directly or raised overtly in the interviews and conversations during site observations. Nevertheless, issues of power – implicit, unspoken, frequently unacknowledged – ran through the data and are touched on in subsequent sections, especially those concerning proximity, belonging and relational roles. In that sense, power, like ideas of identity and boundaries (straddling, crossing, breaching or maintaining), suffused and cut across many of the meaning categories created in the review and scoping work of Chapter 3.
Findings 2: cross-case variations in structures
Scale
The scale of the initiatives ranged from a single embedded researcher to an embedded team of up to five. Some schemes were bounded and time limited; others were more open, emergent and evolving. There were differences in terms of configuration too; for example, the Coxheath initiative was primarily configured and constructed around Jane as an individual, whereas Evansville was configured around an evolving portfolio of projects. Porter was different again, with an overarching strategic priority and focus to the configuration aimed at building research capacity and capability. The descriptive accounts in Chapter 4 outlined some of these differences in more depth.
Despite these differences, there was a sense that, in their timescales and overall life cycles, initiatives had identifiable stages of development: initially an incubation period, followed by formalisation and then a developmental stage, and finally (for fixed-term projects) a conclusion.
In the incubation period, relating to a specific individual, group or intention/idea, the initiative would begin to grow and the foundations would be laid, including the building of important relationships, ‘buy-in’ and extensive discussions over many months or even years. In Bridgetown, for example, this growth started under an earlier role-holder, Jill, developing the intentions of the role and identifying the key academic and practice stakeholders.
The initiative would then be formalised to a certain degree with role specification(s) and appointment(s) to address a given issue or project. This would include many practical considerations and arrangements, surrounding contracts and lines of reporting. At Porter, this phase in Victoria’s role offered helpful insight into the complexities of funding and contractual arrangements, in line with various expectations and intentions of the role:
I was in post for a very long time before I started to work out exactly what I was supposed to be doing, which was . . . kind of typical of the way these structures work; and to some extent, I was involved in writing my own job description . . . I’m employed by one of the directorates, not by the trust per se . . . so, I’m not actually affiliated with the R&D department . . . That made it even more complicated. [It’s] one contract with the university, that the trust buys out for 2 days a week, so it was set up as a university post, with trust partnership. So, the trust pay the university my salary for 2 days a week.
A developmental period would then follow. During this, the embedded researcher would demonstrate their potential and the role would grow; often the details of the role (particularly in terms of the nature of the work) would change and evolve in line with various expectations and demands. In many instances, the role undertaken would be different from that proposed and the overall workload larger than anticipated. Karla’s main duties in Porter related to service evaluation, but, as her role grew, she also became heavily involved in the trust’s quality improvement strategy: extra work that was not part of her original role.
In some instances, the initiative would then come to an end, as in the case of Evansville, where funding stopped at the end of its fixed timescale. Even so, there could be a degree of influence ‘left behind’ by the completed projects, in terms of increased research awareness, cultural change and service improvement. Bella, formerly at Evansville, said she was still contacted by members of the NHS trust for advice and support.
Many initiatives however – like three of the four studied here – would be funded indefinitely, with the embedded researchers on permanent contracts. As a result, the timescale of individual projects and activities would become important, rather than the duration of the overall initiative. Maintaining researcher objectivity over such longer timescales might become a concern. Scale and longevity, then, would sometimes be planned from inception, and, at other times, would emerge contingent on local circumstances.
Involvement
The teams surrounding the initiatives tended to be relatively small, compromising largely line managers and senior academic support, before branching outward to team members. Each initiative and embedded researcher had a distinct configuration or network of relationships. The partner identification tool used in the interviews helped us to identify those involved with the initiative (rather than the projects/work undertaken by the researcher in this instance), and how intrinsic their involvement was. The schematics drawn from use of that tool are displayed in Figure 5.
Each researcher identified two or three key contacts, often those to whom they reported directly or indirectly and who they felt had fundamental involvement with their work. These individuals often held senior positions in either the academic institution or the NHS, and the embedded researcher often described being ‘accountable’ to these individuals. In line with this seniority, such key contacts usually provided the embedded researcher with a degree of protection and a voice at a senior or strategic level. They were often important in ensuring funding and also access to other areas of the organisation, setting the tone from a senior level. There was universal agreement that, without these individuals and their ‘buy-in’, the embedded roles would be unsustainable.
Moving out from the centre of the partner identification tool, the individuals and groups came from a variety of sectors. Unsurprisingly, other members of the NHS team, often clinical specialties and/or professional group members (managers and department leads), were seen as important. In addition, relationships with members of the trust R&D teams were quite often seen as necessary. Again, these individuals brought access to front-line staff and their knowledge or expertise around each specific clinical context.
Academic colleagues were also identified here, showing collaborative relationships. The degree of importance placed on these formal and informal academic connections differed depending on each role. In some cases, senior academics (at dean or professorial level) were responsible for line management, but many of the academic relationships were with peers and brought shared research interests, support and a critical friend for the embedded researcher. In most instances, these relationships had been built over many years and their involvement and inclusion were of mutual benefit. Karla (Porter) offered her experience of the involvement she shared with colleague Victoria (also embedded at Porter):
So another person that I . . . have quite a bit of contact with, but I don’t know, I don’t know how to put her in terms of influence, is [Victoria] . . . Talking to her, she spans across the trust and [place]. So yeah, I suppose she’s, she’s very much like a partner, partner in crime, I guess.
So, between us we run the quality improvement group for the directorate . . . She was brought in as the strategic lead to basically get research on people’s agendas. So she is, in terms of the pecking order, very high up within the organisation, and it was her job to integrate research into people’s day jobs . . .
I’d say [she provides] peer support to be honest, because we meet every couple of months, and we just chew the fat about life in the different organisations and things . . . Because [Victoria] is still quite active in actually doing research, doing evaluation. So, it’s kind of peer support, that she’s done a PhD, I’ve done a PhD, and there aren’t many of us, what are we doing with that now? So, with that one it is more peer support, I guess, and thinking about and talking about the development of the quality improvement agenda.
Other sectors involved with the embedded initiatives included non-profit organisations and wider collaborators including national and international partners (e.g. NHS England and NHS Improvement). In these instances, the embedded initiative might provide consultancy or informal input to other organisations or agencies. Students were also frequently included in the identification tool, their involvement relating to shared interests, capacity- and capability-building, and mentorship.
It is also important to note the involvement of patients. Although patients were noted infrequently during the completion of the partner identification tools, the motivation to improve their experience and the care they received was reiterated often by the embedded researchers and those supporting the initiatives. In terms of formal PPI, the researchers shared diverse experiences, largely related to the nature of their role and the projects in which they were involved.
Generally speaking, PPI was undertaken for specific projects for which it was perceived as relevant, and was accessed through existing PPI groups; this included internal, trust-level service user networks, charity PPI groups (relating to specialism or illness) and national PPI representation through schemes such as NIHR INVOLVE. Other experiences relating to PPI were shared through research capacity-building work; in these examples, the embedded researchers were encouraging and supporting others to employ PPI effectively. Evansville was the only case study to set up a designated PPI panel in relation to the initiative and the role of the embedded researchers more broadly. Despite this variability and the sometimes less tangible examples of patient involvement in initiatives, their inclusion (directly and indirectly) was viewed as fundamental by the embedded researchers.
The networks of involvement then were very specific and bespoke to each initiative and often to each individual embedded researcher. Although some connections were tight (e.g. to line managers and scheme champions), others were loose and informal, forming and reforming as needs were identified and projects developed. In most cases, PPI was limited to projects rather than being linked to the overarching strategic goals of the embedded initiative. In few cases did it seem that the connections described had been fully constructed or specified at scheme inception; instead, they emerged contingently as the schemes evolved.
Proximity
Each initiative varied in terms of how its relationships and activities afforded contact and proximity with different groups. In more obvious ways, this related to where embedded researchers and research activities were physically located. Researchers at Porter were generally located within the NHS organisation, with allocated desk space, whereas researchers at Coxheath, Bridgetown and Evansville had multiple desks (as did Victoria, the only researcher at Porter with a dual contract). Irrespective of desk space, there was a sense that work location was changeable and flexible:
I have to live my life with a pair of trainers and a rucksack, because I’m never in one building long enough to just be able to settle, I have to, I run round [location] – you know, I walk up to [location A] . . . and [location B] site is right over the other side of the city and I walk there; and the university campus is all spread out over here, and in 1 day I can be in four or five different buildings.
Much of the movement between locations was done in a conscious attempt to be more visible. There was a shared sense among researchers that physical visibility, particularly within the NHS organisation, was crucial to the success of their work. Visibility was seen as intrinsic to informal relationship-building, and generated opportunities to reduce the gap between practice and academia. Karla, at Porter, said she spent most of her time at one main location, ‘the main base [because] I like to be in the thick of things’, and went on to link this to perceptions of ‘embeddedness’:
Through writing these narrative reports I’ve started to get a really, really good understanding of the service . . . I’m visible as well, people have got to know me so people will then contact me and say, ‘oh’, you know, ‘you can you help me with this’. And actually, if I was just a name, people wouldn’t necessarily feel confident to do that . . . I think part of the embeddedness as well is actually like physical proximity . . . their hub is sat sort of probably where that couple are over there [about 5 m away], they’re that close to where I am, so quite often I’ll sort of go over and introduce myself, talk to them, work with them, so it’s sort of getting to know them as well.
Embedded researchers in these case studies generally prized close proximity to the organisations in which they were embedded, but often had to work with great flexibility and agility across multiple sites and settings. Being physically co-located and ‘being seen’ were literal manifestations of the metaphorical need to close gaps between research and practice and required considerable effort.
Belonging
All the initiatives brought to the fore the importance of social, cultural and knowledge boundaries, and the degrees to which actors shared particular ways of thinking, values and norms, and a sense of belonging by virtue of their professional socialisation or organisational affiliation.
Much of the hidden or unseen work of the embedded researcher involved working across and mediating the tensions between these boundaries. At the same time, however, this created certain role ambiguities for the researchers themselves. As discussed previously, the original intention of the role could evolve, resulting in researchers taking on different areas of work. In Porter, Karla’s contracted work related to narrative report writing for local service evaluation, but her role had since grown unexpectedly and she was now also heavily involved in setting up a local, trust-wide quality improvement initiative. This ambiguity of role was experienced across all of the case studies and, in some instances, had significant implications for the number of hours worked, with few boundaries in place to limit this:
Technically [I] devote 16 hours [per week] to [the trust], but the reality is I do a bit at home on top, because you can’t squeeze it into 16 hours, that’s a nonsense . . . The other 24 hours then are university hours, but again, it’s impossible to fit my university job into 24 hours . . . I do an awful lot of the evening work, an awful lot of weekend work, a lot of juggling between posts, and a lot, wherever possible, [of] double-weighting, to try and make up some of the shortfall, because I work about 80 hours a week.
The problem is that both organisations [need you] to do the job properly, and I can’t bear not to, [and I] haven’t got enough time to do either job properly unless I put in extra hours.
All of the embedded researchers therefore struggled with role and identity ambiguity relating to their sense of professional self, that is whether they were clinicians or academics (or both), whether they were aligned with the needs of the university or NHS (or both), and how they cultivated relationships with front-line staff or those facing the real-world need for research and evidence.
Each embedded researcher found their own paths to navigate these tensions of belonging, and appeared to find creative ways of resolving, or at least reconciling, multiple competing roles, in turn starting to forge new or distinct identities and belonging. They often found support, and a sense of that belonging, in informal relationships: they spoke at length about these relationships, which were often with other people in hybrid or similarly uncommon roles. These relationships, and even friendships, offered a sense of grounding, and opportunities to work through common problems or difficult experiences. Notably, the workshops and formation of the ‘embedded network’ that formed part of this project (see Chapters 1 and 6) seemed to demonstrate the emergence of a distinct occupational identity or sense of belonging around the embedded researcher role itself; for example, the researchers of Bridgetown and Coxheath came to see that they had much in common.
In summary, then, belonging across multiple sites, roles and identities posed specific challenges and additional work burdens. These were often actively managed, more informally than formally, but still left considerable unaddressed challenges and needs.
Findings 3: cross-case variations in processes
Functional activities
There were substantial variations in the nature of the work undertaken by the embedded researchers and in the scale of that work. The purpose of those activities also differed to some degree. Activities and projects included service evaluation, local and regional quality improvement, and small-scale research projects (often as part of research capacity- and capability-building). In those instances, much of the work was undertaken with a smaller, more local team. In some cases, the embedded researcher might even work one to one with an individual to support or guide a specific project.
Timescale similarly depended on the nature of the work. The heavily weighted service evaluation roles were structured around report deadlines (often annual and with accompanying interim reports). Other work, such as that relating to quality improvement, might be more flexible in its time frame and be part of larger, ongoing agendas or change programmes that took place over many years.
Researchers spoke little of formal training associated with their activities. In Porter, Bev and Katrina’s activities related largely to service evaluation, although they were also involved in recruiting for large-scale clinical trials (for which their service was often a patient recruitment site):
So, we’ll do an annual report on inpatient outcomes, outpatient outcomes, which we then share with the team, look at our outcomes, see what we need to be doing better, if we can do anything differently. So, we use that in conjunction with patient experience questionnaires . . . We’ll look at [them] and say ‘is there any way that we can improve on that?’.
Our bread and butter is service evaluation, so we’re getting outcome data for our service – that’s, that is what we have to do on a daily basis, but in conjunction with that, we do have other projects going on.
For insight into the sheer range of activities undertaken by the embedded researchers, we can explore the work of Jane (Coxheath) and, to varying degrees, that of Anna (Bridgetown), Bella (Evansville) and Rachel and Victoria (both at Porter). To differing degrees, all five were involved in leading more ‘traditional’, original research, including funding applications, conference presentations and publications. Jane offered the most powerful example, with a helpful contextual insight into her activities and workload. Her work often related to relatively large, regional and national research projects, many of which she had secured funding for and that had been ongoing for months or years:
So, the work in the university, there’s two projects, but they’re not always primary research and they’ve been building on, or implementing or impact-based research . . . I’m now leading the work for Health Education England [relating to the] workforce transformation type agenda . . . that’s by July. The second piece of work is for [organisation name], so they want us to help them to develop a . . . career development framework for their mainly nursing staff at this stage, for health-care support workers across hospices . . . in the trust . . .
Two big things I’m working on, one is a single capability framework across the whole system . . . and then I’ve got three priorities, which is emergency department, general ward and children’s to apply that to, so that’s really important and hopefully will link to an IT [information technology] set-up that can be used . . . by Health Education England.
Then the other thing is . . . [a] collaborative research study across the trust to . . . and then I’ve got the ongoing stuff with the clinical leadership programme and the support of the consultant practitioners . . . There’s a conference that we’ve got in June . . .
In short, the portfolios of functional activities of these embedded researchers were often sprawling, diverse and sometimes unmanageable, with variable (and often negligible) amounts of formal training. The activities frequently reflected a large span of scale: from local improvement projects to regional or even national initiatives; from the local, informal and unfunded, to generic knowledge generation from funded formal research.
Skills and expertise
As with the portfolio of functional activities, the skills and expertise needed by embedded researchers were many and diverse. The range described tended, in part, to reflect career backgrounds, local initiative specifications and evolving relationships with wider stakeholders.
However, common skill sets were identified, not only by the embedded researchers, but by members of their wider networks. Notably, there was considerable agreement across the varying levels of seniority; for example, managers largely noted the same necessary skills identified by junior members of the team, such as doctoral students. These skill sets could be summarised as follows:
- possessing expertise and in-depth research knowledge
- making research and research processes accessible
- forming complex networks, connections and collaborations
- building strong relationships and trust
- harnessing the knowledge and potential in others
- working across hierarchies
- being committed and driven, with great stamina.
Unsurprisingly, research expertise and in-depth research knowledge were seen as central to the embedded researcher role. This assumption was shared across almost all participants in the case studies. Embedded researchers themselves naturally saw this as important, and central to their ability to adapt to new areas of research, specialties and topics. Crucially, their research knowledge gave credibility to their role:
Making research and research processes accessible was also seen as fundamental to the embedded role – particularly when working to change perceptions of research and subsequently build research capacity and capability in the NHS organisation. Research was often seen as ‘out of reach’ for many clinicians and managers, and, in some instances, was deemed irrelevant. The embedded researchers’ ability to change these perceptions was seen as a necessary strength:
The ability to form complex networks, connections and collaborations was at the heart of the embedded role. Arguably, this work sometimes received too little weight or acknowledgement relative to its importance (and relative to the more concrete and more obvious research skills). The term ‘bridging’ was used, as was ‘conduit’; however, these suggest the forming of a connection between two parties: the reality is that many of the embedded researchers were at the centre of complex webs of relationships and networks:
I can connect those people because I know what’s going on . . . [before] there just hasn’t been the forum to bring people together . . . to make those connections for people . . . I kind of feel like I’m a bit of a matchmaker essentially, because I’m kind of going round getting all this information, putting it in, and then kind of regurgitating it out in a sort of useful manner.
The ability to build strong relationships and trust was key to maintaining these complex networks, connections and collaborations. There was a strong sense throughout the data that the embedded researchers were seen (and certainly aspired to be seen) as trustworthy and genuine, with no hidden agenda:
People trust her, they’re not wary of her, they believe that she’s going to be there tomorrow and have that feeling about, you know, that solidarity. And I think I’ve got the ability to give people that feeling because I’m an experienced researcher and a nurse, erm, but [embedded researcher] has it . . . So, I think there’s a level of trust . . .
You know, there are barriers that she doesn’t have to go over because she naturally lives with those barriers. Are you with what I’m saying? It’s like she’s already inside the pen, she lives inside the pen.
The skill to harness the knowledge and potential in others linked to some of the preceding skills (e.g. building relationships and making research accessible). Drawing out the potential in others and challenging their thinking (with care) were skills that embedded researchers displayed often. This related to the empowerment of others, not just those in junior positions, but also those working in senior management and leadership roles:
So, in establishing all of my ideas, my thinking, I went and sat with [embedded researcher] and we looked at the ideas, and what I was saying, and we, we had a conversation and she was able to support that conversation, that then helped me . . . articulate and reflect on my vision and then develop my thinking, such that I was able to make the pitch for a Darzi Fellow[ship].
As the quotation above demonstrates, working across hierarchies was an additional interpersonal skill strongly valued by the embedded researchers and by their wider teams. The ability to work with a variety of individuals, with differing roles and seniority, was important for success. Within this, there was the necessary ability to adapt their approach and communication to differing levels and audiences, so that the embedded researcher was seen as credible, relatable and personable:
With [embedded researcher] they know that she’s there, she was part of them, she has that ownership, that belonging. She recognises their significance; she has that continuity that’s really important . . . and before the post she’s in now really because she was embedded even when she worked for [university] – she was embedded in there. So, it’s been over time, exactly, that she’s known now, and she’s understood . . . And she can talk to people at a whole range of levels and she has that ability as well.
What has run across the range of skills and expertise noted so far has been the need for embedded researchers to be committed and driven, with great stamina. These characteristics were seen as crucial, largely as a result of the extensive challenges associated with the embedded role (including role creation, development, workload and competing expectations, alongside the cultural and strategic barriers). There was a sense that the embedded researchers needed to be strong, resilient, adaptable and committed to their role. This was voiced by many of the embedded researchers, especially Anna, Jane, Victoria and Karen:
It’s very time-consuming, and having anything that’s a split sort of role it is exhausting. It’s always more than the sum of its parts really, so it is very dependent on someone who has the energy and the drive and the vision to see something – that’s, you know, that’s not so much the norm.
Overall, there was broad agreement about the wide and diverse skills needed, but the degree to which each skill was important varied according to how the initiative was configured, particularly when looking at strategic involvement and the need for methodological expertise. For example, those roles that included a high proportion of service evaluation required fewer formal methods training and strategic interpersonal skills – these skills were much more relevant for initiatives that required ‘speaking truth to power’ and working across strategic organisational barriers at senior levels. The soft skills of effective communication and relationship-building, however, were always in demand.
Relational roles
The themes described in Findings 3: cross-case variations in processes show the relevance to embedded initiatives of effective communication and relationship-building. This was seen as a fundamental quality and skill of the researcher: the ability to build and sustain relationships, work across hierarchies (by being relatable and adaptable), harness skill and potential in others, and build trust were all crucial and could be described as skill in creative mediation:
Jane’s very approachable; she doesn’t make you feel stupid, she listens, she’s an active listener and then she helps you assimilate it. So . . . her communication skills are brilliant . . . and obviously the knowledge is there.
Diplomacy and negotiation formed a large part of this relational work and were also necessary for managing conflict and balancing competing agendas, expectations and priorities.
The key influences on the nature of the embedded researchers’ relational roles were their professional background and, most crucially, the functional contribution expected of them. For example, three embedded researchers in Porter, Katrina, Bev and Karla, whose roles focused largely on service evaluation, spent less time negotiating agendas and experiencing conflict through different priorities. The boundaries and nature of their work were more tightly and clearly defined from the outset. By comparison, Jane in Coxheath was senior in both her past and current roles, and her work had far fewer boundaries and an ever-expanding remit. Within this work, she had to juggle an array of conflicts, perceptions and agendas. As a result, her extensive experience in creative mediation – the ability to relate to others and work across rigid professional and philosophical boundaries and hierarchies – was much more relevant to the success of her role.
Being accessible and relatable was universally seen as important. The researchers had to ensure that they were not ‘too academic’ in their approach, particularly if they were to make research ‘accessible’ (and change perceptions of research). All of the embedded researchers needed this ability, arguably because there was often a sense of unattainability attached to the term ‘research’, rather than a problem with any of the specific processes in ‘doing’ research (i.e. the term itself is intimidating to many people).
Additional relational aspects of the embedded researchers’ roles were covered in Belonging and Skills and expertise.
Learning mechanisms
As noted in Chapter 4 on the formal evaluation and performance monitoring of embedded initiatives, success was seen as largely challenging to measure, partly because of the ambiguity of roles, but also because of the complexity of the work being completed. There was, therefore, variation in terms of achievement and what was subsequently learnt about what was, or was not, achieved.
For work such as service evaluation reports and specific quality improvement projects, the outputs were relatively easy to monitor and evaluate. Similarly, there were individual research projects that had clear aims or objectives and tangible outputs on completion, such as publications. Such achievements were, by nature, more explicit and therefore easier to ‘see’, making it easier to recognise, in the first instance, and learn from them. Jane in Coxheath and Victoria in Porter, for example, published widely and often attended conferences. These achievements were easier to identify, and learning how to make these accomplishments was relatively straightforward.
Less easy to acknowledge, monitor and learn from, however, were projects around research capacity and capability growth. Often there was no baseline of existing understanding or culture. At the same time, measuring cultural change, particularly when it related to changes in perception and engagement, was, by nature, difficult. Much of the work associated with culture change is subtle and even invisible. Subsequent achievements can be difficult to identify and acknowledge, and subtle small-scale changes are likely to be missed.
Anna in Bridgetown was working to establish a baseline of research activity (as shared by a member of her management team, Gillian) to try to mitigate some of these challenges and offer a starting point from which achievement could be measured:
As a case study that had ended, Evansville showed that learning could be carried over into new initiatives through the accumulated experience of the embedded researcher (Bella). Bella said one initial goal of the embedded research role had been to disseminate findings beyond the immediate team, to ‘create a kind of a research or an evaluation mindset within the [whole] organisation’:
Many times, [people were] just looking at their daily practice, their work, they’re not considering types of evidence out there, they’re not thinking about evaluating what they’re doing, they’re not thinking about quality improvement. So, part of a deal for our team was, you know, to show them different ways in which they could do that.
It was only when the role was redesigned, however, that the team started sharing its findings widely every 2 months, ‘even . . . emerging findings’:
Our findings [now] are going to different implementation groups across the trusts, it’s not just going to three or four teams, they’re being presented at their group executive level in 2 weeks. So . . . the chief exec[utive] knows exactly what we’re doing, and they know what’s coming out of our evaluation so . . . We learned from that, I think, and now we’re really careful not just to focus on data . . . but actually to be able to use this formative design a bit more practically. So, I think that was one of the big issues we had on the team and one of the reasons why I don’t think we [i.e. the original embedded initiative] were funded beyond the 4 years.
Learning within the embedded research initiatives proved difficult then, not least because of the frequent lack of clarity over intentions and the emerging nature of many of the activities. Formative learning was clearly prioritised, alongside the necessary informal support systems for the embedded researchers. Summative learning was less in evidence, notwithstanding the varied governance and accountability mechanisms that played a role in shaping the overall work programmes and the day-to-day activities.
Concluding remarks
This chapter has dug deeper into the data from the four intensive case studies conducted in workstream 3. This cross-case analysis complements the case descriptions of Chapter 4, looking at the attributes and dynamics of these diverse schemes. Using the structured framework laid out in Chapter 3, and the account of knowledge co-production derived from a wide literature summarised in Chapter 2, we have drawn insights from the messy real-world experience of embedded research. As with Chapter 4, only interim conclusions are drawn here, with the main discussion of implications for future scheme development to be found in Chapters 6 and 8.
The analysis in this chapter was arranged under the broad domains of ‘intents’, ‘structures’ and ‘processes’, further broken down into their constituent themes as laid out in Chapter 3. Of course, there is much interplay and linkage between these domains and themes, and so boundaries between them should not be taken as firm or fixed. Nevertheless, this heuristic allows some degree of order to be placed on what otherwise would be a confusing mass of data.
In terms of intent, our in-depth study amply illustrated that there was often no singular or fixed intent to the schemes; rather, purposes evolved and flexed to reflect the priorities of key partners. Intentions were often shaped around building applicable knowledge (collaboratively, if not always co-productively) for the betterment of front-line services, often alongside the creation of new capacities and capabilities. Often discernible, if not always overt, was a desire to reshape patterns of power, especially the relations between academic and service partners, and sometimes also between different professions in the service settings. Intrinsic to this were ideas around the (re)negotiation of knowledge and the capacity of research to inform what counted as knowledge in context.
The structures created to support embedded research involved various practical configurations, but all had to pay attention to the scale of activity envisaged, the nature of the partners involved, and the importance for the researchers of presence and visibility in service settings. These important and relevant features, in turn, created tensions and dilemmas for the embedded researchers over identities and belonging. It was clear that such tensions, ambiguity and unsettlement were in play for many of the embedded researchers for much of the time. Managing these contradictions and tensions was a major preoccupation, drawing attention to the need for deft supervisory arrangements, careful mentoring and supportive networks.
The incredibly varied range of functional activities for many of these embedded researchers highlighted the complex processes that needed to be ‘designed in’ to embedded research schemes, and the extraordinary array of necessary skills and expertise needed to function effectively. The range of skills and styles we observed to be active in embedded research also suggests that these kinds of projects rarely settle on a single recognisable type of knowledge co-production, but fluctuate between (or sometimes blend) different approaches, sometimes more conventional, sometimes more committed. Nurturing and shaping the relational roles through which knowledge was brought to the fore and used mattered at least as much as, if not more than, standard research or project management skills in this context. Finally, given the often ambiguous and shifting nature of scheme goals, it was unsurprising that learning mechanisms, although much needed, often seemed underdeveloped.
Taken together then, the insights derived from these in-depth case studies suggest many areas of potential discussion, development and learning for stakeholders in embedded research, and it is to these issues that we now turn. Chapter 6 will move away from close-up examination of the dynamics of specific embedded cases to bring together the lessons from all four workstreams and introduce a design framework for the specification, management and assessment of embedded research initiatives. Sitting alongside this, a suite of tools, resources and supporting materials will then be described more fully in Chapter 7.