Simulation As A Tool For Developing Knowledge Mobilisation Strategies Innovative knowledge transfer in youth services

Knowledge mobilisation (KMb), also known as knowledge translation or knowledge exchange, is a process that shares academic research and other forms of knowing with the goal of informing service delivery, community practice and public policy (Phipps & Shapson 2009). KMb has been widely developed in the fields of health care, education, international development and climate science (Kothari & Armstrong 2011; Levin 2013), though is less well known among those who design and deliver social services. Where social service providers have engaged in KMb, excellent models have been described that address the challenges of sharing research and evidence to inform policy and practice among human service professionals (Greenhalgh et al. 2004; McLennan et al. 2006). These models, however, have tended to build on research undertaken in more structured settings like hospitals and large government agencies. One focal area for KMb that has had less attention is how human service providers share knowledge in less formal settings like community-based non-government organisations (NGOs). Here, practitioners are charged with applying knowledge to address so-called ‘wicked problems’ where no obvious solution exists and where myriad approaches to the issue may have already been tried (Weber & Khademian 2008). Structural differences in clinical and community-based settings (such as access to research findings published in peer-reviewed journals, time for professional development and the support of staff dedicated to research) make academic research and expertise less accessible to community partners (Hart et al. 2009; Kothari & Armstrong 2011; Phipps 2011). When KMb has been theorised in community-based settings (Bonnie 2010; Estey, Kmetic & Reading 2008; Smylie et al. 2004), the models have most often been adaptations of those that already exist. As Greenhalgh and Wieringa (2011) suggest, the metaphor of knowledge transfer and related concepts like KMb do not include sufficient scope to capture the range of possible activities that are involved. Given the complexity of ‘wicked problems’, iterative designs based on efforts by stakeholders to find solutions to their own KMb challenges in poorly resourced settings are needed. Gateways: International Journal of Community Research and Engagement Vol 8/No 1 (2015): 100–117 © UTSePress and the authors

translation or knowledge exchange, is a process that shares academic research and other forms of knowing with the goal of informing service delivery, community practice and public policy (Phipps & Shapson 2009). KMb has been widely developed in the fields of health care, education, international development and climate science (Kothari & Armstrong 2011;Levin 2013), though is less well known among those who design and deliver social services. Where social service providers have engaged in KMb, excellent models have been described that address the challenges of sharing research and evidence to inform policy and practice among human service professionals (Greenhalgh et al. 2004;McLennan et al. 2006). These models, however, have tended to build on research undertaken in more structured settings like hospitals and large government agencies.
One focal area for KMb that has had less attention is how human service providers share knowledge in less formal settings like community-based non-government organisations (NGOs).
Here, practitioners are charged with applying knowledge to address so-called 'wicked problems' where no obvious solution exists and where myriad approaches to the issue may have already been tried (Weber & Khademian 2008). Structural differences in clinical and community-based settings (such as access to research findings published in peer-reviewed journals, time for professional development and the support of staff dedicated to research) make academic research and expertise less accessible to community partners (Hart et al. 2009;Kothari & Armstrong 2011;Phipps 2011). When KMb has been theorised in community-based settings (Bonnie 2010;Estey, Kmetic & Reading 2008;Smylie et al. 2004), the models have most often been adaptations of those that already exist. As Greenhalgh and Wieringa (2011) suggest, the metaphor of knowledge transfer and related concepts like KMb do not include sufficient scope to capture the range of possible activities that are involved. Given the complexity of 'wicked problems', iterative designs based on efforts by stakeholders to find solutions to their own KMb challenges in poorly resourced settings are needed.
In this article, we build on our experience brokering relationships between research and community-based practice and report on an innovative approach to KMb model development.
This was a knowledge mobilisation simulation designed to engage service providers concerned with the mental health needs of children and youth who have been exposed to chronic or acute adversity, a population we have called children and youth in challenging contexts (CYCC). For three days, 65 policy-makers, senior staff of NGOs, mental health professionals, KMb specialists, and youth participated in a series of interactive exercises to answer the following questions: 1 What are the barriers to KMb in less formal service settings and in settings primarily concerned with services for CYCC?
2 What KMb strategies are already working in these settings?
3 Building on the answers to the first two questions, how do we move knowledge of effective practices between service providers?
Our focus was on sharing both practice-based evidence and evidence-based practice. Practice-based evidence is what is reflected in reports of 'what works' which are shared between practitioners.
Typically, we find practice-based evidence reported at professional workshops and in the grey literature, published sources that appear online or in print but have not gone through a rigorous peer-review process. Practice-based evidence tends to reflect the experience of those delivering programs and to provide anecdotal evidence of program effectiveness. In contrast, evidence-based practice meets the criteria for rigorous evaluation of outcomes, is often published in peer-reviewed journals and meets the criteria for replication, meaning that studies can be repeated with a reasonable expectation that, if there is fidelity to the program design (program providers deliver the program as it was intended), similar results would be expected.
Our collective interest as participants at the simulation was in working with young people in contexts where resources may be poor and the challenges confronting them very complex.
In such contexts, there may be far fewer examples of evidencebased practices as the evidence is time consuming and expensive to produce (Mitchell 2011). Furthermore, much of this work is performed by non-government organisations which rely on practice-based evidence, either through developing their own program solutions for vulnerable child populations or borrowing program elements from others who have reported success. In these less formal contexts where practice-based evidence is more commonly employed as the basis for decision-making, both a lack of resources and social complexity create daily hassles when identifying and delivering effective services and complicate the meta-challenge of figuring out how to mobilise knowledge across service providers. These challenging contexts include remote or culturally and socially marginalised communities, situations of violence or poverty, exposure to extreme forms of stigma, and those where family or community systems present significant levels of adversity that threaten the healthy development of CYCC.
The adversity experienced by CYCC requires service providers to adapt programming to be contextually and culturally responsive (Mitchell 2011).
The heterogeneity of CYCC and the programming they need further complicates KMb, especially when services are delivered in community-based programs. In contexts like these, August, Gewirtz and Realmuto (2010) argue that most often youth services approach program innovation by identifying core elements of a preferred intervention while adapting program elements to local conditions in ways that will not jeopardise fidelity to the principles of intervention. As an alternative, they propose a different strategy for program innovation: an adaptive model of program development that engages clients in the process of decisionmaking. The clients help to decide the needs, dosage and sequence of care based on individual needs. Clients have the support to make deliberate and effective choices about their services. As August, Gewirtz and Realmuto (2010)  KMb, then, for youth services that meet the needs of CYCC may be most useful when it emphasises program adaptation and contextualisation (the second approach). Attention to the practice context has been demonstrated to be critical to effective implementation of health (Jacobsen, Butterill & Goering 2003) and education (Levin 2013) interventions. Without standardisation, however, the exchange of best practices becomes especially complex.
It was in this context of under-resourced services, challenging contexts and the need to adapt programs that we undertook an innovative approach to identify KMb strategies appropriate for youth service providers and other stakeholders. We were looking for ways of addressing the nature of the evidence for effective programming and of facilitating its adaptation in less formal service contexts. Our work was guided by our knowledge of KMb models, notably the Promoting Action on Research Implementation in Health Sciences (PARIHS) model (Kitson, Harvey & McCormack 1998), which is a well-developed approach that we also drew on for the case studies used during the simulation. The PARIHS model has received recent attention as knowledge mobilisers search for a better understanding of the factors that drive research use (Stetler et al. 2011). The PARIHS framework stresses the interplay of three core elements: (1) the level and nature of the evidence; (2) the context or environment into which the evidence is to be placed; and (3) the way in which the process of KMb is facilitated.
The simulation exercise embodied all three of these elements.
The evidence was first summarised in the synthesis reports, then participants were asked to identify the best methods to facilitate the use of this information in different service contexts.

SIMULATION EXERCISES
A simulation may be either an opportunity to assess participants' competencies at specific tasks or, as used here, an experiential learning tool to facilitate problem solving and program codesign in complex environments. As a tool for assessment, the to prepare professionals, soldiers and even politicians for the exigencies of in situ challenges. For example, the CSI has used simulations to train more than 600 military officers from more than 60 countries on how to improve their interactions with child soldiers. The techniques are reported to be effective in preparing soldiers for the challenges that exist in current conflicts, which in turn creates situations where children are more ably protected and the mental health of the soldiers is considered. In situations like this, the complexity of the setting in which knowledge must be applied requires simulation participants to address 'wicked problems' through innovative practices rather than by imitating best practices.
Our KMb simulation was designed to enable participants to experience much the same fluidity in response to the complex service ecologies that provide support to CYCC. As the above examples show, simulations can help to generate changes to individual care practices, program design and policy development, and provide opportunities for values clarification. Based on our experience with simulations in these other contexts, we anticipated that they could be used to good effect to generate change strategies at the level of individual practitioners, service teams and provider organisations unfamiliar with KMb. Our goal was to do more than just ask service providers what they thought they needed to access program knowledge. The simulation added a degree of rigour and creativity to the process of discovering effective KMb strategies that fit the specific needs of a particular group of service providers largely unfamiliar with KMb. In this sense, the activity avoided abstractions and provided us with an opportunity to look critically at KMb strategies. We were able to problem solve how to make KMb most effective with our colleagues present as both commentators and facilitators of innovation.

SIMULATING KNOWLEDGE MOBILISATION
Members of a network of practitioners, academics and researchers (The CYCC Network) concerned with the wellbeing of CYCC designed a simulation exercise to explore innovative ways to both push knowledge from those providing and/or researching effective interventions and pull knowledge from those seeking to expand the scope of their practice. The federally funded network was established both to share innovative practice knowledge between stakeholders and to find ways to engage CYCC service providers in KMb activities in contexts where the concept of knowledge mobilisation was relatively unknown. For this reason, few of the agency staff who participated had experience with KMb or could describe a specific approach, even though almost all had at some point during their careers adapted programming to meet the needs of their focal population. The following is a brief description of the logistics of the meeting and sequence of events: -The three-day event took place at a retreat centre to encourage informal contact between participants outside of the formal work sessions.
-Prior to the meeting, participants were provided with documents summarising evidence-based practices, practice-based evidence and local examples of services thought to be effective with CYCC.
These documents were co-produced by university researchers and community partners and reflected the diversity of sources of knowledge that describe best practices. This co-production was important, not only to the engagement of stakeholders across the network, but also to the identification of best practices, most of which are not described in the academic literature.
-A process of appreciative inquiry (Watkins & Bernard 2001), a methodological approach to focus attention on programs that work, was used to gather from participants examples of successful KMb that they had already experienced.
-Opportunities were provided to work through three scenarios. (An example is presented in Appendix 1.) -Participants were asked to focus on what they could do to effectively share best practices with their colleagues in different contexts based on each scenario.
-After presentation of the proposed KMb strategies by teams of participants, discussions were held to explore the feasibility of the different approaches.
-Detailed process notes were kept and reviewed by meeting organisers. Exit interviews were conducted with participants and summary notes shared with participants for their feedback.

FINDINGS
To report on the process of the simulation and its impact, detailed observation notes were made by graduate students who participated in the event, and exit surveys, both written and video recorded, were conducted to solicit feedback and to assess the potential for future impact. The team that organised the event, including the authors, reviewed these documents for common themes. Results can be grouped broadly into two categories: (1) the experience of the simulation process as a tool to develop innovative KMb strategies; and (2) the KMb strategies themselves.

Process: Successes and Challenges
Participants were introduced to the simulation exercise at the beginning of the session with the following instructions: While all organizations working with children and youth need to understand knowledge mobilization, we will be looking more specifically at NGOs big and small, and government programs that are strongly community-based. We want to explore how knowledge mobilization can work for them. Our goal is to bring together roughly equal numbers of community members, academics, policymakers and youth. Together we will problem solve how we can make the exchange of evidence-based practice and practice-based evidence much easier for organizations with very few resources.
Despite these instructions, it was difficult to keep the focus of the simulation on the exchange of best practice knowledge between service providers. Conversations tended to focus instead on what works for particular vulnerable populations of young people and the contextual factors that impede program effectiveness. We will term these conversations first order (the exchange of immediate, program-level information regarding effective programs) and second order (the exchange of best practice knowledge between service providers that results in improved programming). KMb strategies for the exchange of best practice knowledge between stakeholders and the co-production of strategies to adapt best practices to specific contexts are examples of second order conversations. At this level, the focus is on communication patterns rather than on the specific content that is shared (see Bateson 1972). To explore first and second order exchanges of information, we used case examples borrowed from programs familiar to most participants, though the simulation proved disruptive when it challenged attendees to look at their own knowledge exchange strategies (how they access new sources of information for the programs they offer). In general, participants were more comfortable with first order exchanges of information than with higher order discussions regarding the effectiveness of what they do and the challenges they face sharing best practices between themselves and their colleagues.
Understandably, given the focus was as much on second order strategies as first order descriptions of interventions, several participants felt the meeting had not provided them with details of programming for the at-risk youth with whom they work. This problem was not unexpected, since the idea of KMb was new to most participants. After a second briefing on the simulation's goals, participants engaged much better with the task of coproduction of innovative strategies. The confusion was also addressed through a change midway through the process to create more homogenous working groups. For the second scenario, for example, we invited participants to work with their peers by forming groups of policy-makers, frontline service providers, administrators, and program participants (young people and their advocates). For the third scenario, we asked participants to sort themselves by one of three ways they most liked to receive knowledge: in writing (including websites), through interactive social media (Twitter, Facebook, webinars, etc.), and through creative arts-based forms of communication (photography, videos, dramatic representations of client experiences, etc.). Our experience was that these more homogeneous working groups were able to identify KMb strategies better than the heterogeneous groups where the context in which KMb was to take place was less well defined. It was also suggested by participants that the focus of each KMb simulation be narrowed to a particular type of program for a specific population (e.g. office-based therapies to help youth exposed to community violence). Greater focus, it was felt, would help the working groups explore contextually relevant ways to share knowledge.
Finally, the presence of young people in the simulation was seen as mostly positive as they influenced conversations about both process and content. For example, it was decided that in a field that valued client empowerment and strategies that addressed young people's marginalisation, having young people help with KMb activities could add to the credibility of the knowledge that was transferred (e.g. if young people themselves said a program was effective, other program developers were more likely to believe the evidence). However, some adult participants at the simulation found the presence of the young people distracting. In written exit surveys they expressed concern that the presence of the youth may have caused conversations to focus too much on what works rather than how to share what works between service providers. Youth, however, were very comfortable with the participatory atmosphere created by the simulation.

KMb Strategies
The most common remark in the exit interviews was that the value of the simulation had been the opportunity to meet with peers -People look to people they know or people whom their colleagues know for evidence of best practices. The more credible the individual, the more program-effectiveness data would be perceived as trustworthy.
-While search engines, websites and social media might be used during the preliminary stages of a search for innovative programming, participants preferred to make direct contact with the individuals who were operating the programs. Even better, participants preferred to hear first hand from both service providers and clients regarding the effectiveness of a specific intervention.
-Participants perceived a need for knowledge brokers, individuals and organisations that have the capacity to build bridges between individuals who hold evidence of effective practices and those who need access to that evidence. Funders were perceived as being ideally suited to play this role.
-The more familiar the source of the evidence, the more it was preferred and considered trustworthy. Local wisdom was viewed as more likely to be contextually relevant and easier to locate through professional networks.
-Participants tended to look for information about what they had already heard through the media, professional gatherings or word of mouth. Very little consideration was given to conducting surveys of the extant literature documenting services for a specific population.
-Service providers wanted to talk with other service providers, policy-makers wanted to talk to policy-makers. Each professional cohort wanted to find someone who could get them 'up to speed quickly' on new interventions and share knowledge relevant to their role in the decision-making process.
-Participants wanted others to notice what they were already doing right. A good exchange of knowledge was characterised by recognition that all those involved in the exchange had something positive to share.
-Participants responded best to a new program idea when there was a champion for the idea with whom they could interact. The more credible the champion, the more a program was perceived as effective.
-Participants wanted 'just in time' knowledge and easy to access mentors.
-The sharing of stories describing the effectiveness of an intervention was preferred to the sharing of data.
There were other preferred strategies for KMb that did not emphasise relationships, though all included a relational component. These included the need to evaluate program effectiveness in different contexts and with the participation of multiple stakeholders, the need to employ short-and long-term strategies to share effective practice and the need to account for differences between communities, again with reliance on local stakeholders to help knowledge mobilisers understand the context in which programs worked well.
Participants noted that larger organisations tended to be more data focused than relationship focused when involved in KMb. They relied more on systematic reviews of the evidence and decisions by committees rather than on a single advocate for a new initiative (usually connected to a program champion). Larger organisations also tended to centralise decision-making, which distanced those with evidence of program effectiveness from those who might be interested in adapting a program to their specific context and clientele. Furthermore, in larger organisations, it could be difficult to build commitment for a new approach to practice unless participants could assert, 'I heard it from someone I trust'.
Interestingly, the simulation exercise itself provided a forum for this building of trust. The social dynamic of the simulation generated familiarity between participants, such that some participants emerged as more credible sources of knowledge on program design.

DISCUSSION
The simulation exercise allowed us to answer the three questions with which we began: What are the barriers to KMb in less formal service settings? What strategies are already working? And how do we move knowledge between service providers in settings without familiarity with KMb or the resources to document and share effective practices? Our results highlighted the need for relationships as the basis for good KMb strategies. We note that this emphasis on relationships as the basis for KMb rather than the exchange of knowledge through written or online sources is not unique to providers of youth services. This is similar to the collaborations that underpin knowledge exchange in social work settings (Wilkinson, Gallagher & Smith 2012) and in ongoing knowledge-exchange partnerships between researchers and decision-makers in health services (Mitchell et al. 2009). However, this focus on relationships was more prominent than some might expect given that most of the participants in the simulation were active users of social media and had graduate-level professional credentials. This is a recurrent theme familiar to those who have studied the process of KMb and is a key reason for privileging relationship building when engaging community partners in KMb (Hart & Aumann 2013).
While the simulation activities identified preferences for particular KMb strategies among participants, it remains unclear whether these strategies would be effective in real-world settings. 5 Find ways to minimise the costs of maintaining the network, which encourages sustainability and self-sufficiency.
Where broker involvement is particularly time intensive, such an approach is perhaps best suited to situations in which the broker gains specific benefits from taking on this role. In the community-university partnership context, for example, universities brokering KMb can secure benefits for their research and teaching, providing a win-win solution to the burdens of gatekeeping and network organisation.
Though we used an iterative method of discovery (the simulation exercise) to identify the best strategies for KMb among service providers working with CYCC, our results are similar to those of others who have also found that KMb works best when relationships are emphasised. For example, Nutley, Walter and Davies (2007) have five mechanisms for the sharing of evidence among staff of public service organisations. These include: 1 Dissemination of research findings to potential users in formats tailored to the target audience 2 Interaction, by developing stronger links and collaborations between research and policy or practice communities 3 Social influence, through relying on influential others such as experts and peers to inform individuals 4 Facilitation of resources, to enable the use of research through technical, financial, organisational or social support 5 Use of incentives, reinforcement and rewards, to strengthen appropriate behaviours.
Each of these five mechanisms was reflected in the strategies discussed by participants during the simulation, though attendees emphasised that relationships were necessary for effective dissemination, interaction, social influence, facilitation and the use of incentives. Our experience suggests that in the context of community services for young people, relational factors are the single most important element for an effective KMb strategy.
Of course, this important finding may be a consequence of the methodology we used to explore KMb strategies. A simulation relies on interactions between participants, so it is plausible that, if we had discussed KMb using more didactic means, the KMb strategies preferred by participants may have been less relational.
While this is possible, the evaluations by participants suggest that, when given the chance to engage in KMb activities in an experiential way with colleagues, relationships are given more value than any other knowledge-sharing strategy. This is consistent with the CoP approach discussed above, which understands learning between stakeholders to be a situated social process. Nutley, Walter and Davies (2007), too, have shown that research uptake is a process that needs to be facilitated through interactive methods connecting researchers and research users. This act of facilitation is also one of three elements in the PARIHS framework, along with the nature of the evidence and the context in which that evidence is implemented into policy or practice (Stetler et al. 2011). Interestingly, Levin (2013) recently identified three similar elements for effective knowledge mobilisation in education: production of education research; the 'use' context; and mediation. These similarities arising from independent research in education and in health suggest that context, evidence and facilitation are elements in common across different settings, an important conclusion when considering knowledge mobilisation for CYCC service providers. Specifically, the simulation showed the need for active knowledge brokering.
We also note that, by creating greater homogeneity in the working groups during the simulation, it was easier for participants to work on the task of KMb. It has been shown elsewhere in the KMb literature that homogeneity among co-producers of knowledge facilitates familiarity and trust building, which is a key driver of success for KMb (Bennet & Bennet 2008). Simulation participants preferred having knowledge mobilised through connections with trusted intermediaries or knowledge champions, on whom they could rely to provide evidence of best practices. This emphasis on leadership (i.e. champions) has also been identified as a key element for effective knowledge mobilisation in previous studies of KMb (Hart & Church 2011;Wensing, Bosch & Grol 2009). We suggest that trusted organisations (like those represented at the simulation) can also play the role of knowledge brokers.
The potential of this social and emotional role for individuals and their organisations, rather than the quality of the knowledge being shared, has been growing as a focus in the KMb literature. Recently, the PARIHS framework was challenged to include the role of the individual in the process of implementation, with it being noted that '[a] robust and uncontested evidence base was a necessary, but not sufficient condition for practice change' (Roycroft-Malone et al. 2013, p. 28). Initiatives that seek to translate research into accessible formats, derive actionable strategies for program design from the evidence and make these strategies available in various electronic forms are insufficient on their own to generate engagement amongst stakeholders in the co-production of knowledge processes. The simulation showed that relational factors were very important to participants from community-based NGOs. Despite the high level of professional qualifications and confident social media use of our participants, a trusted intermediary was still their preferred route to knowledge exchange.

CONCLUSION
Simulation is a way of generating solutions to barriers to KMb and identifying innovative strategies for sharing best practices among members of organisations providing services to CYCC. -1 youth (Amin) -1 youth mentor who was a former gang member -2 service providers from the gang prevention program, SafeZone Hint: The service providers want to know where they can go for information that will help them to improve their program -1 lawyer who is also a board member with SafeZone The problem, however, is that Amin has made friends with other peers who are involved in a gang. Amin was already struggling with school and failing his classes because of his years of missed schooling in Somalia, but has recently started to not show up to classes at all. To make matters worse, he is constantly arguing with his parents who do not understand the music, clothes and the 'Western mannerisms' he has started to adopt. His father, who has been unable to find work, and has never quite recovered from the loss of his eldest son, has also started drinking excessively and become increasingly violent towards Amin and his mother.
Because of these troubles at home and at school, joining the gang was easy for Amin. The gang quickly accepted him and now some of the older youth in the gang have become like brother figures to him.
Recently, Amin was arrested by police for being involved in a violent crime and theft. He was referred to a gang prevention program known as SafeZone. This program is run by a group of dedicated community members, some of whom are former gang members themselves. SafeZone receives a small amount of funding from the Department of Justice to provide recreational activities and a safe space for the youth to hang out. The staff's emphasis has been on early prevention by getting the youth involved in recreation and sports, and encouraging young leaders to serve as role models for other youth in the area.
Increasingly however, the staff are encountering youth like Amin, who have complex case histories and are displaying signs of trauma. None of the staff has had formal training to deal with trauma or mental illness and they often feel that this therapeutic component is a major gap in their programming.
Although SafeZone's emphasis on youth leadership and recreation has definitely lead to positive and noticeable results with some of the youth, the staff realise that what is missing is a more comprehensive strategy to address the mental health needs of youth like Amin. They want to begin a dialogue with the school and police department in the community about how they can further expand and improve their program.

The Simulation
Remember, the goal of this role play is not to solve Amin's problems, but to: 1 Help the service providers from SafeZone access information that will help them to improve their program.
2 Get a better understanding of how information can be shared with service providers in a way that is relevant and accessible.

Role Play
The service provider from SafeZone is trying to find strategies that better serve the complex mental health needs of gang-involved youth like Amin. The service provider has arranged a community meeting.
-Where can SafeZone get the information they need to improve their program and meet the mental health needs of gang-involved youth?
-Once the information has been found, what does it have to look like for the service providers at SafeZone to be able to implement it and use it (i.e. how should the information be formatted)? How can the information be made useful to the service providers?

Barriers or Challenges
1 There is a lot of promising work being done on youth gang prevention and mental health promotion, but the problem is that this information is not always shared or easily accessible.