Perspectives of applied collaborative sport science research within professional team sports

Abstract The purpose of the study was to examine the perspectives of both academics and practitioners in relation to forming applied collaborative sport science research within team sports. Ninety-three participants who had previously engaged in collaborative research partnerships within team sports completed an online survey which focused on motivations and barriers for forming collaborations using blinded sliding scale (0–100) and rank order list. Research collaborations were mainly formed to improve the team performance (Academic: 73.6 ± 23.3; Practitioner: 84.3 ± 16.0; effect size (ES = 0.54), small). Academics ranked journal articles’ importance significantly higher than practitioners did (Academic: Mrank = 53.9; Practitioner: 36.0; z = −3.18, p = .001, p < q). However, practitioners rated one-to-one communication as more preferential (Academic: Mrank = 41.3; Practitioner 56.1; z = −2.62, p = .009, p < q). Some potential barriers were found in terms of staff buy in (Academic: 70.0 ± 25.5; Practitioner: 56.8 ± 27.3; ES = 0.50, small) and funding (Academic: 68.0 ± 24.9; Practitioner: 67.5 ± 28.0; ES = 0.02, trivial). Both groups revealed low motivation for invasive mechanistic research (Academic: 36.3 ± 24.2; Practitioner: 36.4 ± 27.5; ES = 0.01, trivial), with practitioners have a preference towards ‘fast’ type research. There was a general agreement between academics and practitioners for forming research collaborations. Some potential barriers still exist (e.g. staff buy in and funding), with practitioners preferring ‘fast’ informal research dissemination compared to the ‘slow’ quality control approach of academics.


Introduction
The appreciation and application of sport science support within team sports has grown exponentially over the past few decades. Support structures traditionally involved one sport science practitioner having a plethora of roles within a team, such as physical trainer, nutritionist and even sport psychologist.
The growth within the sport science sector is concurrent to the increased financial wealth of teams (Doust, 2011), allowing investment in both support staff and technology. The substantial growth in technology and data available to teams has led to an increase in the number of different support roles within a team. It is now commonplace for professional teams to have several sport science support staff in roles across the four disciplines of sport science: physiology, biomechanics, nutrition and psychology. Practitioners typically adopt roles such as strength and conditioning coach, data scientist, sport psychologist and rehabilitation fitness coach. Combined with colleagues from other disciplines, such as performance analysis and medical services, there is upwards of ∼15 support staff for one team, notwithstanding the team's technical coaching staff (Eisenmann, 2017).
Team sport practitioners work within a result-based environment and as such are faced with a high amount of pressure to deliver positive outcomes that enhance the team performance. Coutts (2016) recently proposed a conceptual model within applied sport science which involves both 'fast' and 'slow' methods of working. The 'fast' approach is often adopted by the practitioners working at the 'coal face' in which they have to make immediate decisions that have a direct impact on practice. Whilst this approach has short-term benefits, due to the applied nature of data collection and analysis, the quality control checking of the information provided can be of a lower standard. This has led to a number of collaborations between teams and universities, with the academics adopting a 'slow' approach in terms of quality control, critical analysis and validation of methods used. This concept of knowledge transfer has been defined as 'the process through which one unit (e.g. group or department) is affected by the experience of another' (Argote & Ingram, 2000). The successful implementation of such strategies on a long-term basis could lead to potential enhancement of the sport science support programme (Coutts, 2016).
In order to bridge the gap between both approaches, it is now commonplace for teams to employ both university research consultants and student interns within the organization (Jones et al., 2017). This 'embedded scientist' approach combines the roles of 'research practitioner' in which academic principles are used on a daily basis within practice. Such approaches provide further insight into which of the day-to-day performance questions need answering through scientific rigour. Bishop (2008) developed an Applied Research Model for the Sport Sciences (ARMSS) which aimed to provide a guide for those looking to undertake this collaborative approach. The ARMSS model is broken down into eight stages: (1) defining the problem, (2) descriptive research, (3) predictors of performance, (4) experimental testing of predictors, (5) determinants of key performance predictors, (6) efficacy studies, (7) examination of barriers (and motivators) to uptake and (8) implementation studies in a real sporting setting. This approach has become more popular despite sport performance research being seen as underfunded and with underutilized impact potential (Beneke, 2013).
Despite the increase in the amount of applied research being conducted by sport scientists, there still appears to be a gap when translating into practice with key stakeholders (i.e. coaches and athletes). Reade, Rodgers, and Hall (2009) examined the transfer of sport science knowledge to high-performance coaches and found that coaches still prefer informal conversations with fellow coaches to gain knowledge of sport science. It may also be the case that sport scientists often research what is relevant to themselves rather than the key stakeholders, recently defined as 'interesting' as opposed to 'useful' (Jones et al., 2017). Williams and Kendall (2007) found that coaches perceived a requirement for further research in sport psychology, which is often undervalued within the professional setting. Bishop, Burnett, Farrow, Gabbett, and Newton (2006) revealed the need for sport scientists to work on the communication of results to both coaches and athletes using their terminology rather than through traditional methods (e.g. journal articles). It may be the case that some lesser experienced sport scientists have a high level of theoretical knowledge but lack the 'soft skills' that come with more experience. Therefore, despite the increase in the number of collaborations within professional team sports, the efficacy of such programmes has not been examined.
Given the ever-growing competition for higher education institutions to attract prospective students to enrol upon sport degree programmes, there is a necessity for institutions to excel in higher education league table assessed criteria. For example, the Higher Education Funding Council for England (HEFCE) and Australian Research Council (ARC) have developed frameworks designed to assess the quality of research outputs from academic institutions (ARC, 2017;HEFCE, 2017). Outputs submitted for this review process are categorized using a tier structure based on the research quality and impact (e.g. from 'world leading' to 'below national standard'). Such assessment processes have placed pressure on academics to 'publish or perish', with a particular focus on attaining higher tier research outputs with public impact linked to funding opportunities. Such studies typically involve invasive, mechanistic-type research in order to be highly recognized from the research councils (e.g. 'four star' research rating). Although not empirically proven, such paradigms are likely to have important implications for the nature (descriptive or mechanistic), duration (fast or slow) and subsequent overall impact (interesting or useful) of collaborative opportunities that academics decide to pursue with team sport practitioners.
The purpose of the present study was to examine the perspectives of both academics and practitioners in relation to forming applied collaborative sport science research within team sports. Specifically, the study aimed to identify the outcomes and any potential barriers relating to collaborations.

Participants
Ninety-three participants (male = 82, female = 11) who stated that they had engaged in a collaborative research partnership within the previous 18 months of receiving an invitation to participate, voluntarily completed the survey between July and September 2017. The participants consisted of both academics (n = 57) and practitioners (n = 36). Although it must be acknowledged that participants may have been involved in both roles (i.e. as academics and practitioners), we defined each group based on their main job profession and source of income. All procedures were submitted and approved by the host institution's Ethics Committee (ref: 1617153) and conformed to the principles of the Declaration of Helsinki. Each invitation to participate was accompanied by a study information cover letter and participants were provided informed consent.

Procedure
The survey was distributed by the researcher team electronically using an online platform (Survey-Monkey, California, USA). A link for the online survey was emailed to potential participants and was then accompanied by a second email invitation to those who had not previously responded during the latter weeks of this period (September 2017). This resulted in a 43% and 56% survey completion rate for academics and practitioners, respectively.

Survey design
A survey consisting of 106 items was developed to gather information around academics and practitioner's perspectives to forming applied collaborative sport science research within team sports. The survey was specific to either academics or practitioners, but the number of items remained equal across groups. Items were developed by the lead researcher based on previous research and experience, which was then distributed to the research team for critique and further development. The survey was then pilot tested with a small sample of both academics and practitioners (n = 7) to establish its feasibility. This resulted in a positive response based on verbal feedback, with the use of the 'slider scale' function being commended in making the responses clear. In addition, the use of a progress bar within the online survey and organization of the survey by sections helped to alleviate survey fatigue based on pilot testing feedback.
Seven sections were developed for the survey: general information (Section 1: 25 items), motivations (Section 2: 17 items), formation (Section 3: 15 items), design (Section 4: 11 items), dissemination (Section 5: 17 items), overall perceptions (Section 6: 9 items) and barriers (Section 7: 13 items). The general information (Section 1) part of the survey comprised multiple-choice questions designed to ascertain the eligibility, suitability and additional information. Responders were required to use blinded, sliding (0-100) scales to evaluate the level of motivation (Section 2), responsibilities during collaboration formation (Section 3), research design (Section 4), preferred dissemination of findings (Section 5), overall perceptions (Section 6) and perceived barriers (Section 7) they apportion to discrete components of applied team sport research collaboration. This was followed by Perspectives of applied collaborative sport science research 149 an opportunity for the responder to expand upon their perceptions within an open-text box. For Section 5 (dissemination), respondents ranked which method of dissemination they would like to be used using a rank order list (1 = most preferred, 8 = least preferred).

Statistical analysis
Only fully complete returned surveys were used for the data analysis (n = 93, 45.2%). Preliminary analyses screened data for outliers using Q-Q plots and normal distribution using skewness and kurtosis values. All variables demonstrated acceptably normal distribution with values reasonably close to 0 (skewness < 2, kurtosis < 5), with no outliers identified (Field, 2017). Data were corrected for type 1 errors using false discovery rate (Benjamini & Hochberg, 1995). Null hypotheses were rejected if p < q and the 95% confidence interval did not contain 0. χ 2 analysis compared groups to determine even distribution of demographic variables within academic and practitioner groups. Independent-sample t-tests were used to compare responses between groups for motivation, responsibility, perceived importance of research facets, current and past research collaboration, and barriers to collaboration. Mann-Whitney tests examined the rank order variables of methods of research dissemination for practitioners and for academics. For each parametric test, 1000 bootstrapped samples were ran to generate mean survey scores ± standard deviation (SD), mean difference (M diff ) with 95% confidence intervals (95% CI), accompanied by relevant effect sizes (ESs) (<0.2 trivial, 0.2-0.6 small, 0.6-1.2 moderate, 1.2-2.0 large and >2.0 very large) (Hopkins, Marshall, Batterham, & Hannin, 2009).

General information
Data from respondents showed that 57% of respondents had participated in funded research, which tended to be equally financed (52.3 ± 36.8%). However, less than half (48.2%) declared that they used mutually agreed research contracts.

Dissemination of research findings
Academics and practitioners demonstrated some variation in identifying a rank (1most preferred to 8 least preferred) order of methods of perceived preference for research dissemination (Table II). Specifically, academics ranked journal articles significantly higher than practitioners did (Academic: M rank = 53.9; Practitioner: 36.0; z = −3.18, p = .001, p < q). However, practitioners rated one-to-one as more preferential (Academic: M rank = 41.3; Practitioner: 56.1; z = −2.62, p = .009, p < q). There was little difference between groups when identifying player preference.

Overall perceptions of research collaboration
In general, both academics and practitioners stated little agreement (≤ 50 (0strongly disagree to 100strongly agree)) to statements relating to their perceptions of current and past collaboration. The lowest scoring area for academics was their motivation to seek future collaborations (19.5 ± 24.9), and that practitioners had developed own knowledge (29.1 ± 28.5). Both academics and practitioners showed that the  Perspectives of applied collaborative sport science research 151 completion of the survey helped them to reflect upon research collaboration (Academic: 38.5 ± 24.5; Practitioners: 50.3 ± 24.5; ES = 0.48, small).

Discussion
The present study examined the perspectives of both academics and practitioners in relation to forming applied collaborative sport science research partnerships within team sports. In general, there appears to be agreement in motivations between academics and practitioners for research collaborations. Potential barriers that were identified include funding, time to dedicate towards the research and staff buy in. Differences existed in terms of how research should be disseminated, with academics preferring more formal outputs (e.g. journal articles and conferences) compared with practitioners preference for more informal methods (e.g. one-to-one conversations and infographics). Both groups reported low motivation for conducting invasive mechanistic research, with practitioners favouring 'fast' type research that has immediate impact on practice. Applied sport science research aims to produce an outcome that is relevant to sport and can be applied  (Bishop et al., 2006). In order for this to be achieved, relevant information generated from applied studies must be communicated effectively to the key stakeholders involved in the performance process (Martindale & Nash, 2013). The present study revealed that academics have a preference for research dissemination in journal articles and conference proceedings compared with practitioners who favour a more informal approach. Reade et al. (2009) found that coaches were least likely to gain sport science knowledge from academic journals due to lack of time and ability to interpret findings. Practitioners in the present study reported a higher preference towards infographics as a method of dissemination. The use of infographics is now commonplace on social media platforms, such as Twitter, with practitioners preferring their ease of access and simplicity in relaying information (Burke, 2017). Such methods may be useful to simplify the overall message to key stakeholders (e.g. coaches and athletes). However, as they only provide a 'snapshot' of the research study, practitioners and academics should critique the original research before then feeding forward. It may be the case that academics feel pressure to disseminate findings using established methods that can be used as part of university research quality metrics, such as the Research Excellence Framework (REF).
Whilst some publishers are now allowing the publication of informal methods such as infographics in their journals (see Heron et al. (2017) for example), their lack of ability to score high on the tier structure of research assessment frameworks will likely deter academics from this approach if key assessed metrics remain unchanged. One possible solution is for academics to be evaluated more clearly on their 'impact' (e.g. REF impact case studies) that results in a positive change to policy and practice.
According to the ARMSS model developed by Bishop (2008), applied research should aim to solve problems encountered in the applied setting through description, experimentation and implementation. It was found in the present study that both academics and practitioners had low motivation to conduct experimental research. By limiting this type of research, the projects may only reach stage 2 of the ARMSS model (i.e. descriptive) rather than being experimental to develop practice. Eisenmann (2017) refers to applied sciences as 'translational science' with the aim of bridging the gap between the laboratory and playing field. The main barriers for preventing invasive research appeared to relate to budget restriction and player/coach buy in.
Although it may be difficult to carry out laboratorybased methods in an applied setting, this should be seen as an interesting challenge for academics and practitioners rather than a hindrance. Recent studies have shown that it is possible to carry out invasive research designs within the applied setting, utilizing typically viewed 'laboratory methods' such as muscle biopsies (Bradley et al., 2016) and doubly labelled water method (Anderson et al., 2017) with elite team sport athletes. Whilst it has been acknowledged that sport performance research is underfunded (Beneke, 2013), both academics/ practitioners and external bodies (e.g. sporting teams, league representatives) should both look to contribute to finding solutions in order to overcome the potential barrier of funding to enhance our understanding of sport science.
In terms of potential barriers that may exist with establishing applied collaborative research, both academics and practitioners reported that funding and staff buy in were major challenges. One of the issues that may result in a lack of staff buy in is due to a lack of importance that non-scientific staff place upon sport science as a practice (Eisenmann, 2017). Whilst sport science has been adopted within coach education programmes for those currently coming through the system, some coaches may dismiss the usefulness of sport science research as it could expose a weakness in their current knowledge base. This finding was evident in the present study, with practitioners perceiving inferior knowledge as a greater barrier than academics (ES = 0.28, small). However, recent research has shown that coaches find sport science support useful, although the perception of purpose may differ between coach and practitioner (Weston, 2018). The issue around funding as a potential barrier may relate to who feels ultimately responsible for providing the finance for research projects. Only 48% of respondents used a mutually agreed research contract prior to commencement, with academics seen as responsible for the majority of the process. It may be speculated that some of the potential issues regarding funding may be due to a lack of ownership, with both parties having a difference in opinion in terms of who should ultimately be responsible for leading the collaborative projects. It would be recommended that both parties sign a research contract agreement when establishing collaborations to clearly outline the roles and responsibilities from both sides.
For the practitioner who works day to day in performance-based sport, the environment can be high paced and often demanding in terms of time commitment (Coutts, 2016). This type of industry can result in short-term planning amongst practitioners who may be concerned about the next result in order to keep themselves in employment rather than thinking long term. The present study supported this notion, with practitioners favouring the 'fast' type approach Perspectives of applied collaborative sport science research 153 to research projects rather than the 'slow' deliberate and focused approach. Whilst the 'fast' approach can be useful in the applied setting to get quick buy in from staff and athletes, ultimately the 'slow' research improves the quality control of data produced which then allows for long-term implementation. McCall et al. (2016) discussed the need for sport teams to adopt the 'research and development (R&D)' approach as used within the business world to generate new ideas and technology. The use of in-house research projects may potentially lead to competitive advantage with input from 'off-field brains' (Buchheit, 2017). However, the research conducted must be relevant to the team, rather than academics conducting research solely for personal interest reasons (Jones et al., 2017). One possible solution may be the increased use of 'embedded scientists' who work as part of the team and therefore can communicate information between the key stakeholders using their own practical language. This may also help to generate contextually relevant research questions that address 'real-world' practical issues (Buchheit, 2017).
One of the main issues that exists is the time frame involved from initiation of a project idea through to the final end product. Burgess (2017) describes the need for balance between using 'slow' type research and the practical realization of trying to implement approaches. Whilst this is a pertinent point raised, practitioners are sometimes guilty of ignoring the science component of sport science and adopting new methodologies without quality control and validation (Burke, 2017). Conversely, academics must look to improve the process in which research is administrated and disseminated (Buchheit, 2017). For example, peer review in scientific journals is a slow and inconsistent process that deters many practitioners from publishing their work (Smith, 2006). The promotion of relevant submission types (e.g. case studies), faster turnaround and accountability of reviewing and making content freely accessible may help with this process (Buchheit, 2017). It could also be argued that research should be disseminated in multiple ways across the continuum of science to practice, in order for all key stakeholders to feel involved (Jones et al., 2017). In addition, if practitioners and academics agree on the research objectives at the beginning of a project, this may allow for realistic expectations to be managed. The use of 'embedded scientists' allows research to be disseminated during the process, rather than waiting until the end of a research study cycle (Jones et al., 2017).
Whilst the information gathered from the present survey provides useful insight into the perceptions and potential barriers of collaborative research, several areas still require further investigation. The sample of respondents was mainly from Europe and Australia, with the majority working in soccer and rugby union. Differences in perceptions may exist in other regions across the world. For example, Asia is an emerging team sport market in which sport science is still in its relative infancy. It would be interesting to have a larger sample across other team sports to see if perceptions differ depending on the sport (including level of competition). Future research should also focus on strategies to overcome some of the potential barriers raised in the present study. It must be noted that whilst we have attempted to define academics and practitioners based on their main job role, both types sit on a continuum of practice (Jones et al., 2017). Further investigation into how people interact along this continuum would provide useful information about how we can maximize applied collaborative sport science research.
In summary, the present study found that there appears to be a general agreement in motivation between academics and practitioners for forming research collaboration. However, potential barriers still exist when forming such collaborations, most notably staff buy in and funding sources. Practitioners favoured more 'fast', informal methods of research dissemination (e.g. one-to-one conversations and infographics) compared to academics who preferred 'slow' scientific outputs (e.g. journal articles and conferences). Both groups were pessimistic about conducting invasive type research, mainly due to the barriers previously mentioned. Whilst difficult to conduct in the applied setting, such research can identify which interventions work with specific athletes and the potentially underlying reasons. We would recommend that both parties sign research contract agreements when establishing collaborations to outline the roles and responsibilities, whilst also managing the expectations across the research timeframe. The future of applied sport science research should look to develop research-active practitioners through academic collaboration and challenge the 'status quo' to achieve the highest standards of scientific rigour.