The Long Conversation Between Practice and Research

The question was not fixed at the beginning

A research question can look like a starting point.

Choose it carefully. Define it. Then go and answer it.

One of the older pieces of work Keith Lyons brought back into Clyde Street tells a less tidy story.

In April 2011, while supervising a growing number of Masters and PhD students, Keith republished the introduction to his own doctoral thesis. The writing itself came from 1989.

His original intention had been to research the teaching of boys’ physical education in secondary school. He had arrived with a sociological education, a commitment to qualitative research and, as he later admitted, more confidence in the language of research than in what he actually wanted the research to do.

Practice complicated that confidence.

From 1986, Keith worked with five physical-education teachers in two schools. He described that experience as transforming his understanding of qualitative research. Among the things he had not previously understood clearly was the difference between doing research with teachers and doing research on them.

His research question changed too.

He had begun by asking whether children learned in spite of or because of the teacher. After years of reading and empirical work, his central question had become how teachers constructed the contexts in which teaching and learning became possible.

The shift matters because Keith did not present the question as something protected from what happened next. The questions sat inside a process he described as emergent, fallible and reflexive. They were open to revision and replacement.

Research, in this account, does not simply begin with a question.

The research can change the question.

So can the people involved in it. So can practice. So can biography. So can a body of reading that makes an earlier way of seeing difficult to sustain.

That does not make personal experience evidence enough. Keith’s own account keeps a distinction between where a question comes from and what would be needed to investigate it. But it does make the origin of a research question less clean than the textbook version.

The question can be one of the outcomes.

When thought has to meet action

There is another historical layer in Clyde Street that makes the relationship more demanding.

Four months before the 2011 PhD republication, Keith had posted a paper written in 1993: Working with Coaches and Performers: Notational Analysis as Liberational Praxis?

The publication dates can mislead here. The 2010 Clyde Street post contains 1993 thinking; the 2011 post contains 1989 thinking. Read at their historical layers, the doctoral questions come first.

By 1993, in a different professional setting, Keith was working with governing bodies in Welsh rugby union and women’s lacrosse. The paper asks a different but related question: what happens when analytical knowledge enters practice?

Praxis gave him one way of thinking about it. Drawing on Paulo Freire, Richard Tinning, Patti Lather, Linda Bain and others, Keith used the idea to hold thinking and action together. At one point he described praxis as the fusion of thinking and practice.

That was an aspiration, not a guarantee.

The analytical work was defined closely with coaches. Sometimes weeks or months of observation produced information they found useful and occasionally helped them recast what they knew about performance. At other times, similar amounts of work did little. Analysis could block action. It could make decision-making worse. It could become an academic exercise whose liberational language was more rhetorical than real.

That is important for the conversation between practice and research because it prevents an easy division of labour.

Research does not become useful simply by arriving in practice.

Practice does not become research simply because it is reflective.

The interesting work happens in what each forces the other to confront: what counts as a useful question, what kind of evidence is needed, what can be acted upon, and what has been made harder rather than easier by the intervention.

The coach becomes a researcher

By July 2012, the relationship was visible in a much more concrete form.

Finn Marsland had presented an upgrade seminar at the University of Canberra on cross-country skiing performance analysis using micro-sensors. Keith was one of his supervisors. Finn was also the National Coach for Cross Country Skiing.

Keith called the project a great example of praxis: theoretically informed action.

The significance was not that a researcher had finally managed to reach a coach. Finn occupied both roles at once: national coach and doctoral researcher. The source does not tell us that every research question arose directly from coaching practice, but the inquiry was being conducted from inside a professional coaching role while also being subjected to formal research, publication and scrutiny.

Keith hoped Finn’s work would encourage other coaches to research their practice and share openly their findings.

That hope should not be turned into a rule. A coach does not become a good researcher merely by investigating a familiar environment. Proximity can create insight and blind spots. Formal research does not become more valid because the researcher is also a practitioner. Nor does one coach-researcher project demonstrate that the profession changed.

But Finn’s work alters the geometry of the familiar “research-to-practice” picture.

There is no clean handover point at which research finishes and practice begins.

In Finn’s case, a practitioner was also the person conducting formal research. That is enough to disrupt a one-way transfer picture without claiming that every research question began in practice or that every result returned to it.

Practice asks back

A year later, the direction changed again.

In September 2013, Darrell Cobner asked a practical question on the Visual Performance Analysis blog: “Any performance analysts actually use statistical tests in the applied world?”

Keith’s short answer was that he had not.

But he did not stop there.

Darrell’s question sent him back through his own methodological history. Keith described three decades of descriptive-analytic systematic observation, his qualitative turn, his reliance on purposive sampling, his concern with validity and reliability, and his interest in narrative knowing and storytelling.

He was explicit about limits. He did not claim random sampling. He did not claim representativeness. He said he hoped what he shared would resonate with others and that he had tried to make the constraints visible.

At the same time, he was encountering other kinds of work. A PhD student with decades of coaching and analysis experience was learning to use statistical tests. Keith had reviewed a paper in which statistical work and coaching interpretation sat together. Younger analysts were blending statistical insight, algorithms, visualisation and applied work.

The point is not that Darrell’s question converted Keith from qualitative to quantitative research. It did not.

The interesting movement is that a practitioner’s question made Keith reopen what counted.

What kinds of evidence belonged in performance analysis? What could statistical insight add? What should remain descriptive? How could analysis become a story without hiding its methodological constraints?

Practice was not waiting at the end of the research process to receive an answer.

It was asking a question back into it.

The next question changes the pedagogy

By December 2018, the conversation had reached teaching.

Keith had been reading several posts about pedagogy and thinking about what a new age of performance analytics might require. The prompts came from outside sport and from inside high-performance practice.

Laura Seth had shared news of work on effective observation. Mladen Jovanović had published on training-load data and machine learning. Sam Robertson had listed kinds of sport-science and analytics research he wanted to see more of: better practitioner-team structures and communication, collaboration across clubs and universities, longitudinal interventions, better coach–athlete communication, raw tracking analysis, interdisciplinary work, cognitive science, field application, automation and human–machine integration.

Keith noted that these were people actively engaged in high-performance service delivery.

Their questions did not become his findings. They became part of the problem he thought education needed to answer.

He began imagining a pedagogy of praxis for performance analytics and said his own pedagogy was moving more strongly toward an “unmeeting” approach.

Again, this was prospective. Clyde Street does not show that the proposed pedagogy produced better analysts, better learning or better performance.

What it does show is another turn in the conversation.

If the practical and research problems of a profession change, the way new practitioners are educated may have to change too.

The next research question can become a curriculum question.

A long project returns

In September 2019, Finn Marsland returned to Clyde Street.

Seven years had passed since the upgrade seminar. Keith now described a completed PhD in macro-kinematic performance analysis in cross-country skiing using micro-sensors. The graduation ceremony was still forthcoming in this source; the thesis work itself was complete.

Keith again framed the project through praxis, this time emphasising Finn’s practice as a coach alongside a theoretical understanding of emerging micro-sensors developed with Colin Mackintosh.

The thesis had become substantial: multiple studies, published papers, algorithms for identifying skiing sub-techniques, comparisons between sensor- and video-derived measures, competition analyses and a final case study concerned with what coaches might learn from individual macro-kinematic variation.

It would be easy to make this the success-story ending.

That would give us a neat ending: the early coach-researcher idea matured, the PhD was completed, and research and practice were united.

The source is more useful than that because the limits remain visible.

Some sub-techniques were under-reported. Turn detection caused misclassification. Manual reclassification was needed. Samples were small. Several practical applications were still described as possibilities for future work.

The research had not abolished uncertainty. It had made some of the uncertainty more precise.

And it had generated more things that could now be asked in practice: how to monitor training, profile courses, evaluate sub-technique efficiency, adapt strategies to terrain and other skiers, and improve the algorithms themselves.

The answer had become another set of questions.

The long conversation

“Research and practice” are often pictured as two places with a gap between them.

Sometimes that picture is useful. Institutions can be disconnected. Research can be inaccessible. Practitioners can be asked to use findings produced without much knowledge of their context. Academic work can fail to answer the problems that matter in practice.

But these Clyde Street sources do not form a story about finally building a bridge.

They show something more recursive.

In the 1989 doctoral writing, practice, biography, reading and relationships change the research question itself. In the 1993 paper, research knowledge has to survive contact with action and can fail when it obstructs the people it is meant to support. Finn Marsland’s work makes a coach a producer of formal research rather than merely its recipient. Darrell Cobner’s practical question sends Keith back to method and evidence. Later practitioner questions help reshape what Keith thinks future analysts should learn. Finn’s completed thesis then places new methods and new uncertainties back beside questions of coaching practice.

These six sources are not a continuous diary of the thirty years between the earliest writing and Finn’s PhD. They are selected points where the relation becomes unusually explicit. The gaps between them matter.

None of that makes research and practice equivalent.

These sources also keep differences visible. Formal research carries methodological questions about validity, sampling, misclassification and limits. Practice carries context, consequences and decisions that cannot always wait for perfect evidence. Neither cancels the other.

That friction is not necessarily a defect in the relationship.

It may be part of what keeps the conversation useful.

The question to carry from Clyde Street is therefore not simply: how do we get research into practice?

It is also: what has practice made research ask next?

Question to carry

And after the answer comes back, what has changed — the question, the method, the representation, the pedagogy, the practical problem, or perhaps only our confidence that we understood it in the first place?

The conversation remains open because either side can still answer back.

Keith Lyons, “Of Personal Experience And Research Questions” — 11/04/20; republishes the introduction to his 1989 PhD

Keith Lyons, “Working With Coaches And Performers” — 10/12/02; republishes the 1993 paper “Working with Coaches and Performers: Notational Analysis as Liberational Praxis?”

Keith Lyons, “Analysing Cross Country Skiing Performance” — 12/07/04

Keith Lyons, “What Counts” — 13/09/19

Keith Lyons, “Performance Analytics and Pedagogy” — 18/12/20

Keith Lyons, “Finn’s Graduation” — 19/09/13

Source and ownership note

“The Long Conversation Between Practice and Research” is an editorial framing, not the name of a Keith Lyons model and not a claim that research and practice became equivalent or that a field-wide research-to-practice gap was solved. Of Personal Experience And Research Questions republishes 1989 doctoral writing; its teacher participants retain their voices and the cited qualitative, ethnographic and sociological authors retain their intellectual contributions. Working With Coaches And Performers republishes a 1993 paper; Paulo Freire, Richard Tinning, Patti Lather, Linda Bain and the other praxis sources retain their concepts, while the performance-analysis literature retains its own authorship. Finn Marsland owns the cross-country-skiing research described in Analysing Cross Country Skiing Performance and Finn’s Graduation. Darrell Cobner owns the initiating question in What Counts; the students, reviewed-paper authors and statistical researchers retain their methods and findings. The teaching and practitioner sources in Performance Analytics and Pedagogy retain their ideas and research priorities. Keith owns his source-visible acts of returning to old work, changing and exposing research questions, supervising, connecting practice with inquiry, reconsidering method and altering his pedagogical response.

This article does not claim that practice is evidence, that practitioner involvement guarantees useful or valid research, that coach-researchers are inherently superior researchers, that supervision transfers research ownership, that publication proves implementation, that pedagogy proves learning, that open sharing proves uptake, or that praxis certifies a successful integration of thought and action.