Knowledge Foundations
The knowing–doing gap is the transfer failure at the centre of most training investment: learners complete a program, demonstrate learning, and then do not change what they do at work. Closing it draws on ten interconnected bodies of knowledge, ordered below by how load-bearing each is for the transfer problem specifically.
The ten knowledge foundations
1. Transfer of Learning
The home discipline. Baldwin and Ford's model frames transfer as both generalisation (applying learning to the job) and maintenance (sustaining it over time), shaped by trainee characteristics, training design, and — critically — the work environment. Most failures live here: not in whether people learned, but in whether learning travelled.
2. Motivation and Self-Determination
Whether a learner chooses to apply something is a motivational question. Deci and Ryan's self-determination theory identifies autonomy, competence, and relatedness as the conditions for durable, self-directed action; Bandura's self-efficacy explains the belief in one's own capacity that precedes attempting a new behaviour. This is the engine behind relevance and volition.
3. Behaviour Change Science
The gap is fundamentally a behaviour problem, not a knowledge problem. The COM-B model treats behaviour as the product of Capability, Opportunity, and Motivation; nudge theory addresses the choice architecture around the act; and implementation intentions — specific if-then plans — close the space between intention and action.
4. Universal Design for Learning
UDL supplies the design levers most relevant to transfer: its engagement principle builds relevance and autonomy into the experience, and its action-and-expression principle gives learners authentic, varied ways to apply and demonstrate capability. Flexible design manufactures transfer rather than hoping for it.
5. Adult Learning Theory
Knowles' andragogy holds that adults apply what is relevant to their immediate role, learn from accumulated experience, and engage most when learning is problem-centred rather than content-centred. The more a design mirrors the real task and its real stakes, the shorter the distance from classroom to job.
6. Transformative and Experiential Learning
The "doing" side, made explicit. Kolb's experiential cycle and Schön's reflection-in-action describe how behaviour becomes encoded through practice and reflection rather than transmission; Mezirow's transformative learning explains how adults revise the assumptions that otherwise keep old behaviour in place.
7. Authentic and Performance Assessment
If a program assesses recall, it produces recall. Authentic, application-based assessment — real tasks performed under realistic conditions — is both how transfer is built and how it is evidenced. What is measured signals to learners what actually matters.
8. Design for Application
The learning science of practice. Whole-task models such as Four-Component Instructional Design keep practice integrated rather than fragmented; spaced practice, retrieval practice, and deliberate practice are the evidence-based mechanics that move capability from the session into durable, retrievable skill.
9. Training Evaluation
Transfer that cannot be shown will not be funded. Kirkpatrick's Levels 3 (Behavior) and 4 (Results), and Phillips' ROI extension, provide the frameworks for evidencing that behaviour actually changed and that the change mattered — the measurement discipline that converts executive scepticism into confidence.
10. Change Management
New behaviour that the workplace does not reinforce is extinguished, however good the course was. Structured models such as ADKAR and Kotter's framework describe the manager support, reinforcement, and organizational conditions that determine whether a changed behaviour survives contact with daily work.
Use transfer science to frame the whole problem. Use motivation and behaviour-change science to move learners from intention to action. Use UDL, andragogy, and experiential learning to shape an experience worth applying. Use design-for-application methods to build durable skill, authentic assessment to signal what matters, and evaluation to prove it worked. Use change management to make the new behaviour stick. Beneath all of these runs cognitive load theory (Sweller, Mayer): overload during instruction is a quiet reason capability never consolidates into usable behaviour in the first place.
Transfer, not transmission
The distinguishing feature of this work is that the deliverable is not knowledge but changed behaviour. A content-focused approach optimises for whether information was presented and passed; a transfer-focused approach optimises for whether performance changed where the work happens. These produce materially different designs — in how outcomes are written, how practice is structured, how success is measured, and how the surrounding organizational conditions are engaged.
The transfer-of-training literature — beginning with Baldwin and Ford and developed through the applied work of practitioners such as Wick, Pollock, and Jefferson — consistently finds that post-training conditions, including manager reinforcement, opportunity to practise, and organizational climate, are stronger predictors of behaviour change than the learning event itself. A rigorous design therefore treats the environment around the learner as part of the design surface, not as something outside its scope.
This is also why the problem is named as a gap rather than a deficit. As Pfeffer and Sutton observed in their study of why organizations fail to act on what they know, the obstacle is rarely a shortage of knowledge; it is the set of conditions that allow talk, analysis, and completion to substitute for action. Learning design that closes the gap is designed backwards from the moment of application and forwards into the conditions that sustain it.
A collaborative, context-specific approach
Every organization presents a distinct set of tasks, constraints, incentives, and cultural conditions that determine whether a behaviour will transfer. Effective practice is developed collaboratively — with the people who perform the work, the managers who reinforce it, and the leaders who resource it — rather than applied as a generic template. The craft lies in diagnosing why a known behaviour is not being performed, and then designing both the experience and the surrounding conditions so that doing becomes the path of least resistance.
Design Process
Closing the gap is a deliberate sequence, not a property that appears if a course is engaging enough. The stages below describe a comprehensive approach; in practice, scope and depth are calibrated to each organization's capacity. Understanding the full sequence clarifies what is gained or lost when particular stages are abbreviated.
Diagnose the Gap
Is this a knowing problem, or a doing problem?
Before designing anything, it is essential to establish why a known behaviour is not being performed. The COM-B model provides a structured diagnosis: is the barrier capability (people cannot yet do it), opportunity (the environment or workflow does not permit it), or motivation (people are not disposed to)? Each points to a different intervention, and only capability gaps are primarily a training problem.
This is the stage that prevents the most common and expensive failure in the field — building a course for a gap that training cannot close. Where the barrier is opportunity or motivation, the appropriate response involves workflow, incentives, or management, with learning playing a supporting role at most.
- A behavioural diagnosis identifying capability, opportunity, and motivation barriers
- A clear determination of whether — and where — learning is the right lever
- A shared understanding with stakeholders of what training can and cannot resolve
Programs that skip diagnosis frequently deliver competent training against a gap that was never a training gap — a well-built solution to the wrong problem.
Define Behaviour-Level Outcomes
What learners will do, not what they will know
Outcomes are written as observable on-the-job performances rather than knowledge states. "Understands the escalation policy" is a knowing outcome; "escalates a qualifying incident within the defined window, following the documented steps" is a doing outcome. The distinction determines everything downstream: what is practised, what is assessed, and what counts as success.
Behaviour-level outcomes are agreed with the stakeholders who own the performance, so that the definition of success is shared before design begins rather than contested after launch.
- Outcomes expressed as observable, job-embedded behaviours
- Agreement with performance owners on what changed behaviour looks like
- The criteria that later evaluation at the behaviour level will measure against
Design for Relevance and Autonomy
Making application something learners choose
Transfer depends on volition: a learner has to want to apply the behaviour when no one is watching. Drawing on self-determination theory and andragogy, this stage builds in the conditions that support intrinsic motivation — clear relevance to the learner's actual role, meaningful autonomy over how they engage, and early experiences of competence that build self-efficacy.
UDL's engagement principle operationalises this: multiple ways to connect with the material, calibrated challenge, and choice in how learners participate. Motivation designed in at this stage is what carries a behaviour across the threshold from the session into the workflow.
- Explicit relevance framing tied to the learner's real responsibilities
- Autonomy-supportive structures: choice, self-direction, and calibrated challenge
- Early competence-building experiences that raise self-efficacy for the target behaviour
Build Authentic Practice
Rehearsing the real task under realistic conditions
Practice is the mechanism by which knowing becomes doing, and its design is where transfer is won or lost. Whole-task approaches such as Four-Component Instructional Design keep practice integrated with the real work rather than fragmented into isolated sub-skills, directly addressing the transfer paradox in which efficient, compartmentalised practice fails to generalise.
The evidence-based mechanics of durable skill are applied here: spaced practice distributed over time rather than massed; retrieval practice that requires learners to produce the behaviour rather than review it; and, where expertise is the goal, deliberate practice with specific feedback. Cognitive load is managed so that practice challenges the target skill rather than overwhelming working memory with extraneous demands.
- Authentic, whole-task practice that resembles the job rather than a quiz
- A practice schedule using spacing and retrieval rather than single-exposure review
- Feedback structures aligned to the target behaviour
Engineer the Transfer Climate
Designing the conditions the behaviour returns to
The strongest predictors of whether a behaviour transfers are found not in the learning event but in the environment the learner returns to. This stage engages that environment deliberately: securing manager reinforcement, ensuring genuine opportunity to perform the behaviour soon after learning, and identifying incentives or workflow frictions that would otherwise extinguish it.
This is where learning design meets change management. Manager briefings, reinforcement structures, and the removal of practical barriers are treated as part of the design, because a behaviour with no supportive climate to return to will not survive regardless of how well it was taught.
- A transfer climate plan: manager support, practice opportunity, and reinforcement
- Identification of environmental barriers that would undermine the behaviour
- Defined roles for managers and leaders in sustaining the change
Research on transfer consistently finds work-environment factors to be decisive. A program that addresses only the learning event has designed only part of the system that determines its own success.
Support the Moment of Application
Closing the space between intention and action
Even a motivated, well-practised learner faces a gap between intending to act and acting in the moment. This stage builds the supports that close it: implementation intentions — specific if-then plans that pre-commit the learner to act on a defined cue — together with job aids, prompts, and well-designed nudges positioned at the point of performance.
These supports are deliberately lightweight and located in the workflow rather than the course. Their purpose is to make the newly learned behaviour the easiest available option at the moment it is needed.
- Implementation-intention prompts linking specific cues to the target behaviour
- Point-of-performance job aids and reminders positioned in the workflow
- Choice-architecture adjustments that lower the effort of doing the right thing
Measure Behaviour and Results
Evidencing that the gap actually closed
Evaluation is designed in from the start and reaches past reaction and learning to the levels that define this work. Kirkpatrick's Level 3 measures whether behaviour changed on the job; Level 4 measures whether organizational results moved; and Phillips' ROI extension relates the value of that movement to the cost of achieving it. These levels require data beyond the learning platform — observation, performance metrics, and business indicators.
Designing for these levels requires measurement infrastructure and organizational agreement established before launch. It is also what allows the value of learning to be demonstrated on evidence rather than advocacy — the shift that changes how executives regard the function.
- An evaluation plan reaching behaviour and results, not only reaction and learning
- Behaviour-level evidence drawn from observation and performance data
- A defensible account of impact, and of return where the organization requires it
The great majority of workplace training is evaluated only at the reaction level. Designing for behaviour and results is precisely what distinguishes closing the gap from reporting completion.
Key Decisions That Shape Transfer
No two transfer problems are identical. The decisions below determine what kind of intervention is appropriate and how much of the surrounding system must be engaged. They also shape how the work is structured collaboratively with the organization.
Nature of the barrier: is this actually a capability gap?
The COM-B diagnosis determines the entire response. A capability gap is a learning problem; an opportunity or motivation gap is not, and treating it as one guarantees a program that cannot succeed. This decision precedes every other and prevents the field's most common failure mode.
Strength of the transfer climate: what conditions will the behaviour return to?
Where managers reinforce and opportunity to practise is immediate, the learning event carries more of the load. Where the climate is indifferent or hostile, the design must extend well beyond the event — or it will not transfer regardless of quality.
Transfer distance: how near is practice to the real task?
Near transfer, where the practice context closely resembles the job, is more forgiving; far transfer, where learners must adapt principles to varied situations, demands more authentic, whole-task practice. The greater the distance, the more the design must invest in realistic rehearsal.
Manager involvement: are managers part of the design?
Manager reinforcement is among the strongest predictors of transfer, yet it is frequently left to chance. Deciding early whether managers are briefed, equipped, and accountable determines whether the surrounding climate supports the behaviour or quietly erodes it.
Time to application: how soon can learners perform the behaviour?
Behaviour practised soon after learning transfers more reliably than behaviour deferred. When application must wait, the design compensates with spacing, retrieval, and point-of-performance supports to bridge the interval without decay.
Evaluation depth: how will changed behaviour be evidenced?
Measuring at the behaviour and results levels requires infrastructure agreed before launch. Organizations that define these measures early design differently — and are positioned to demonstrate value to leadership on evidence rather than advocacy.
A well-built learning experience is necessary but rarely sufficient. The decisions above share a theme: the further the transfer climate sits from ideal, the more the design must engage the system around the learner. Acknowledging this openly produces better outcomes than an excellent course that is expected, on its own, to change what people do.
Core Competencies
The following competencies form the professional foundation of transfer-focused learning design — spanning diagnosis, motivational and practice design, environmental engineering, and evaluation. Together they distinguish designing for behaviour change from producing training content.
Behavioural diagnosis
The ability to determine why a known behaviour is not being performed — separating capability gaps from opportunity and motivation gaps using structured models such as COM-B — and to recognise when learning is, and is not, the right lever. This is the analytical judgment that anchors everything downstream.
Motivational design
Designing for volition rather than compliance: building relevance, autonomy, and early competence into the experience so that learners are disposed to apply the behaviour unprompted. This draws on self-determination theory, self-efficacy, and adult learning theory, and underpins UDL's engagement principle.
Practice architecture
Structuring practice so that capability becomes durable, retrievable skill — through whole-task design, spacing, retrieval, and, where appropriate, deliberate practice — while managing cognitive load so that difficulty challenges the target skill rather than overwhelming the learner.
Authentic assessment design
Constructing assessments that require the actual behaviour under realistic conditions, rather than proxy performances that measure recall. Because what is assessed signals what matters, assessment design is itself a driver of transfer, not only a measure of it.
Transfer-climate engineering
Treating the work environment as part of the design surface: securing manager reinforcement, ensuring opportunity to perform, and addressing the incentives and frictions that determine whether a behaviour survives. This is where learning design and change management meet in practice.
Point-of-performance support
Building the lightweight supports that close the intention–action gap in the moment — implementation intentions, job aids, prompts, and nudges positioned in the workflow — so that the learned behaviour is the easiest available option when it is needed.
Behaviour and results evaluation
Designing evaluation that reaches past reaction and learning to on-the-job behaviour and organizational results, using the Kirkpatrick levels and, where required, the Phillips ROI extension. This includes building measurement infrastructure before launch and connecting learning data to performance data.
Stakeholder and manager partnership
Engaging performance owners, managers, and leaders as co-owners of transfer rather than recipients of a finished course. This requires translating the evidence on transfer into commitments those stakeholders can act on, and sustaining design integrity under organizational pressure.
Evidence fluency
The sustained ability to ground design choices in the transfer, motivation, and learning-science literature — and to explain, in terms a non-specialist stakeholder can act on, why a given move makes application more likely. This is what allows the practice to be advisory rather than merely executional.
Outcomes
Transfer-focused design creates change at two horizons. The shorter-term outcomes are tangible and measurable within a program cycle. The longer-term shift affects how an organization understands the relationship between learning and performance.
Shorter-term outcomes: what becomes visible and actionable
| Outcome | What this looks like in practice |
|---|---|
| The real barrier is correctly identifiedDiagnosis distinguishes capability gaps from motivation and opportunity gaps. | Effort is directed at a problem learning can actually solve, rather than a course built against a gap training cannot close. |
| Behaviour changes on the jobOutcomes are behavioural, practice is authentic, and the transfer climate is engaged. | Application at work becomes substantially more likely than with transmission-focused training. |
| Managers become part of the systemReinforcement is designed rather than left to chance. | The behaviour has a supportive environment to return to, and managers hold part of the outcome. |
| Impact is demonstrated on evidenceEvaluation reaches the behaviour and results levels, not completion. | Leadership receives a defensible account of what changed and why it mattered. |
Longer-term change: what becomes embedded in practice
| Outcome | What this looks like in practice |
|---|---|
| Learning is judged by performanceThe implicit measure shifts from delivered-and-liked to behaviour-and-results. | Programs are assessed by whether they changed what people do, not by completion and satisfaction alone. |
| Transfer conditions are designed by defaultManager reinforcement, practice opportunity, and point-of-performance support become standard. | These elements are built into design routinely rather than added occasionally. |
| The function earns strategic standingDemonstrated impact repositions learning from a cost centre to a capability. | Learning is resourced as something that moves organizational outcomes rather than as course production. |
| Executive confidence is restoredEvidence that training changes behaviour addresses completion-only scepticism. | The value of the function is argued on evidence rather than advocacy. |
Transfer research consistently finds that post-training conditions — manager reinforcement, opportunity to practise, and organizational climate — predict behaviour change more strongly than the learning experience itself. The practical implication is not that the learning event is unimportant, but that a design addressing only the event has engaged only part of the system that determines whether the gap closes.
Organizational Reflection
The questions below are intended to help surface useful considerations about how your organization currently approaches transfer. They are not a formal assessment. Take your time with them — the most useful answers are honest ones, not aspirational ones. Working through them with colleagues who hold different roles and perspectives tends to be more productive than working through them alone.
On diagnosing the gap
- How does your organization currently establish whether a performance problem is a learning problem — and how often is training the assumed answer regardless?
- Where are capability gaps, motivation gaps, and opportunity gaps distinguished from one another before a program is commissioned?
- How are your learning outcomes written — as knowledge states, or as observable behaviours learners will perform on the job?
On practice and durable skill
- How closely does practice in your programs resemble the real task, under realistic conditions?
- Where do spacing and retrieval appear in your programs, as opposed to single-exposure review?
- What do you know about whether skill built in a session is still retrievable weeks later — and how did you find that out?
On the transfer climate
- To what extent do your programs engage the environment learners return to — manager reinforcement, opportunity to practise, and reinforcement structures?
- Where are managers currently briefed, equipped, and accountable for reinforcing new behaviour, rather than left to support it informally?
- What point-of-performance supports — implementation-intention prompts, job aids, or nudges — currently exist in the workflow?
On evaluation and strategy
- How is the impact of learning currently evidenced — completion, satisfaction, learning gains, behaviour, or results?
- At which of Kirkpatrick's levels does most of your measurement happen, and what would it take to reach the behaviour and results levels?
- What would a realistic next step look like for your organization — given current resources, constraints, and the people with authority to move things forward?
The organizations that close the gap tend to be those willing to look honestly at where behaviour actually changes on the job — and where it quietly does not — rather than where they would like it to.
Citations
The frameworks and evidence in this resource draw on established scholarship and professional practice. Sources are grouped by the body of knowledge they primarily support, and are provided in full so that each may be independently verified.
Baldwin, T. T., & Ford, J. K. "Transfer of Training: A Review and Directions for Future Research." Personnel Psychology, 41(1), 63–105, 1988. Establishes transfer as generalisation and maintenance, shaped by trainee characteristics, training design, and work environment.
doi.org/10.1111/j.1744-6570.1988.tb00632.xBroad, M. L., & Newstrom, J. W. Transfer of Training: Action-Packed Strategies to Ensure High Payoff from Training Investments. Addison-Wesley, 1992. Source for manager, learner, and trainer roles across the transfer timeline.
Pfeffer, J., & Sutton, R. I. The Knowing-Doing Gap: How Smart Companies Turn Knowledge into Action. Harvard Business School Press, 2000. Origin of the term and its account of why organizations fail to act on what they know.
Deci, E. L., & Ryan, R. M. Intrinsic Motivation and Self-Determination in Human Behavior. Plenum Press, 1985. Source for autonomy, competence, and relatedness as the conditions for self-directed motivation.
selfdeterminationtheory.orgBandura, A. Self-Efficacy: The Exercise of Control. W. H. Freeman, 1997. Source for the belief in one's own capacity that precedes attempting a new behaviour.
Michie, S., van Stralen, M. M., & West, R. "The Behaviour Change Wheel: A New Method for Characterising and Designing Behaviour Change Interventions." Implementation Science, 6, 42, 2011. Source for the Capability–Opportunity–Motivation model of behaviour.
doi.org/10.1186/1748-5908-6-42Gollwitzer, P. M. "Implementation Intentions: Strong Effects of Simple Plans." American Psychologist, 54(7), 493–503, 1999. Source for if-then planning as a means of closing the intention–action gap.
Thaler, R. H., & Sunstein, C. R. Nudge: Improving Decisions About Health, Wealth, and Happiness. Yale University Press, 2008. Source for choice architecture and nudging at the point of decision.
CAST. Universal Design for Learning Guidelines, Version 2.2. CAST, 2018. Source for the engagement and action-and-expression principles applied to transfer.
udlguidelines.cast.orgKnowles, M. S., Holton, E. F., & Swanson, R. A. The Adult Learner: The Definitive Classic in Adult Education and Human Resource Development. 8th ed. Routledge, 2015. Source for relevance, experience, and problem-centred adult learning.
Kolb, D. A. Experiential Learning: Experience as the Source of Learning and Development. Prentice Hall, 1984. Source for the concrete-experience-to-active-experimentation learning cycle.
Schön, D. A. The Reflective Practitioner: How Professionals Think in Action. Basic Books, 1983. Source for reflection-in-action as the encoding of professional behaviour.
Mezirow, J. Transformative Dimensions of Adult Learning. Jossey-Bass, 1991. Source for the revision of assumptions that maintain existing behaviour.
van Merriënboer, J. J. G., & Kirschner, P. A. Ten Steps to Complex Learning: A Systematic Approach to Four-Component Instructional Design. 3rd ed. Routledge, 2018. Source for whole-task design and the transfer paradox.
Roediger, H. L., & Karpicke, J. D. "Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention." Psychological Science, 17(3), 249–255, 2006. Source for retrieval practice as a driver of durable retention.
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. "The Role of Deliberate Practice in the Acquisition of Expert Performance." Psychological Review, 100(3), 363–406, 1993. Source for deliberate practice with specific feedback.
Sweller, J. "Cognitive Load Theory, Learning Difficulty, and Instructional Design." Learning and Instruction, 4(4), 293–312, 1994. Source for intrinsic, extraneous, and germane load in practice design.
Mayer, R. E. Multimedia Learning. 2nd ed. Cambridge University Press, 2009. Source for reducing extraneous load in instructional materials.
Kirkpatrick, D. L., & Kirkpatrick, J. D. Evaluating Training Programs: The Four Levels. 3rd ed. Berrett-Koehler, 2006. Source for the Behavior (Level 3) and Results (Level 4) evaluation levels.
Phillips, J. J. Return on Investment in Training and Performance Improvement Programs. 2nd ed. Butterworth-Heinemann, 2003. Source for the ROI extension relating impact to cost.
Wick, C., Pollock, R., & Jefferson, A. The Six Disciplines of Breakthrough Learning: How to Turn Training and Development into Business Results. 2nd ed. Pfeiffer/Wiley, 2010. Source for practitioner disciplines linking learning to on-the-job results.
Hiatt, J. M. ADKAR: A Model for Change in Business, Government and Our Community. Prosci Learning Center Publications, 2006. Source for individual-level reinforcement of new behaviour.
Kotter, J. P. Leading Change. Harvard Business Review Press, 1996. Source for the organizational conditions that sustain change.