Kolb's Experiential Learning Cycle

6 min read · SessionData

In 1984, David Kolb published Experiential Learning: Experience as the Source of Learning and Development. The central argument was stark: learning is not the acquisition of knowledge. It is the transformation of experience.

This was a deliberate departure from the transmission model — the idea that learning happens when an expert deposits information into a learner. Kolb drew on John Dewey, Kurt Lewin and Jean Piaget to build a model where learning is a process, not an event. A cycle, not a transfer.

The four stages

Kolb proposed that learning moves through four stages, each qualitatively different from the last.

Concrete experience

The learner encounters something. A simulation, a real task, a difficult conversation, a failure. The experience is specific, situated, and felt — not abstract. It provides the raw material that the rest of the cycle works on.

Reflective observation

The learner steps back from the experience and examines it. What happened? What did I notice? What surprised me? Reflection is not passive — it is an active process of making sense of experience by reviewing it from multiple angles.

This is where Argyris and Schön's double-loop learning intersects with Kolb. Reflection that stays within existing frames is single-loop. Reflection that questions the frames themselves — the assumptions, values and mental models that shaped the experience — is double-loop. Kolb's cycle accommodates both, but the depth of what happens at this stage determines the quality of everything that follows.

Abstract conceptualisation

The learner forms a theory. Not necessarily a grand one — a working hypothesis, a principle, a revised mental model. "When I do X in context Y, the result seems to be Z." Conceptualisation turns the particular (one experience, one reflection) into something generalisable.

This stage connects to Senge's discipline of mental models. The models we carry — about how teams work, how people respond to feedback, how organisations change — are the product of prior conceptualisation. New learning either reinforces them or, in the best cases, revises them.

Active experimentation

The learner tests the theory. They try a new approach, adjust their behaviour, design a different intervention. Experimentation is the bridge back to concrete experience: the test produces new experience, and the cycle begins again.

The cycle is the point

The four stages are not a checklist. They describe a continuous process. Learning happens in the movement between them — not in any single stage.

A programme that provides rich experience but no time for reflection produces activity without insight. A programme that is all conceptualisation — frameworks, models, best practices — but no experimentation produces knowledge without capability. The cycle needs to complete.

This has a direct bearing on how programmes are designed. If Kirkpatrick's four levels describe what to evaluate, Kolb's cycle describes what to design for. A session that stops at experience (a team-building exercise with no debrief) or stops at theory (a lecture with no application) is an incomplete cycle. It is not that learning cannot happen — it is that the conditions for deep learning are absent.

Learning styles — the useful part and the overreach

Kolb also proposed that individuals have preferred learning styles — converging, diverging, assimilating, accommodating — based on which stages they gravitate toward. This became enormously popular in corporate training and enormously controversial in academic psychology.

The controversy is warranted. The evidence for fixed learning styles as a basis for instructional design is weak. People are not neatly sorted into types, and matching instruction to supposed style does not reliably improve outcomes.

But the underlying insight — that different stages of the cycle require different cognitive work, and that groups benefit from having people who are strong at different stages — remains useful. A team composed entirely of theorisers will struggle to act. A team composed entirely of doers will struggle to learn from what they did. The cycle, not the typology, is the durable contribution.

Where the cycle breaks

Kolb's model assumes the learner has time and space to complete the cycle. In practice, this is often what is missing.

The debrief gets cut because the schedule overran. The reflection happens but there is no forum to share it. The new approach is tried but the environment punishes deviation from the norm — which is why psychological safety matters. Without it, experimentation is too costly, reflection stays private, and the cycle stalls.

Programmes that produce lasting change tend to be ones that protect the full cycle: structured reflection, time to conceptualise, supported experimentation, and then — crucially — a return to experience to see whether the new approach worked. The cycle is not a metaphor. It is a design requirement.

Kolb, D. A. (1984). Experiential Learning: Experience as the Source of Learning and Development. Englewood Cliffs, NJ: Prentice-Hall.

Dewey, J. (1938). Experience and Education. New York: Macmillan.