Skip to content

Experiential Learning

Principle · All principles
Evidence · 29 claims (20 for, 9 mixed) · 37 studies (10 causal, 8 quant-synthesis, 7 qualitative, 5 review, 4 theoretical, 2 associational, 1 design), q1–q4 · 7 of 37 report an effect size · 23 claims rest on one study

Conditional relationship

Experiential learning means learning from something the learner does or undergoes (a placement, a simulation, a lab, a field or community project, a hands-on task) rather than only from being told. This page owns the general relationship: an experience becomes learning when it is followed by structured reflection, connection to a concept or criterion, and a further attempt with feedback; experience alone, given to learners who lack the knowledge to interpret it, is the condition under which the relationship fails.

No claim in the wiki tests experiential learning as a whole, a full experience–reflection–conceptualization–experimentation cycle against an alternative of matched time. The only claim naming it directly is a second-hand report, inside one program-design article, that experiential learning increased engagement, practical skill and ethical reasoning in human services education (Reviewed studies report experiential learning benefits [+W]; no design, sample or effect size reported). What the wiki does test are components of the relationship, each in its own population:

  • Structured reflection after doing beats unstructured reflection: in school writing-to-learn and in a first-year engineering course (below).
  • Repeated, feedback-rich practice in a simulated experience beats traditional clinical education for procedural clinical skills in medical education (below).
  • Doing with materials beats symbol-only instruction in mathematics from kindergarten to college, and manipulation, physical or virtual, beats no manipulation for undergraduate physics concepts (below).
  • Unguided discovery learns less than explicit instruction, across domains and in an experiment with child learners who were novices on the target procedure, while discovery with feedback, worked examples or scaffolding does better than other instruction (below). One of the reviews behind that claim names experiential teaching among the minimally guided forms it argues against.

So the expected change is in skill-performance, conceptual-understanding and near-transfer, conditional on: the learner's task-specific knowledge of what the experience is meant to teach; whether the experience is guided and coached; whether reflection is structured by specific prompts rather than invited; whether the experience is connected to a named concept, rule or criterion; whether learners try again in a changed situation with feedback; the outcome chosen; and the horizon. No claim here reports a delayed outcome of an experiential cycle, an effect on attitudes or dispositions beyond self-report, or an effect on far transfer.

Siblings and pattern. Inquiry-based learning is one experiential form in which the experience is an investigation of a question; its conditional model (guided inquiry in school science, unguided discovery as the failure condition) sits inside this one and should be read for question-driven designs. Problem-based learning sits inside inquiry, organized around an authentic problem. Placements, simulations, service and practice-based designs are experiential without being inquiry, and are what this page adds. The experiential learning cycle is the reusable design built on this relationship; it has not yet been converted to this format, and its sequence is a proposal.

Default design, while the relationship is untested

The earlier page's guidance, with the pattern's sequence, kept as a concrete default a designer can act on and revise. Each step is a proposal; its evidence status is stated beside it, and no claim tests the sequence as a whole.

  1. Name the capability before choosing the experience: what the learner should be able to do afterwards (a clinical procedure, a judgement, an explanation, a way of working with people), in what situation, and how it will be checked. Untested; from the earlier page's learning objectives.
  2. Check what the learner can already do on this task, and teach what is missing first. Ask for the key idea or first move in a short probe. Where it is absent, demonstrate or explain it, with a worked example, before the experience, and narrow the experience so only one or two things can vary. Evidence: unassisted discovery learned less than explicit instruction, and discovery with feedback, worked examples or scaffolding did better (Unassisted discovery produces less learning than explicit instruction, while discovery enhanced with guidance outperforms other instruction [~S], below). Carried from discovery tasks to experiential tasks by analogy; not tested on placements or service.
  3. Give a meaningful experience tied to the goal: a task, role or situation the learner acts in, with consequences that can surprise them, and a stated noticing focus ("watch what happens to the patient's oxygen saturation when you..."). Untested; from the earlier page's requirements and the pattern's preparation step.
  4. Coach during the experience in response to what the learner does, raising support after a failure and lowering it after a success. From the earlier page; the contingent scaffolding [+M] claim it cited rests mostly on one-to-one tutoring of school pupils and has not been read against this model (Further evidence, below).
  5. Debrief straight after with specific prompts, not "reflect on the experience": what did you expect, what happened, where did they differ, what would you do differently and why. Evidence: specific prompts beat generic ones with the number of reflections held constant (the claim page reports 52 reflections apiece over a semester in the engineering study), and metacognitive-reflection prompts were the strongest content moderator in a 48-study writing-to-learn synthesis (Reflective Practice Improves Outcomes When Structured [+M], below). Neither study debriefs an experiential task; how soon after the experience is not tested.
  6. Connect the experience to the concept: the learner names the rule, principle or criterion the experience illustrates and produces an artifact (a stated rule, a written account, a revised plan). Untested; from the earlier page's requirement that facilitation connect experience to concepts, and the pattern's conceptualization stage.
  7. Try again in a changed situation, with feedback, and repeat: rehearse to a stated standard, raising difficulty across attempts. Evidence: simulation with deliberate practice (repetition to mastery standards, immediate feedback, increasing difficulty) beat traditional clinical education on procedural skills (Simulation Based Education With Deliberate Practice [+M], below). Tested on procedural and psychomotor clinical skills; untested for interpersonal or judgement goals.
  8. Judge the design on separate outcomes: performance on a new, unaided task; explanation of the concept; and, where it matters, a delayed check, not on satisfaction, journals completed or time spent doing. Untested; from the earlier page's constraint that experience can be mistaken for progress.

Constraints from the earlier page, still untested: experience alone can let learners repeat ineffective habits; high-complexity experiences can overwhelm novices; experiential tasks can become performative or logistically heavy if not tied to the goal; and reflection must go beyond narrating what happened toward interpretation and a next step. If a step cannot be provided (no facilitator, no second attempt), keep the steps that remain in the same order and shorten the experience rather than dropping reflection or the second attempt (untested proposal).

Fitting the design to a situation

The three facts that most change the decision:

  1. The learner's starting knowledge of what the experience is meant to teach. A novice cannot tell which parts of an experience matter, so an unguided experience is mostly noise; a learner with prior knowledge can interpret more of it unaided. If unstated, ask: have they done or seen a task like this before, and can they say what to watch for?
  2. The kind of goal. A procedure or psychomotor skill gains from repeated attempts to a standard; a concept gains from manipulation linked to its symbolic form; a judgement or an attitude-motivation goal has, in this wiki, only self-reported evidence. If unstated, ask: what will the learner do differently afterwards, and how would we see it?
  3. Who can structure reflection and feedback, and when. The tested components all depend on specific prompts and feedback; a placement with no one to debrief is experience without the rest of the relationship. If unstated, ask: who will be with the learner afterwards, for how long, and can learners try again?
If the brief says… Then change… Basis
Learners are novices on the target task Teach or demonstrate the key idea first, with a worked example; narrow the experience to one or two variables; supply the reflection prompts and the vocabulary for naming the concept; check an unaided step before removing support Unassisted discovery produces less learning than explicit instruction, while discovery enhanced with guidance outperforms other instruction [~S]; tested on discovery tasks, school and mixed domains, not on placements
Learners are advanced or experienced practitioners Shorten the preparation, make the experience less structured, and have learners write their own debrief questions; keep the second attempt with feedback untested proposal; the guided-inquiry claim's narrative review argues that the advantage of guidance recedes with prior knowledge, which no recorded study tests here
Learners are children, or school-age Use short, concrete hands-on tasks and connect each to its symbol or written form in the same session; tell them what to try, not only to explore Hands-on learning improves achievement [+M] (mathematics, kindergarten to college); Unassisted discovery produces less learning than explicit instruction, while discovery enhanced with guidance outperforms other instruction [~S] (third and fourth graders)
Adult professionals, workplace-clinical, a procedure or psychomotor skill Replace one-off exposure with simulated practice repeated to a stated mastery standard, immediate feedback after each attempt, rising difficulty, and a debrief; add booster sessions if the skill is not used soon Simulation Based Education With Deliberate Practice [+M]; medical education, procedural skills only; boosters untested
The goal is a science or mathematics concept Have learners manipulate the system (physical or virtual), predict before acting, and map the result onto the formula or diagram afterwards; if a physical lab is unavailable, a virtual one is a fair substitute Physical Experience Enhances Science Learning [~M] (undergraduate physics); Hands-on learning improves achievement [+M]
The goal is an attitude or disposition (civic responsibility, empathy, respect for diversity) Use service or community experience with structured reflection, and measure a behaviour or a decision in a new case, not only self-report; treat any benefit as unconfirmed Service-learning is associated with improvements in self-efficacy and social responsibility [+W]; second-hand and self-reported
online-self-paced, no facilitator present Replace the live debrief with specific written prompts after each task and a model answer to compare against; let learners retry a changed task with automatic feedback Reflective Practice Improves Outcomes When Structured [+M] for specific written prompts (engineering undergraduates, in a classroom course); untested online
One session (single-session) Run one complete loop (brief experience, prompted debrief, named rule, one retry) rather than a long experience with no debrief; cut loops, not stages untested proposal, from the experiential-learning-cycle pattern
Large group, one facilitator Run debriefs in pairs or small groups against a shared written protocol; spend facilitator time on the conceptualization step, where a wrong rule is hardest for peers to catch untested proposal, from the pattern; the reflection synthesis reports peer interaction as a moderator but the claim page does not say in which direction
Assessment is high-stakes, or credit attaches to reflection Assess performance on a new unaided task; do not grade journals for completion; score explanation and performance separately untested; from the earlier page's constraint that reflection must go beyond narration
Learners reflect in a second language, or are uncomfortable reflecting in public Offer written or recorded reflection, in the language they choose, before any group debrief; give sentence starters for the prompts untested proposal, from the pattern's personalization

Six rows rest on a claim (novices, children, clinical procedures, concepts, attitudes, self-paced online), each only as far as its population; the rows for advanced learners, one session, large groups, assessment and language are untested proposals.

Observation, state and explanation

What can be observed is what a learner does in and after an experience under stated conditions: the actions they take, what they say they noticed, the rule or reason they state in the debrief, the help and feedback they received, and what they do on the next, changed attempt. "Learned from the experience", "reflective" and "competent" are inferred states. Enthusiasm, time on task and a fluent journal are observations of participation; none shows that the intended capability changed.

Response Plausible explanations Observation that could change the interpretation
Debrief narrates what happened, names no rule or next step The prompt invited narration; the learner lacks the concept needed to interpret the event; reluctance to reflect in front of others Repeat with a specific prompt ("what did you expect, what happened, why the difference?") in writing, privately; if still no rule, probe the concept directly.
States the right rule in the debrief, repeats the error on the next attempt Rule held verbally but not yet proceduralized; next situation differed in more than intended; time pressure or load on the second attempt Give a component probe of just that step, then a matched retry with the same load; record the feedback given between attempts.
Performs well in the simulation, poorly on placement Simulation fidelity differs on the features that matter; anxiety or social demands of the real setting; skill not used since training Compare the two tasks feature by feature; observe a second real attempt with a known stressor removed; note the interval since practice.
A novice is busy throughout but draws the wrong lesson Unguided search without a schema; experience did not contain the target contrast; misreading surface features as causes Ask what they watched for; check whether the experience as run could show the target relation; demonstrate it and retry.
Reports large personal growth, little change in observed behaviour Self-report sensitive to the experience's emotional weight; change real but not in what was observed; observation too soon Observe a decision in a new case at a stated later point; compare self-report with a peer's or supervisor's record.

These rows are proposals, untested with learners. A probe can shift confidence among the explanations; it does not identify a cause on its own, and each probe is itself an experience that may change the learner.

Evidence and qualifications

  • Reflective Practice Improves Outcomes When Structured [+M]: a meta-analysis of 48 school-based writing-to-learn interventions found writing overall produced a small gain (d = 0.22) and prompts for metacognitive reflection were the only content-type moderator that reached significance (b = 0.48); a quasi-experiment in two sections of a first-year engineering course (N = 208, adult) found specific reflection prompts beat generic ones on exams, project scores and two problem sets with the number of reflections held constant, no effect size reported. Not settled: the abstracts could not confirm the entries. Neither study reflects on an experiential task; it supports the structuring of reflection, not experience itself.
  • Reflection interventions in higher education have a medium positive average effect on learning that varies with duration, peer interaction and the reflective activity used [~M]: a random-effects meta-analysis of 23 controlled studies of reflection interventions in higher education (2,010 participants) found a medium positive pooled effect (g = 0.56) that depended on intervention duration, whether reflection included peer interaction, and the reflective activity used. The claim page records that this one synthesis does not show mixed evidence, so its title is unconfirmed. Marked ~ because the average hides variation by implementation; it does not say which configuration to choose. Not settled: abstract only.
  • Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M]: a meta-analysis of 14 studies (mostly randomized trials and pre/post comparisons) of medical education found simulation with deliberate practice (repetition, mastery standards, immediate feedback) beat traditional clinical education on clinical skill acquisition (d = 0.71). Procedural and psychomotor skills only; the claim page notes patient-level outcomes and team or decision-making skills are less established. Not settled: abstract only. It supports a structured, feedback-rich experience, not experience as such.
  • Hands-on learning improves achievement [+M]: a meta-analysis of 55 studies (N = 7,237, kindergarten to college) found mathematics instruction with concrete manipulatives beat abstract-symbol-only instruction, with moderate-to-large effects on retention and small effects on problem solving, transfer and justification, moderated by instructional and methodological features. Mathematics only; not yet checked against its source. It does not separate manipulation from the prediction, discussion and reflection that usually accompany it.
  • Physical Experience Enhances Science Learning [~M]: a brief physical experience of angular-momentum forces improved college physics quiz scores (lab experiments and a randomized classroom experiment, no effect size in the abstract); in a second study, 234 undergraduates learning heat and temperature through physical, virtual or combined manipulation did equally well, and all beat traditional instruction without experimentation. Marked ~ because it suggests manipulation, not physicality, carries the gain. One of two entries passes the judge.
  • Unassisted discovery produces less learning than explicit instruction, while discovery enhanced with guidance outperforms other instruction [~S]: a meta-analysis of 164 studies found explicit instruction beat unassisted discovery across 580 comparisons (d = 0.38 favouring explicit instruction), while enhanced discovery (feedback, worked examples, scaffolding, elicited explanations) beat other instruction across 360 comparisons (d = 0.30); an experiment with 112 third- and fourth-grade novice children found many more mastered the control-of-variables procedure under direct instruction, and they did as well on a later transfer task as the few who discovered it; a narrative review argues minimally guided teaching, experiential teaching among its forms, is less effective for learners without enough prior knowledge. Partly checked: 1 of 3 entries passes; one judge failure was reviewed and dismissed. This is the boundary of the principle: it predicts poor learning from unguided experience for novices, not that guided experiential designs fail. The abstract does not break results down by prior knowledge.

How far each may be carried: the reflection evidence is school writing and one engineering course; the simulation evidence is procedural medical skills; the hands-on evidence is mathematics manipulatives and undergraduate physics; the discovery evidence is mostly school learners and mixed domains, read from abstracts. None is a placement, internship, service project or fieldwork study. Retain learner age and starting knowledge, the comparator (traditional clinical education, generic prompts, symbol-only instruction, explicit instruction), the amount of guidance and feedback, time on task, outcome type and horizon when transporting these results. A pooled effect over unlike implementations is not a forecast for one course or learner, and a nonsignificant difference does not establish equivalence.

The evidence by question

Further evidence, not yet read against this model

Claims this page cited before the 2026-10-02 rewrite. They are evidence about the relationship, but none has yet been re-read against the model above; each says how far its own sources have been checked.

Objective and learner-valued goal

Ask what the learner wants from the experience, and why, separately from the designer's objective. Experiential designs are often chosen because they feel relevant, and relevance to the designer is not the same as value to the learner. Agreement, divergence and uncertainty are observations; enjoyment of a placement or a simulation does not establish that the learner values the target capability, and a learner who values the experience may value something other than what is assessed. Choose a representative instrument, criterion, permitted help and horizon for each outcome that matters: skill performance, explanation, transfer to a new case and changed practice are different claims.

For example, a social-work student on a first placement may want to be trusted by the team and to feel able to sit with a distressed client, while the designer's objective is that the student can run a structured risk assessment and justify its conclusion. A confident debrief establishes neither. Record both aims, use the student's wish to be trusted to frame the debrief questions ("what would make the team rely on your assessment?"), and check the designer's objective separately on a new case, unaided.

What would revise this model?

The expectation should weaken if a full experiential cycle (experience, structured reflection, conceptualization, a changed retry with feedback) fails to beat a comparison of matched time on aligned performance or conceptual measures for comparable learners. If structured reflection adds nothing once feedback on a second attempt is given, the reflection step is not carrying the effect and should be shortened. If unguided experience is found to work for learners with stated prior knowledge, record that boundary rather than generalizing it. If attitude or disposition gains appear only in self-report and not in observed behaviour, say so rather than claiming them. Do not protect the model by relabelling every failure as a shallow debrief or an inauthentic experience after the fact.

An engaging experience, an immediate skill gain, an explanation, near transfer, delayed retention and changed practice in the learner's own setting are separate claims. The evidence recorded here establishes no optimal number of cycles, no fixed order of Kolb's four stages, no readiness threshold for unguided experience, and no effect of placements, internships or service as such.

Claims

Examples

Key Sources

  • Dernova, M. (2015). Experiential learning theory as one of the foundations of adult learning practice worldwide. Comparative Professional Pedagogy, 5(2). doi:10.1515/rpp-2015-0040
  • Hansman, C. A. (2001). Context-based adult learning. New Directions for Adult and Continuing Education, 89, 43-52. doi:10.1002/ace.7
  • Zeivots, S. (2016). Emotional highs in adult experiential learning. Australian Journal of Adult Learning, 56(3), 353-373.
  • Knowles, M. S., Holton, E. F., & Swanson, R. A. (2015). The adult learner (8th ed.). doi:10.4324/9781315816951
  • Torkington, Kate. (1996). The Rationale for Experiential/Participatory Learning. Working Papers in Early Childhood Development 16. Bernard van Leer Foundation. https://eric.ed.gov/?id=ED392940
  • Patil, T., Hunt, M., Cooper, K., & Townsend, R. (2020). Developing a case-based experiential learning model at a program level in a regional university: Reflections on the developmental process. Australian Journal of Adult Learning, 60(2). https://eric.ed.gov/?id=EJ1267932
  • Watkins, J. (2023). "That is Still STEM": Appropriating the Engineering Design Process to Challenge Dominant Narratives of Engineering and STEM. Cognition and Instruction, 41(4), 405-435. https://doi.org/10.1080/07370008.2022.2156512
  • Eggleton, Patrick J. (1995). Evolution of a Mathematical Philosophy: The Story of One Secondary Mathematics Preservice Teacher. https://eric.ed.gov/?id=ED389592
  • Hyde-Hills, I. (1998). It Is Better To Learn To Fish: Empowerment in Adventure Education. https://eric.ed.gov/?id=ED424058
Page metadata
Field Value
type principle
id experiential-learning
aliases experiential-course-design-practice-centered-diagnostic, kolb-derived-curriculum-guidelines, experiential-learning-as-program-philosophy, appropriating-disciplinary-tools-to-challenge-restrictive-narratives, mediate-cultural-choices-with-personal-meanings, primary-and-secondary-program-themes
title Experiential Learning
description For learners building applied, professional or interpersonal capability, an experience followed by structured reflection and a further attempt with feedback is expected to improve performance and transfer, while unguided experience for novices learns less than guided instruction.
status review
canonical True
generated by claude/unspecified on 2026-10-02
trust unverified — no one has confirmed this page's content
sources 5 — torkington-1996, patil-2020, watkins-2023, eggleton-1995, hyde-hills-1998

Frontmatter is the page's machine-readable record — see the Schema & Guide.