Skip to content

Debriefing

Principle · All principles
Evidence · 23 claims (15 for, 7 mixed, 1 against) · 39 studies (11 causal, 11 quant-synthesis, 8 review, 4 qualitative, 2 associational, 2 theoretical, 1 design), q1–q4 · 9 of 39 report an effect size · 12 claims rest on one study

Conditional relationship

This page owns the facilitated conversation after a shared experience. For learners who have just taken part in a simulation, a case, a team exercise, a field task or a real event that they experienced together, a debrief led by someone who knows the target performance (a facilitator, a teacher, a senior colleague, or a written protocol standing in for one) is expected to improve their next comparable performance (immediate-performance, near-transfer) more than the experience alone or an unstructured recap ("how did it go?"), when the debrief does four things: it establishes what actually happened; it asks learners to explain their reasoning at a few decision points tied to stated criteria; it gives feedback about the task and the process, not the person; and it ends with a change each learner can try at a next attempt. The expected gain is in the next performance and in the learner's explanation of it; a fluent debrief is not itself the outcome.

The relationship is conditional on the learners' expertise on the task (novice learners need a more explicit, facilitator-led debrief focused on process and cues; advanced learners can lead more of it themselves), on whether learners can judge their own performance (untrained self-assessment is unreliable), on psychological safety (whether learners will admit errors in front of this group), on whether a next attempt follows, and on time. No claim in this wiki compares the same experience with and without a debrief, or one debriefing method with another. The nearest evidence tests the debrief's ingredients in other settings (structured reflection prompts, self-explanation, task- and process-level feedback, peer plus instructor explanation, contingent support), and a simulation meta-analysis whose subgroup analyses could not isolate feedback as the active feature. So the whole relationship is a proposal built from tested parts.

How the neighbouring pages sit beside this one. Experiential Learning owns the general relationship (experience followed by structured reflection, a stated principle and a changed attempt); the debrief is its reflection step, and the Experiential Learning Cycle pattern carries the repeated loop. Reflection owns a learner examining their own past performance against a criterion, often alone and in writing; this page adds what changes when the examination is a facilitated, usually group, conversation about one shared event: someone leads it, several people saw the same thing differently, and errors are discussed in front of others. Error Analysis owns working on mistakes as information. The element page Debrief holds the design decisions (whether a debrief adds anything, how structured its questions are, what feedback is about, who explains, length and timing), each with its evidence; this page does not repeat them.

Default design, while the relationship is untested

The earlier page's guidance, made concrete as a default a designer can act on and revise. Each step is a design proposal; its evidence status is stated beside it. Durations are untested proposals.

  1. Before the experience, name what the debrief will examine. Write down one to three performance criteria or learning objectives (for a resuscitation scenario: recognition, call for help, roles; for a negotiation role-play: what each side wanted and what was conceded) and appoint someone, or a recording, to capture what happens at the moments that bear on them. Tell learners the ground rules: what is said stays in the room, errors are what the debrief is for, the debrief is not graded. Untested; from the earlier page's requirements (a concrete shared experience, psychological safety).
  2. Start straight after, and protect the time. Begin within minutes, while details are still accessible. Plan the debrief to run about as long as the experience for a short scenario (a 15-minute simulation, 15–20 minutes of debrief) and 5–10 minutes after a class activity. A debrief cut to a few minutes after a rich experience becomes a recap; if time is short, shorten the experience, not the debrief. Untested; from the earlier page.
  3. Open with reactions, briefly (2–3 minutes). Ask how it felt and let everyone say a sentence. After a high-stress event, give this more time and settle the group before analysis (Check-ins); after a distressing real event, emotional support comes first and is not this debrief. Untested; from the earlier page.
  4. Establish what happened before asking why. Learners reconstruct the sequence together, checked against the recording, the observer's notes or the monitor log, so the analysis is about events, not impressions. Untested here. Learners' unaided judgments of their own performance are often inaccurate and improve with criteria and models (Self Assessment Accuracy Is Low Without Training [~M], secondary and university students).
  5. Analyse two or three decision points, not everything. At each, ask the learner who acted to explain it: what were you seeing, what did you decide, why. Prompting learners to explain their reasoning improves understanding and problem solving on immediate tests (Prompting learners to self-explain improves understanding and problem solving on immediate tests by a moderate average amount, with little evidence yet on delayed or classroom gains. [+S]; tested while studying examples and text, carried here to explaining one's own actions). Use specific questions tied to the criteria, not "what did you learn?": specific prompts beat generic ones with the amount of reflection held equal (Reflective Practice Improves Outcomes When Structured [+M]; a classroom engineering course and school writing-to-learn studies, not post-simulation debriefs).
  6. Ask how sure they were, and stay on confident errors. Before correcting a wrong decision, ask how confident the learner was; give the correction clearly and check it is understood. Errors held with high confidence were remembered correctly more often after correction than low-confidence ones (High-confidence errors are more likely than low-confidence errors to be corrected on a later test after feedback. [+S]; studied with factual answers and immediate corrective feedback, carried here to decisions in a performance).
  7. Let peers explain first, then the facilitator. Ask other participants or observers what they saw and would have done; then the facilitator says what the criterion requires and where the performance fell short. In two classroom studies, peer discussion followed by instructor explanation gave larger gains than either alone, for weak, medium and strong students (Peer Discussion Improves Conceptual Understanding [+M]; undergraduate genetics concept questions, no randomised control).
  8. Make the facilitator's feedback about the task and the process, never the person. Name what was done, what the standard is, and the strategy that would close the gap. High-information feedback on the task and process outperformed reinforcement or punishment, and feedback lost effect as it drew attention to the self (Feedback Most Effective At Task And Process Levels [+S]; education studies and laboratory and field studies, not debriefs specifically).
  9. Follow what learners noticed and missed. Ask open questions first; if learners do not find the gap, narrow the question, then point to the moment, then state it. Become less directive as they find gaps themselves. Contingent support beat fixed or absent support in one-to-one tutoring (Contingent scaffolding improves learning more than fixed or absent support. [+M]; children, long division, 8 per condition; untested in group debriefs).
  10. With novices, debrief the process, not only the result. Ask which cue they used and which step they took, and say explicitly what the expert would have attended to; leave outcome-focused questions ("did the patient survive?") until the procedure is secure. Novices who set process goals outperformed those who set outcome goals (Process goals lead to better skill acquisition for novices than outcome goals. [+M]; high-school writing revision, carried here to what a debrief focuses on).
  11. Close with one change per learner, and use it. Each learner says or writes one specific thing to do differently ("call for help before the second check", not "communicate better"). The next experience opens by reading those changes, and its debrief starts by asking whether they were tried. Untested as a loop; simulation with deliberate practice (repetition to a mastery standard with immediate feedback) beat traditional clinical education (Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M]), which treats the debrief as the feedback step of a repeated attempt. If no next attempt exists, shorten the debrief to the criteria and the correction.
  12. Across sessions, hand the debrief over to the learners. A default arc over a course of six or more debriefed experiences: early sessions (first two): the facilitator leads with a written protocol (reactions, what happened, two or three decision points, one change each), names the gaps learners miss, and supplies the vocabulary; middle sessions: learners answer the protocol's questions in writing for two or three minutes before the group discussion, a learner leads one phase, and the facilitator adds only what was missed; late sessions: learners run the debrief, choose the decision points themselves and write their own questions; the facilitator checks the stated changes and corrects any wrong rule. Move a group to the next stage when, on two consecutive debriefs, learners find the main gap themselves before the facilitator names it. Untested proposal; guidance that helps novices can become redundant as expertise grows (Instructional guidance that helps novices can become redundant or counterproductive as expertise grows. [~M]), and the facilitator's role fades as in Scaffolding and Fading.

When a condition is not met: if no facilitator is available, use a written self-debrief with the same questions and a model performance to compare against (see the table); if learners will not speak about errors in front of this group, take the first answers in writing and privately; if there will be no next attempt, keep steps 4, 5 and 8 and drop the rest.

Fitting the design to a situation

The three facts that most change the decision:

  • The learners' expertise on this task and whether they can judge their own performance. Novices miss the cues that matter and misjudge their own performance; the facilitator must lead, focus on process and supply the criterion. Advanced learners can run the debrief. If unstated, ask: have they done this task before, and can they say, before the debrief, which decision mattered most and whether they got it right?
  • Who will lead the debrief, for how long, and for how many learners. The claims behind the default rest on someone who can ask specific questions, respond to answers and give task-level feedback; a facilitator with five minutes for thirty learners, a volunteer, or no one at all changes the design. If unstated, ask: who is with the learners straight after, how long do they have, and how many learners per leader?
  • Whether a next attempt follows, and what counts as success. The debrief is expected to pay off through a changed next performance. If unstated, ask: when will learners next do something like this, and what observable change would show the debrief worked?
If the brief says… Then change… Basis
Learners are novice on the task (first simulations, first placements) Facilitator-led throughout; give the criterion or a model performance before analysis; ask about cues and steps, not outcomes; name what the expert would have noticed; one change each, stated as an action Process goals lead to better skill acquisition for novices than outcome goals. [+M] (high-school writing revision, carried to debrief focus); Self Assessment Accuracy Is Low Without Training [~M]
Learners are advanced, or experienced practitioners Learners choose the decision points and lead; the facilitator asks for alternatives considered and the reasoning behind them, and adds only what was missed; keep it shorter Instructional guidance that helps novices can become redundant or counterproductive as expertise grows. [~M] (cognitive-load studies; no study of debriefs)
Learners are children (primary age) Five minutes, oral, teacher-led, three questions (what happened, what was tricky, what we will do next time), straight after the activity; one class change written on the board and read out next time untested proposal
adult learners in a work setting, after a real shift, project or incident Use an after-action format (what was planned, what happened, why the difference, what to keep and what to change); keep it separate from performance appraisal and from any incident investigation; write the changes into the team's next briefing untested proposal; from the earlier page's examples
Goal is a procedure or psychomotor skill in workplace-clinical training Treat the debrief as the feedback step of a repeated attempt: debrief against a stated mastery standard, then rerun the scenario or a changed one until the standard is met Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M] (14 studies, health professions); Simulation Based Education Improves Outcomes [~M]: could not isolate feedback as the active feature
Goal is judgement, a concept or a principle (diagnostic reasoning, a negotiation, an ethical decision) Spend most of the time on the learners' reasoning at two or three decision points; end with the general rule stated in the learner's words and tested on a new case Prompting learners to self-explain… [+M] (immediate tests, study of examples and text; carried to debriefs)
The experience was emotionally charged or high-stress Lengthen the reactions phase; check everyone is settled before analysis; move personal distress to a separate, private conversation; never analyse the performance of someone visibly upset in front of the group untested proposal; from the earlier page's constraints
Some learners join remotely while others are in the room, or some are on phones only Everyone writes their first answers (what happened, one decision point) in a shared chat or document before talking; the facilitator calls on remote learners first at each phase; share the recording or the timeline on screen; phone-only learners send a voice note or a two-line text untested proposal
online-self-paced, no facilitator Replace the live debrief with a written self-debrief: three to five specific questions tied to the criteria, then a model performance or expert commentary to compare with, then one change for the next module, shown back at its start Reflective Practice Improves Outcomes When Structured [+M] (classroom engineering course); Metacognitive prompts improve learning [+M] (computer-based environments)
One facilitator for a large group Debrief in threes or fours against a written protocol for 10 minutes; then the whole group on the most common gap, with the facilitator's explanation after the groups have reported Peer Discussion Improves Conceptual Understanding [+M] (peers then instructor; undergraduate classes); peers who share a misconception may reinforce it, untested
Scarce facilitator time (volunteers, a busy shift, a supervisor who can give five minutes) A three-question debrief straight after, in five minutes: what happened at the key moment, one thing to keep, one change; the leader corrects any wrong rule in one sentence; changes go in a shared log; a full debrief only for critical or new events untested proposal
One session only (single-session) One short experience, a debrief of at least equal length, then a two-minute retry of the key moment with the change; do not claim retention untested proposal
The brief names its own debrief model (plus/delta, PEARLS, GAS, an after-action review, the "What? So what? Now what?" questions) Keep it. Check only that it reaches the learners' reasoning at specific decision points, gives task- and process-level feedback against a stated criterion, and ends with one written change for a next attempt; add whichever of those is missing as one question Reflective Practice Improves Outcomes When Structured [+M]; Feedback Most Effective At Task And Process Levels [+S] (neither compares debrief models)
The performance is assessed or high-stakes, or the team must keep working together Separate the formative debrief from grading in time and by person where possible; talk about actions and decisions, never traits; record what was agreed, not who erred Feedback Most Effective At Task And Process Levels [+S] (feedback aimed at the self lost effect); separation from grading untested
The product or performance was for a real audience (a public performance, a client deliverable, a community event) Debrief the team privately before any public review; use audience responses as evidence in the debrief, not as its verdict; keep the analysis out of anything the audience sees untested proposal
Learners have low confidence or literacy, or speak in a second language Take first answers privately, orally or drawn, with sentence frames ("I saw…", "I decided… because…"); in the group, ask for actions, not explanations, from anyone not ready; judge the analysis, not the language untested proposal

Eight of the sixteen rows rest on a claim, and none of those claims studied a debrief: each carries a finding from tutoring, writing-to-learn, classroom discussion, feedback or simulation studies to this setting and says so. The rows for children, workplace incidents, stressful events, mixed or remote modes, scarce facilitator time, single sessions, real audiences and low-confidence learners are untested proposals.

Observation, state and explanation

What can be observed in and after a debrief: what learners say happened, which decision points they pick, the reasons they give, their stated confidence, the corrections given and by whom, the change each names, how much of the time the facilitator spoke, and what each learner does at the next attempt. "Understood the lesson", "reflective" and "psychologically safe" are inferred states. A fluent, agreeable debrief does not show a changed next performance.

Response Plausible explanations Observation that could change the interpretation
Learners narrate the sequence and name no reason or change The question invited narration; learners lack the concept to interpret the event; reluctance to discuss errors in front of others Ask the decision-point question in writing, privately ("at minute 6, what did you see and decide, and why?"); if reasons appear in writing, the setting was the barrier; if not, teach the concept.
Learners say "it went fine" about a performance that missed the criterion They cannot judge the performance; they judge it on a different criterion; they are protecting themselves or a colleague Show the recording or the log at the key moment and ask them to score it against the written criterion; compare with an observer's score.
A learner defends a wrong decision with confidence A misconception; the learner saw a cue others did not; the criterion is contested in practice Ask what they were seeing; check the recording; if the decision was wrong, correct it plainly and test it on a new case in the next session.
States the right change, repeats the error at the next attempt The change was held as words, not practised; the next situation differed; load or stress on the next attempt Rerun just the key moment with the change, with the same load; record whether the change was named at the start of the attempt.
One person dominates, or the facilitator talks most of the time Status differences in the group; the facilitator explaining before learners have tried; time pressure Time each speaker for one debrief; take first answers in writing; ask the quietest person first at each phase.
Learners are silent, or blame others or the scenario Low psychological safety; the scenario was unrealistic or unfair; fatigue or distress Ask privately afterwards what made it hard to speak; check the scenario against the criteria; restate the ground rules and lengthen the reactions phase next time.

These rows are proposals, untested with learners as diagnostics. A probe asked in the debrief is itself feedback and changes the state it measures, so record each one, and keep the explanations that remain open rather than choosing the one whose name fits.

Evidence and qualifications

No claim in the wiki compares an experience with and without a debrief, compares debriefing methods, or sets a debrief's length or timing. The claims below test ingredients of the default design in other settings.

  • Prompting learners to self-explain improves understanding and problem solving on immediate tests by a moderate average amount, with little evidence yet on delayed or classroom gains. [+S]: across 69 effect sizes, self-explanation prompts while studying or solving problems improved learning (g = .55); in mathematics the gains are on immediate tests, and evidence for classrooms or after a delay is much more limited. It supports asking learners to explain their decisions; it was tested on studying examples and text, not on explaining one's own past actions, and the claim's Discussion warns that learners without the background can explain wrongly and reinforce a misconception, which is why the facilitator corrects.
  • High-confidence errors are more likely than low-confidence errors to be corrected on a later test after feedback. [+S]: a narrative review and an experiment report that errors held with high confidence are more often corrected and remembered when corrective feedback follows; no effect size is recorded. It supports staying on confidently held mistakes. The evidence is about answers to factual questions with immediate correction; its Discussion says the effect depends on clear, non-punitive feedback, and that without feedback errors can be reinforced.
  • Contingent scaffolding improves learning more than fixed or absent support. [+M]: one-to-one tutoring studies (fourth and fifth graders on long division, 8 per condition; eighth graders with tutors who prompted rather than explained, who did better on transfer) and a review that finds most evidence one-to-one; a meta-analysis of dynamic assessment ranked explicit strategy training above scaffolding. It supports questioning that follows what learners noticed; group debriefs are untested, and for a structured skill explicit teaching may do better.
  • Process goals lead to better skill acquisition for novices than outcome goals. [+M]: high-school students learning writing revision with process goals outperformed those with outcome goals, and a theoretical chapter synthesises similar findings; no effect size is recorded. It supports a process focus in novice debriefs; it is about goals set during practice, carried here to the focus of the debrief.
  • Reflective Practice Improves Outcomes When Structured [+M]: in 48 school writing-to-learn studies, metacognitive-reflection prompts were the only content type associated with larger effects (b = 0.48); 208 first-year engineering students given specific prompts scored higher on exams and projects than those given generic prompts, with 52 reflections apiece in both groups (quasi-experimental, no effect size). It supports specific, criterion-tied questions; neither study is a debrief.
  • Feedback Most Effective At Task And Process Levels [+S]: high-information feedback on task and process gave d = 0.99 against d = 0.24 for reinforcement or punishment across 435 education studies, and feedback lost effect as it drew attention to the self in 607 effect sizes across settings; both support the claim indirectly, since neither coded the four levels directly. It supports the content of the facilitator's feedback.
  • Peer Discussion Improves Conceptual Understanding [+M]: in undergraduate genetics classes, peer discussion alone and instructor explanation alone gave similar gains, and the two combined gave larger gains for all students (no randomised control). It supports peers first, then the facilitator.
  • Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M]: 14 studies, mostly procedural and psychomotor skills, found simulation with deliberate practice (repetition, mastery standards, immediate feedback) better than traditional clinical education (d = 0.71). The debrief is part of the package, not tested on its own; the claim's Discussion calls debriefing quality "a key moderator" without recorded evidence for it.

What limits and qualifies the model:

Learners, tasks, settings and outcomes differ across these claims (children with a tutor, engineering undergraduates, health-professions trainees, laboratory recall), so do not rank them by effect size. None measures delayed retention or workplace performance after a debrief.

The evidence by question

Further evidence, not yet read against this model

Claims about debriefs or group reflection that the earlier page did not cite, and that this model does not rest on. Each says what it is and how far it bears on this page.

Objective and learner-valued goal

The designer's objective is usually a changed next performance: a procedure done to standard, a decision made on the right cue, a team that communicates at the moment it matters, at a stated horizon. Elicit separately what the learners want from the debrief: to know whether they "passed", to avoid looking incompetent in front of peers or a senior colleague, to understand a decision that puzzled them, to be reassured after a stressful event. Agreement and divergence are both observations. A learner who wants reassurance may agree quickly with every point and change nothing; one who wants to understand a puzzling moment will steer the debrief to it, and that may be the best decision point to analyse.

For example, a newly qualified nurse after a simulated deterioration may most want to know whether calling the senior doctor early would be seen as weakness on the ward, while the designer's objective is recognising early warning signs. Debriefing the recognition step serves the objective; naming the escalation question and answering it (what the ward expects, what the criterion says) serves the learner's. Record both, so that a correct account of the warning signs is not read as the confidence to escalate.

What would revise this model?

The expectation should weaken if studies comparing the same experience with and without a structured debrief, with comparable learners and next-performance outcomes, find no advantage, or an advantage only on self-report. If the gain from simulation with deliberate practice is carried by repetition to a standard and immediate feedback during practice, and a post-event debrief adds nothing beyond them, the debrief should be cut to a short correction and the time given to another attempt. If learner-led and facilitator-led debriefs give the same next performance for novices, the facilitator's role can be reduced earlier. If peer discussion in debriefs spreads a shared misconception more often than it corrects one, the facilitator should explain before peers, not after. Do not protect the model by calling every failure "a poorly facilitated debrief" after the fact.

A good debrief conversation, a correct stated change, the next performance, delayed retention and use in the real setting are separate claims. The present evidence establishes no debrief length, no best debriefing model and no timing rule.

Claims

  • Purposeful Reflection — debriefing is one of the most structured forms of reflection after action
  • Error Analysis — strong debriefs convert mistakes into inspectable evidence
  • Check-ins — emotionally charged debriefs may need brief regulation or climate-setting before analytic discussion
  • Explaining Their Thinking — debriefing depends on making reasoning visible
  • Experiential Learning — the debrief is the reflection step that turns an experience into a stated principle and a changed attempt
  • Scaffolding and Fading — how the facilitator's lead is handed to learners across sessions

Examples

Key Sources

  • Cheng, A., Grant, V., Dieckmann, P., Arora, S., Robinson, T., & Eppich, W. (2015). Faculty development for simulation programs. Simulation in Healthcare, 10(4), 217-222. doi:10.1097/SIH.0000000000000090
  • Paige, J. T., Arora, S., Fernandez, G., & Seymour, N. (2015). Debriefing 101. The American Journal of Surgery, 209(1), 126-131. doi:10.1016/j.amjsurg.2014.05.034
  • Secheresse, T., Lima, L., & Pansu, P. (2021). Focusing on explicit debriefing for novice learners in healthcare simulations. Nurse Education in Practice, 51. doi:10.1016/j.nepr.2020.102914
  • INACSL Standards Committee. (2016). INACSL standards of best practice: SimulationSM debriefing. Clinical Simulation in Nursing, 12, S21-S25. doi:10.1016/j.ecns.2016.09.008
Page metadata
Field Value
type principle
id debriefing
title Debriefing
description For learners who have just shared an experience (a simulation, case, team exercise or real event), a facilitated debrief that establishes what happened, asks for their reasoning at a few decision points against stated criteria, gives task- and process-level feedback and ends with one change to try is expected to improve the next comparable performance more than the experience alone or an unstructured recap, though no claim here compares an experience with and without a debrief.
status review
canonical True
generated by claude/unspecified on 2026-10-05
trust unverified — no one has confirmed this page's content
sources 3 — cheng-2015, paige-2015, secheresse-2021

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