Skip to content

Gagné's 9 Events of Instruction

Pattern · All patterns
Evidence · 13 claims (4 for, 8 mixed, 1 against) · 29 studies (10 causal, 8 review, 6 quant-synthesis, 3 theoretical, 1 associational, 1 qualitative), q1–q4 · 9 of 29 report an effect size · 6 claims rest on one study

Description and scope

A reusable policy for one bounded lesson on a structured target (a concept, procedure or rule with a defined correct performance): direct attention, say what the learner will be able to do, surface relevant prior knowledge, present the content, guide the first attempts, elicit performance, give feedback, check performance, and set up retention and transfer. It instantiates the Direct Instruction principle: explicit explanation, modelling and guided practice with checks for a learner whose response shows no working method for the target. Two other converted principles govern single events inside it: Activation for event 3 (stimulate recall) and Retrieval Practice for events 6 and 9 when the target includes later recall. The Direct Instruction pattern is a sibling policy built on the same principle; this page is the general nine-event lesson architecture, and treats each event as a candidate rather than a requirement.

The nine-event sequence itself is a design framework, not a tested intervention. No claim in this wiki tests the full nine-event lesson against a different lesson design. The nearest test is one experiment in which events were removed one at a time from a computer-based lesson (below): it found that removing practice lowered the posttest and removing objectives, examples or review individually did not. The study configurations below are evidence; the response-dependent policy on this page is an untested design proposal, and the default sequence keeps the earlier page's nine steps, each labelled with what supports it.

Inputs: establish the design brief

Field Record or elicit
Learner and starting observation A first attempt at a representative task before instruction, with the reasoning given, the help used and the time taken. Record task-specific expertise as observed (novice, intermediate, advanced, or unknown), age band and setting. Preserve unknowns; experience in the wider subject does not settle readiness for this task.
Target and outcome The capability the lesson is for and a representative task that shows it; which aids are permitted; the instrument, the locally chosen criterion and the horizon (immediate-performance, delayed-retention, near-transfer, far-transfer). Events 8 and 9 cannot be designed until this is stated.
Knowledge type and task structure Whether the target is a structured concept or procedure, or a complex-skill that has to be practised as a whole; whether the aim is open inquiry, in which case this pattern is likely the wrong frame.
Context and time Lesson length, delivery (classroom, online-self-paced, online-instructor-led), who can give feedback and how fast, and which events earlier or later lessons already carry (a review elsewhere, a later transfer task).
Learner-valued goal Ask what the learner wants from the lesson and why. Record disagreement with the designer objective; negotiate rather than assume agreement.
Designer objective What the lesson is meant to change and how it will be judged; whether the objective will be stated to the learner, and in what words.

A request for "a lesson using Gagné's events" does not specify these inputs. Ask for the target, the starting response and the horizon before deciding which events to elaborate, merge or omit.

Sequence and conditional policy

The default sequence keeps the earlier page's nine events; the label after each says what the evidence on this page bears on it.

  1. Gain attention. Open with a problem, case or question tied to the target. Untested here; from the earlier page.
  2. State objectives. Say what the learner will be able to do and how it will be checked. Removing objectives alone did not lower the posttest in the one experiment recorded (single-event removal [~M]); keep it short and decide by whether learners need it to direct their own work.
  3. Stimulate recall of prior knowledge. Ask for the relevant idea in a form the next instruction will use. Governed by the Activation principle, whose evidence is split; activation alone has often shown little effect.
  4. Present new content. Explain and model the target, with examples. Supported for learners who cannot yet start (Direct Instruction principle); removing examples alone did not lower the posttest in the one experiment (single-event removal [~M]).
  5. Provide learning guidance. Prompts, cues or a partly worked attempt while the learner works. Supported for learners without a working method; may become redundant as task expertise grows (expertise reversal [~M]).
  6. Elicit performance. The learner attempts representative tasks. The one event whose removal lowered the posttest (practice present [+M]).
  7. Provide feedback. Say what was right, what was not and what to do next. The practice versions in that experiment are titled "practice with feedback", so practice and feedback were not separated; see Formative Assessment.
  8. Assess performance. An unaided task matched to the target, under the agreed aids. A check, not a treatment: the evidence here does not test it as an event.
  9. Enhance retention and transfer. Later retrieval, a changed problem, or use in a fuller task. Untested as an event here. Removing a review alone did not lower the posttest (single-event removal [~M]), but that posttest's horizon is not reported, so it says nothing about delayed retention. For a complex skill, a design argument favours whole-task practice over a part-by-part sequence (whole-task practice [~M]).

Between events, choose the next activity provisionally from what the learner does:

Observation Discriminating probe Proposed action, with its uncertainty
Cannot begin the representative task before instruction Ask for a first step on the same relation in a familiar and in a changed representation, with an accessible response mode, recording any hint. If reasoning is absent, run events 4–5 in full. If only recognition or access fails, change the representation or the response mode first. Neither branch is a validated diagnosis.
Already performs the target unaided at the start A changed problem and a request to say why each step is licensed. Shorten or skip events 4–5 and move to performance, feedback and transfer; full presentation may be redundant for this learner on this task. One success does not establish the threshold.
Succeeds during guided practice, fails the unaided check (event 8) A matched item with the model removed but the prompt kept, then one with both removed. Fade guidance in steps and add practice with feedback before the check is repeated; do not repeat the whole presentation by default.
Unaided success at the end of the lesson, weaker on a later or changed task Re-test with a new item at the delay; record what practice and exposure happened in between. Add spaced retrieval or a transfer task as event 9 in a later session rather than lengthening this lesson. Whether a review event would have prevented the drop is untested.
Says the lesson's aims were clear, but cannot state what they were or use them Ask the learner to state the target in their own words and to judge a worked attempt against it. Treat attitude ratings as a separate outcome from learning: in the one experiment, the group with no objectives rated goal clarity highest. Revise the objective's wording or drop it, and judge by performance.

These rows are proposals for local observation, untested with learners. A probe is itself a learning exposure, and one contrast does not identify a cause.

Choosing configurations from evidence

  • Keep the performance event. In an experiment with 256 undergraduate computer-literacy students, randomly assigned within pretest blocks to six versions of a computer-based lesson (randomized, adult, postsecondary), the four versions that included practice scored significantly higher on the posttest than the two without (practice present [+M]). The claim page prints the omnibus F and group means (above 17 against below 15) but no standardised effect size, and does not report the posttest's horizon, what the feedback consisted of, or the lesson's length. One study, not yet checked against its source. It supports keeping practice; it does not say how much, or test the order of events.
  • Do not read single-event removal as permission to drop everything else. In the same experiment, versions without objectives, without examples or without review scored 17.16 to 17.36 against the full program's 17.61, a non-significant difference (single-event removal [~M]). The claim page says equivalence was not formally tested, so this is a failure to detect a difference, not evidence that those events do nothing. It concerns one well-designed lesson with the other events present, a horizon the claim page does not report, and one population. Not yet checked against its source.
  • Learners' ratings do not track which events they received. In the same study, the group with no objectives gave the highest ratings on the goal-clarity items, though not significantly higher than most groups (no-objectives attitudes [~M]), while groups without practice or without examples rated those items lower (missing practice and examples noticed [~M]). So learner satisfaction is a separate outcome (attitude-motivation), and a positive rating of the lesson is not a check that an event did its job. Both not yet checked against their source.
  • Fade guidance as the learner's task-specific competence grows. A review of cognitive-load studies argues that integrated explanations, worked examples and step-by-step guidance help novices by reducing search and can become redundant or depress performance for more knowledgeable learners (expertise reversal [~M]). Read from an abstract, no effect sizes recorded, not settled against its source. It qualifies events 4–5 for advanced learners; it does not say when to fade or how to measure the point.
  • For a complex skill, prefer whole tasks to a part-by-part pass. A design argument, not an experiment, holds that load-reducing methods suited to practising complex tasks to retention (low variability, complete guidance and feedback) can hinder transfer, and proposes high variability and limited guidance with simpler whole tasks early for novices (whole-task practice [~M]). It qualifies a single linear pass through the nine events for a complex-skill target with a transfer goal; it is not evidence of an effect.

Unguided discovery against explicit instruction, which bears on events 4–5 for novices, is read on the Direct Instruction principle and not repeated here. None of these entries tests the nine events as a sequence, their order, or event 1 or 9 as such. A precise abstention names the missing comparison, target or observation and the next way to resolve it.

Fitting the design to a situation

Ask for three facts when the brief omits them: what the learners can do on the target before the lesson, the outcome and the horizon at which it is checked, and the lesson's length and delivery.

If the brief says… Then change… Basis
Time is short and events must be cut Keep practice with feedback (event 6–7) above all; it was the one event whose removal lowered the posttest Practice raises computer-based lesson posttest scores [+M]
Objectives, examples or a closing review would be cut One of them may be trimmed if time forces it; removing any one alone did not significantly lower an immediate posttest, but that is not evidence they do nothing Removing single events did not reduce achievement [~M]
The lesson will be judged on learner satisfaction ratings Also check performance on an unaided task; ratings did not track which events learners received The no-objectives group rated goal clarity highest [~M]; Students notice missing practice and examples [~M]
Learners cannot begin the target task Run events 4 and 5 (presentation and guidance) in full Minimal guidance is less effective for novices [+S]
Some learners already perform the target unaided Shorten events 4 and 5 for them and move to performance, feedback and transfer Expertise reversal effect [~S]
Learners have little relevant prior knowledge for event 3 Replace open "what do you know" recall with a few questions on the coming content; open activation showed little effect Pretesting enhances learning [+M]; Prior-knowledge activation had little effect on retention [-M]
The target must be recalled weeks later Make event 9 retrieval practice in later sessions rather than a review at the end of this lesson Retrieval practice improves long-term retention [+S]
The target is a complex skill with a transfer goal Replace a single linear pass through the events with whole tasks of increasing complexity (see 4C/ID) Whole-task practice improves transfer (design argument) [~W]
online-self-paced, with learners directing their own work State the objective briefly in performance terms and let learners check their own attempt against it untested proposal

The evidence by question

Further evidence, not yet read against this model

Claims this page cited before it was rewritten as a conditional model. 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.

Illustrative design instance and observation record

Invented illustration, not a tested lesson. A 40-minute online module teaches adult warehouse staff to read a stock-rotation label and decide which pallet ships first. Before any instruction, each learner is shown two labels and asked which ships first and why. Most choose by the larger date number; two already apply the rule and are routed to three changed labels and the transfer task. For the rest, the module states the target ("choose the pallet to ship from any two labels"), asks what "best before" means to them, explains and models the rule on three labels, gives four guided items with a hint available, then four unaided items with immediate feedback, then a mixed check. One learner says he wants to stop being corrected by his supervisor; the designer's objective is zero rotation errors on the floor. A week later, a short set of new labels is sent to the same learners. These are local design choices; the evidence above does not establish their optimality.

Record: target and permitted aids → starting attempt and reasoning → which events were run, merged or skipped, and why → help used during guided items → unaided responses and feedback given → check result → candidate interpretations (rule learned, date heuristic still in use, label format unfamiliar) and their basis → next observation at the stated horizon. Preserve the learner's stated purpose and any change in it. An agent may summarize this record but must not fabricate missing inputs or assign numerical probabilities of mastery without a calibrated model.

Elements and limits

Attention, learning objectives, prior knowledge activation and activation, demonstration, guided practice, practice, provide feedback, assessment and transfer tasks. Principles the policy draws on: Direct Instruction, Activation, Retrieval Practice and Formative Assessment.

This pattern is scoped to a bounded lesson on a structured target. Open inquiry, design work, and complex skills that have to be practised as whole tasks need other configurations (see Merrill's First Principles of Instruction for a task-centred alternative). The policy supports observation and design reasoning; the effect of the full sequence, of event order, and of any single event other than practice remains to be tested with learners.

Examples

  • A short online lesson that activates prior knowledge, teaches a new concept, and checks application with feedback.
  • Instructor-led technical training that sequences demonstration, practice, feedback, and transfer tasks.

Key Sources

  • Gagne, R. M. (1985). The conditions of learning (4th ed.). Holt, Rinehart and Winston.
  • Gagne, R. M., Wager, W. W., Golas, K. C., & Keller, J. M. (2005). Principles of instructional design (5th ed.). Wadsworth.
Page metadata
Field Value
type pattern
id gagnes-9-events-of-instruction
aliases gagnés-9-events, gagnés-9-events-of-instruction
title Gagné's 9 Events of Instruction
description A reusable lesson policy that moves a learner from an observed starting response through presentation, guided and independent performance with feedback, to a check at a stated horizon, keeping or dropping each event according to what the learner's responses show rather than running all nine by rote.
status review
grain_size lesson
author Robert Gagné
canonical True
generated by claude/unspecified on 2026-10-02
trust unverified — no one has confirmed this page's content

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