Instructional design principles: definition, examples, and how to apply them

What are instructional design principles?

An instructional design principle tells you what a lesson must do for learning to happen, for example "give learners practice with feedback." Principles come from learning research, and they hold whatever tool or project model you use. The term is old enough to be a textbook title: Gagné, Briggs, and Wager's Principles of Instructional Design reached its fourth edition in 1992, and Northern Illinois University's guide to Gagné's nine events still cites it.

People often mix up principles, theories, and models. The difference matters when you choose what to read or which framework to adopt.

Principles, theories, and models in instructional design

Term	What it tells you	Examples
Principle	What effective instruction does	Align objectives and tests; give practice with feedback
Theory	Why it works	Cognitive load theory; dual coding theory
Process model	The steps a team follows to build a course	ADDIE, SAM, backward design
Lesson framework	The order of events inside one lesson	Gagné's nine events; Merrill's cycle of activation, demonstration, application, and integration

10 core principles of instructional design

These ten appear across the frameworks in this guide, from Merrill and Gagné to Rosenshine and cognitive load theory.

1. Start with an observable objective

Write what the learner will be able to do, under what conditions, and how well. Robert Mager popularized this three-part objective (behavior, condition, standard) in 1962, according to the EBSCO history of instructional design. Use a verb you can observe. You can test "identify the four shutdown conditions," but you cannot test "understand safety." The revised Bloom's taxonomy (Anderson and Krathwohl, 2001) gives six levels, from remembering to creating, with verbs for each level (NIU guide). Our free learning objective generator turns a topic into measurable objectives, each with a Bloom level and a matching assessment type.

2. Align practice and assessment with the objective

John Biggs called this constructive alignment (1996): the learning activities and the assessment ask learners to do what the objective says. Grant Wiggins and Jay McTighe built backward design on the same idea. In their Understanding by Design framework, what Stage 1 names as the goal is what Stage 2 assesses and Stage 3 teaches. To check alignment, read each objective next to its test item. If the objective says "troubleshoot" and the test asks for a definition, change one of them.

3. Analyze the learners and the task first

Find out what learners already know and what the job requires before you write content. Gagné introduced task analysis, which breaks each task down to its basic components. Prior knowledge changes the design: support that helps novices, such as worked examples, can stop helping more experienced learners or even slow them down. Slava Kalyuga and colleagues named this the expertise reversal effect (2003). A short pretest tells you which group you have.

4. Build the lesson around real tasks

Merrill puts the task-centered principle first: teach through whole, real-world tasks, arranged from simple to complex, and teach component skills inside those tasks. Merrill's 2007 paper on task-centered instruction shows a course built as a progression of five business cases, each more complex than the last. A customer service course can work the same way. It opens with a real ticket, moves to harder tickets in later modules, and brings in policy content when a ticket needs it.

5. Activate what learners already know

Start each lesson by connecting new material to something learners know. Gagné's third event is "stimulating recall of prior learning." Barak Rosenshine's first principle of instruction (American Educator, 2012) is to begin with a short review of previous learning. In the studies Rosenshine summarized, the most effective teachers opened lessons with a five- to eight-minute review.

6. Show the skill before you ask for practice

Learners need to see a task performed before they try it. Merrill calls this demonstration, and Rosenshine lists it as "provide models." A worked example counts, and so does a screen recording with each step called out. In a 2006 review titled "Why Minimal Guidance During Instruction Does Not Work", Paul Kirschner, John Sweller, and Richard Clark concluded that minimally guided instruction is less effective and less efficient than guided instruction. The advantage of guidance shrinks only when learners have high prior knowledge.

7. Teach in small steps and cut avoidable load

John Sweller introduced cognitive load theory in the 1980s, starting with a 1988 paper on problem solving. Its core claim, as Sweller, van Merriënboer, and Paas restate it in their 2019 review, is that all new information passes first through a working memory with limited capacity and duration. A design that makes learners spend that capacity on searching for a label or reading the same words twice leaves less of it for learning. Rosenshine's second principle is to present new material in small steps, with practice after each step. In practice, that means one idea per screen or video segment, labels placed on the diagram instead of in a legend, and no on-screen text that repeats the narration word for word. The demo above tests this principle. Richard Mayer's research turns it into specific rules for slides and video, which we cover in our guide to multimedia learning principles.

8. Make learners recall and apply, with feedback

Learners remember more when they pull information from memory and use it. Henry Roediger and Jeffrey Karpicke found that students who took recall tests on a passage retained more after two days and after one week than students who studied it again, even though restudying made students more confident (2006). Feedback is part of the practice. John Hattie and Helen Timperley's 2007 review of feedback found it among the major influences on learning, with effects that depend on the type of feedback and how it is given. Rosenshine reports that the best success rate during guided practice appears to be about 80% of answers correct, which shows that learners are learning and still challenged. A question that nearly every learner gets right may only check attention.

9. Scaffold, then fade the support

Give learners supports such as checklists and partly worked examples, and remove them as skill grows. Rosenshine lists "provide scaffolds for difficult tasks" and "require and monitor independent practice" as separate principles, in that order. The expertise reversal effect explains why you fade: a support that helps a novice can get in the way of someone who no longer needs it.

10. Space the practice and plan for transfer

Spread practice and review over time, and end with use on the job. In a 2013 review of ten study techniques, John Dunlosky and colleagues gave a high utility rating to only two: practice testing and distributed practice. Merrill's integration principle and Gagné's ninth event, enhancing retention and transfer, both ask learners to use the skill in their own work. Rosenshine adds weekly and monthly review. For a schedule you can use, see our guide to knowledge retention strategies.

Merrill's 5 principles of instruction

The best-known set of five instructional design principles is M. David Merrill's first principles of instruction. Merrill reviewed many instructional design theories and models and kept the principles they share. He published them in Educational Technology Research and Development in 2002:

1. Task-centered: learners work on real-world problems or tasks.
2. Activation: learners recall relevant knowledge or experience as a base for the new material.
3. Demonstration: learners watch the new skill performed on specific examples.
4. Application: learners use the new skill on further examples and get feedback.
5. Integration: learners carry the skill into their own work and reflect on it.

Merrill describes the last four as a cycle that runs inside each real-world task. A lesson that skips demonstration or application usually shows it: learners can repeat the content but cannot do the task.

What should a practice exercise for a principle ask learners to do?

Here, "principle" means a cause-and-effect rule in the subject matter, such as "a clogged filter lowers the pressure after it." Merrill's 2007 paper calls this "what-happens" knowledge. To practice it, give learners the conditions and ask them to predict the consequence, or give them a consequence and ask them to find the missing or faulted conditions. A question that asks learners to state the rule tests only whether they remember it.

Gagné's nine events of instruction

Gagné described nine events in The Conditions of Learning (1965). They give the order of one lesson, and the table shows each event in a short training video.

Gagné's nine events of instruction, with a video lesson example for each

Event	What you do	In a video lesson
1. Gain attention	Open with a problem, a question, or a result	Show the mistake the lesson prevents
2. Inform learners of objectives	Say what they will be able to do	One line on screen before the content
3. Stimulate recall of prior learning	Connect to what they know	A short question on the last module
4. Present the content	Teach in small steps, with examples	Short segments with labeled visuals
5. Provide learning guidance	Give cues, examples, and memory aids	A worked example with callouts
6. Elicit performance	Have learners practice	An in-video question or decision
7. Provide feedback	Say what was right or wrong, and why	Feedback for each answer, or a jump back to the explanation
8. Assess performance	Check against the objective	Scored questions or a final scenario
9. Enhance retention and transfer	Space review and connect to the job	A follow-up check a week later and a job aid

Principles, facets, and events: how the frameworks fit together

You will often use several frameworks on one project. A process model runs the project, a planning model sets the goals and the evidence, and principles and events shape each lesson.

Wiggins and McTighe's backward design has three stages: identify desired results, determine assessment evidence, and plan learning experiences and instruction. For Stage 2 they define six facets of understanding. A learner who understands can explain, interpret, apply, shift perspective, empathize, and show self-knowledge. They say a single assessment does not need all six; in math, explanation, interpretation, and application are the natural ones.

ADDIE comes from the Interservice Procedures for Instructional Systems Development, which Florida State University's Center for Educational Technology wrote for the US armed services in 1975. Its five phases were Analyze, Design, Develop, Implement, and Control. Today the model is taught with Evaluate as the fifth phase. SAM, the Successive Approximation Model from Michael Allen, replaces ADDIE's long phases with short cycles of design, prototype, and review. It starts with a one- to two-day "Savvy Start" and moves through alpha, beta, and gold releases (Allen Interactions).

Common instructional design frameworks and when to use each

Framework	Type	Use it for
ADDIE	Process model	Large projects with formal reviews and sign-off at each phase
SAM	Iterative process model	Projects where stakeholders need to see a working prototype early
Backward design	Planning model	Courses where the assessment must show transfer to new situations
Merrill's first principles	Lesson principles	Checking that a lesson has a real task, a demonstration, and practice
Gagné's nine events	Lesson sequence	Putting the parts of one lesson in order
Rosenshine's principles	Teaching practices	Live and virtual instructor-led sessions
Cognitive load theory	Learning theory	Deciding what to cut from a screen, a slide, or a step

Where instructional design principles came from

The field took shape in World War II, when the US military had to train personnel quickly and efficiently (EBSCO Research Starters: history of instructional design). After the war, psychologists who had worked on that training began to treat training as a system, with procedures for analysis, design, and evaluation. The lasting idea is analysis: break a job into smaller tasks, make each task its own learning goal, check performance, and correct errors before the learner moves on. Gagné, who tested aviation cadets for the Army Air Forces during the war, later formalized this as task analysis and learning hierarchies, in which learners master prerequisite skills before the skills that depend on them.

The milestones after that, as Robert Reiser traces them in his 2001 history of instructional design:

• 1950s: B. F. Skinner's programmed instruction used small steps, frequent questions, immediate feedback, and self-pacing (see Skinner's 1958 article "Teaching Machines"). Benjamin Bloom's committee published its taxonomy of educational objectives in 1956.
• 1960s: Mager's behavioral objectives (1962) and Gagné's The Conditions of Learning (1965), with the nine events.
• 1970s: systems models multiplied in the military, universities, and industry, including the 1975 interservice model that became ADDIE.
• 1980s to 2000s: designers drew more on cognitive psychology. Sweller's cognitive load theory, Mayer's research on multimedia learning, and Merrill's first principles (2002) come from this period.

How to apply instructional design principles to course creation

Use this sequence for a new course or module. It applies the principles above in the order a project needs them.

1. Define the performance problem. Write down what people should do differently on the job and how you will know they do it.
2. Analyze the learners and the task. List the steps of the task and the decisions inside it, then mark what learners already know.
3. Write observable objectives, one per task or decision.
4. Design the assessment next. For each objective, write the question, scenario, or observed task that will prove it. Our free course outline generator works backward from the outcome you state and marks assessment points for each module.
5. Choose real tasks and order them from simple to complex.
6. Plan each lesson with Gagné's nine events. At minimum, include a review, a demonstration, practice with feedback, and a check.
7. Build the materials and cut what the objective does not need.
8. Test a draft with a few learners from the real audience, watch where they hesitate, and revise.
9. After launch, compare assessment results and job results with the objectives. Our guide on how to measure training effectiveness covers the methods.

At step 7 you also decide how long each video should be. Length alone does not settle it. Across 8.07 million views of 18,315 Mindstamp videos (2018 to October 2026), videos of 10 to 20 minutes averaged 53.8% of each video watched, and videos of 2 to 5 minutes averaged 39.2%. The longer group also averaged 6.70 viewer actions per view, against 1.16 for the shorter group. Many factors differ between these videos, so treat this as a pattern, not proof. Divide a lesson where the content divides, and put practice at each break. Our post on microlearning strategies covers when short modules work best.

Applying the principles in video lessons

A linear training video covers events 1 through 5 at most: it gets attention, states the goal, and presents content. Practice and feedback, events 6 and 7, happen somewhere else or not at all. We built Mindstamp to put them inside the video, with one setting for each principle:

• Practice and retrieval: add questions at the point where a concept is taught. A question pauses the video until the learner answers.
• Feedback: give each answer its own response. A Change Time action can send a wrong answer back to the explanation, so the learner watches that part again and meets the question again (quiz scores and retries).
• Application in context: use video branching to let learners make a decision and see its consequence. Our guide to scenario-based training shows how to write these.
• Required practice: on supported plans, Prevent passing required interactions stops learners from seeking past a required question (prevent skipping ahead).
• Assessment and records: each answer is stored against the learner. A SCORM 1.2 or 2004 package reports completion and score to a compatible LMS (SCORM export).

Place the first question soon after the first new concept, and one after each later segment, in line with Rosenshine's "practice after each step." Make most questions ones a prepared learner can answer, so the success rate stays near the level Rosenshine's research recommends.

Common questions

What is the most important instructional design principle?

Alignment comes closest. If the objective, the practice, and the assessment ask for the same performance, most other problems are easy to fix. If they do not match, a well-produced course still trains the wrong thing.

Are instructional design principles the same as best practices?

Principles are the research-based rules. Best practices are the habits that teams use to apply them in projects, such as storyboards and pilot tests. For the corporate training side, read our instructional design best practices.

Should you design for learning styles?

The evidence does not support it. A 2008 review by Harold Pashler and colleagues found virtually no evidence that matching instruction to a learner's preferred style, such as visual or auditory, improves learning. Match the format to the content instead: show a spatial process with a diagram, and teach a conversation with dialogue. Our post on visual learning strategies shows how to use visuals well.

Can AI apply these principles for you?

AI tools can draft objectives, outlines, and questions quickly. Mindstamp, for example, can generate multiple-choice questions from a video transcript. You still check the result against the principles. A generated question may test recall when the objective asks for a decision, and only you know whether the task matches the job.

To start, take one existing lesson and check it against principles 2, 6, and 8: does the practice match the objective, does the lesson show the skill before asking for it, and do learners recall and apply it with feedback? Fix the first gap you find, then test the lesson with a few learners. For more ways to keep learners doing the thinking, see our guide to active learning strategies.