How to use self-explanation is to make your reasoning explicit while learning: explain why a step follows, what principle justifies it, how the example connects to the rule, and what would change if one condition were different.

Self-explanation is especially useful with worked examples and problem solving because it turns a visible solution into a reasoning task. Instead of simply following each line, the learner asks why that line exists, what rule controls it and how the same principle could appear in a different problem.

This guide explains how to self-explain without simply narrating obvious actions, how to use prompts, how to compare correct and incorrect solutions, and how to transition from worked examples toward independent performance.

This guide answers the practical question how to use self-explanation as a learning strategy for students, parents and educators. The method is treated as an operating system: define the learning job, create an independent attempt, compare against a reliable standard, repair the first meaningful weakness, and return later to test whether the improvement survives without the original support.

The 60-Second Answer

Study one step or idea. Cover the explanation. State what the step is doing, why it is valid, which principle supports it and how it connects to the previous step. Check accuracy. Then apply the same principle to a fresh case without the model.

Definition: Self-explanation is the learner’s deliberate generation of explanations that connect a solution, example or new idea to underlying principles, prior knowledge and reasons.

Why Self-Explanation Matters

Worked examples can be deceptively easy to follow because the correct route is already visible. Self-explanation requires the learner to reconstruct why the route makes sense.

This can expose missing principles. A student may know what step comes next but be unable to explain why it is allowed or when the same step should be used.

The method is strongest when explanations are principle-based rather than merely descriptive. ‘I divided both sides by three because I need to isolate x and equality is preserved by applying the same operation to both sides’ is more useful than ‘then I divided by three’.

The useful question is not whether a student has heard of self-explanation. It is whether the student can use it at the right time, on the right material, with enough independence that the method produces stronger later performance rather than only a feeling of activity.

The Operating Model

Observe → Explain → Justify → Connect → Check → Fade Support → Solve → Transfer.

Each stage has a separate purpose. Preparation defines the target. The learning action forces a decision. Feedback reveals what the learner misunderstood or omitted. Correction changes that specific weakness. Delayed return tests memory. Transfer checks whether the principle survives changed wording, examples, representations and time pressure.

How to use self-explanation: Step by Step

1. Choose a meaningful step or idea

Do not self-explain every trivial motion. Focus on decisions, principles, transitions and places where learners commonly imitate without understanding.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

2. Ask what the step accomplishes

State the immediate purpose of the step in the solution or explanation.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

3. Ask why the step is valid

Name the rule, theorem, causal principle, grammatical relation or evidence standard that justifies it.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

4. Connect to prior knowledge

Identify the prerequisite idea that makes the step intelligible.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

5. Compare with an alternative

Ask what other move might look plausible and why it would be weaker or wrong.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

6. Check the explanation

Use a reliable source or teacher to correct misconceptions in the self-generated reasoning.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

7. Fade the worked example

Hide or remove later steps so the learner must complete more of the process independently.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

8. Transfer to a fresh problem

Use the same principle in a new case and self-explain the method choice before solving.

Make the decision visible. Ask the student to state what they are trying to do, what evidence will show success, and what a plausible wrong version would look like. A study technique becomes teachable when the learner can describe the controlling decision rather than merely imitate the surface form.

What Strong Performance Looks Like

• Explanations include reasons.
• Principles are named.
• Steps connect to prior knowledge.
• Plausible alternatives are considered.
• Worked examples are gradually faded.
• Learners solve fresh problems independently.
• Wrong solutions are analysed.
• Explanations become shorter as expertise increases.
• Transfer improves.

These are behaviours rather than labels. They can be observed, practised and improved. That distinction matters because the aim is not to classify students as naturally good or bad learners; it is to improve the sequence of decisions that produces learning.

Common Failure Modes

1. Narration without reasoning

The student says what they are doing but not why it is valid.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

2. Parroting the model

The explanation repeats teacher language while the learner cannot generate it independently.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

3. Explaining every trivial step

The process becomes exhausting and attention is wasted on actions that no longer require analysis.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

4. No accuracy check

A misconception is explained confidently and repeatedly.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

5. Worked-example dependence

The learner can explain visible steps but cannot choose a method on a blank problem.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

6. No fading

Fully worked solutions remain present indefinitely, preventing independent completion.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

7. No incorrect examples

Students only explain correct work and never practise diagnosing why a plausible solution fails.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

8. No transfer

Self-explanation stays tied to one example instead of the underlying principle.

Repair: isolate one example, identify the missing decision, require an independent attempt before revealing the model, and then use a changed example. Finish with a delayed return so the learner has to reconstruct the principle rather than recognise the correction beside the answer.

How the Method Changes Across Subjects

Mathematics

Explain why each transformation is valid, why the method was chosen, what clue in the question supports the choice, and how to verify the result.

The subject boundary matters. A general learning strategy can organise study, but subject knowledge still determines what counts as relevant evidence, a valid step, a complete explanation or a precise answer. The strategy should make disciplinary thinking more visible, not replace it.

Science

Explain how a causal mechanism produces the observed result, why a variable change matters and how evidence supports the claim.

The subject boundary matters. A general learning strategy can organise study, but subject knowledge still determines what counts as relevant evidence, a valid step, a complete explanation or a precise answer. The strategy should make disciplinary thinking more visible, not replace it.

English

Explain why a quotation supports an inference, why a grammatical choice is appropriate, why one word is more precise and how a paragraph advances the task.

The subject boundary matters. A general learning strategy can organise study, but subject knowledge still determines what counts as relevant evidence, a valid step, a complete explanation or a precise answer. The strategy should make disciplinary thinking more visible, not replace it.

Humanities

Explain why evidence supports a claim, why one cause is more significant under a stated criterion and how alternative interpretations differ.

The subject boundary matters. A general learning strategy can organise study, but subject knowledge still determines what counts as relevant evidence, a valid step, a complete explanation or a precise answer. The strategy should make disciplinary thinking more visible, not replace it.

Primary School, Secondary School and Examination Years

Primary school

Use short prompts such as ‘Why did you do that?’, ‘How do you know?’, and ‘What rule are you using?’ Keep self-explanation focused on one important decision at a time.

Keep the routine small enough to repeat. Adults can model and prompt at the beginning, but the child should perform an increasing share of the deciding, explaining and checking.

Secondary school

Students can annotate worked examples with reasons, explain incorrect solutions and gradually take over more steps. They should learn to distinguish explanation from simple narration.

Secondary learners should begin carrying more of the planning, monitoring and correction themselves. The method becomes especially valuable when subjects are dense enough that rereading or copying can no longer keep pace with assessment demands.

Before major examinations

Use self-explanation during learning and correction, then compress it during timed performance. The student should internalise the reasoning enough that full verbal narration is no longer necessary for every familiar step.

Near examinations, reduce novelty in the study system and increase authenticity in the tasks. Use the established method on mixed questions, timed work and past papers, then repair the small number of weaknesses with meaningful mark return.

A Focused 60-Minute Practice Session

• 10 minutes — retrieve the principle behind self-explanation without notes.
• 10 minutes — inspect one strong model and identify the decisions that make it effective.
• 15 minutes — complete a guided task with targeted prompts rather than full answers.
• 15 minutes — complete a fresh task independently with the model removed.
• 5 minutes — compare against a reliable source and identify the first meaningful weakness.
• 5 minutes — decide the next practice task and when it will return.

The exact timing can change. The sequence matters more than the clock: independent effort, feedback, targeted repair and later return. Without those elements, a session can feel productive while providing little evidence that learning will survive tomorrow.

Diagnostic Checklist

• Can the student explain why a step is valid?
• Can they name the underlying principle?
• Can they connect the step to prior knowledge?
• Can they explain a plausible wrong alternative?
• Can they check the explanation accurately?
• Can support be faded?
• Can the student solve a fresh problem?
• Can they analyse an incorrect worked example?
• Does explanation become more efficient with expertise?
• Does method selection improve?

Use the checklist to decide what to teach next. If several items are weak, begin with the earliest failure in the process or the one that causes the largest downstream cost.

Practice Laboratory

Practice 1: Why-this-step

Pause at each important worked-example step and explain why it belongs.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 2: Cover-the-next-line

Hide the next step and predict it before revealing the model.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 3: Principle label

Write the theorem, rule or concept beside the step that uses it.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 4: Wrong-solution analysis

Inspect a clearly marked incorrect example and identify the first invalid step and why it fails.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 5: Alternative-method compare

Explain two possible approaches and why one is more efficient or appropriate.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 6: Faded example

Study a solution with progressively more steps removed and complete the missing reasoning.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 7: Think-aloud problem

Solve one fresh question while verbalising method choice and key checks.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 8: Teach-back

Explain a worked example to another student without reading the annotations.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 9: Boundary question

Change one condition in the problem and explain which steps would need to change.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Practice 10: Delayed self-explanation

Return after several days and reconstruct the reasoning without the original example visible.

After the attempt, ask the learner to name the rule or relationship that controlled the answer. Record whether the performance was independent, prompted or modelled. That distinction is more informative than a simple correct/incorrect mark because it shows how much of the thinking the learner currently owns.

Three Student Cases

Case 1: Maren follows worked examples

Maren can complete exercises when a model is nearby but stalls when the question format changes.

Her tutor makes her explain why each major line follows and gradually removes later worked steps.

She begins recognising the principle rather than the visual pattern of the sample solution.

The case is useful because it shows a process change: diagnosis → targeted intervention → independent retest → delayed verification. A good teaching case explains which decision changed, not merely that the student later performed better.

Case 2: Iona narrates only

Iona says, ‘then I multiply, then I simplify’, but cannot say why those operations are justified.

Prompts shift from ‘what next?’ to ‘why is that allowed?’ and ‘what rule makes this valid?’

Her explanations become principle-based, and method errors become easier to diagnose.

The case is useful because it shows a process change: diagnosis → targeted intervention → independent retest → delayed verification. A good teaching case explains which decision changed, not merely that the student later performed better.

Case 3: Leonie explains everything

Leonie spends so much time verbalising routine steps that problem solving becomes slow.

Self-explanation is limited to method selection, unfamiliar transitions and errors.

The strategy becomes efficient because explanation is targeted where reasoning still needs support.

The case is useful because it shows a process change: diagnosis → targeted intervention → independent retest → delayed verification. A good teaching case explains which decision changed, not merely that the student later performed better.

How Parents Can Help Without Taking Over

Parents can support self-explanation by asking for the learner’s current decision before giving the answer. Useful prompts include: “What is the idea here?”, “Why does this step make sense?”, “What example would show the rule?”, “What would be different in a non-example?”, and “How will you check whether you still know this tomorrow?”

When the learner can perform a step reliably, remove the prompt. When the learner cannot, restore only enough support to restart useful thinking, then fade it again. The purpose of support is transfer of control.

How Tutors Can Use a Three-Student Small Group

In a three-student tutorial, self-explanation becomes visible because students can compare reasoning without disappearing into a large class. One learner can explain a rule, another can challenge it with a boundary case, and the third can apply it to a fresh example. The tutor can then hear whether identical answers came from understanding, imitation or guessing.

Small-group teaching should increase the density of useful explanation and feedback. Discussion is followed quickly by independent work so that social understanding is converted into individual performance.

How to Measure Progress

• Students justify key steps more accurately.
• Worked examples require less external explanation.
• Method selection becomes clearer.
• Plausible wrong moves are rejected earlier.
• Support can be faded sooner.
• Fresh-problem performance improves.
• Explanations become more concise without losing reasons.
• Transfer increases.

Marks matter, but they are a lagging indicator. Early progress may appear first as better starts, more accurate explanations, fewer repeated misconceptions, stronger retrieval or better strategy selection. Track the behaviours expected to produce later performance, then verify that the marks eventually follow.

A Four-Week Implementation Plan

Week 1 — Make the process visible

Choose one subject and one recurring task. Model self-explanation, let the learner attempt it, and record where the first breakdown occurs.

Week 2 — Reduce prompts

Use fresh examples. Remove one layer of support and require the learner to explain the reason for key decisions before checking.

Week 3 — Mix and delay

Return after several days, vary wording or representation, and require the learner to recognise when the method applies without a heading announcing it.

Week 4 — Perform under realistic conditions

Use authentic school questions, assignments, timed sections or a past paper where appropriate. Keep what works and identify the next bottleneck instead of adding complexity automatically.

Evidence and Responsible Use

Cambridge’s Handbook of Cognition and Education describes self-explaining as a powerful method when learners connect problem cases to underlying domain principles. Cambridge’s multimedia-learning handbook similarly reports benefits from encouraging learners to generate self-explanations. The Education Endowment Foundation’s worked-example guidance discusses explanation, fading and think-aloud prompts as ways to support pupils moving toward independent problem solving.

No single study technique replaces teaching, prior knowledge, sleep, practice volume appropriate to the learner, or professional support when needs extend beyond ordinary study strategy. The methods here are educational routines, not guarantees of a particular grade.

• Cambridge Handbook: Self-Explaining
• EEF: Mistakes and Explanations
• EEF: Worked Examples

Frequently Asked Questions

Is self-explanation just talking to yourself?

Not exactly. The valuable part is explaining underlying reasons, principles and connections rather than narrating obvious actions.

Should I explain every step?

No. Focus on important decisions, unfamiliar transitions, common errors and method selection. Routine expertise should become efficient.

Can self-explanation be wrong?

Yes. Check explanations against reliable sources or instruction.

Does self-explanation work with worked examples?

Yes. Worked examples are especially useful when learners explain why each important step follows.

What are faded examples?

Examples where some steps are progressively removed so the learner takes over more of the solution.

Can I use wrong examples?

Yes, if they are clearly identified as incorrect and the learner has enough knowledge to analyse the failure safely.

How is self-explanation different from elaboration?

They overlap. Self-explanation focuses strongly on explaining one’s understanding, steps and principles; elaboration is broader and includes making connections, comparisons and examples.

When should I stop self-explaining aloud?

As a process becomes fluent and accurate, reduce explicit narration and reserve it for difficult choices, errors and new situations.

Helpful Reading on eduKateSingapore

• How to Use Retrieval Practice
• How to Use Interleaving
• How to Make a Concept Map
• How to Make a Study Schedule
• How to Use Practice Exams

Do Not Just Follow the Step—Explain Why It Belongs

A visible solution can carry more reasoning than the learner realises. Self-explanation moves some of that reasoning back into the student’s own mind.

Ask what the step does. Ask why it is valid. Name the principle. Fade the model. Solve a fresh case. When the reasoning survives without the example, the explanation has done its job.

Properly taught kids shine a bright light into the future.

Extended Diagnostic Workshops

Workshop 1: Choose a meaningful step or idea × Parroting the model

Set up one fresh task in which the learner must practise choose a meaningful step or idea while watching specifically for parroting the model. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Do not self-explain every trivial motion. Focus on decisions, principles, transitions and places where learners commonly imitate without understanding.

The explanation repeats teacher language while the learner cannot generate it independently. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Write the theorem, rule or concept beside the step that uses it. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 2: Ask what the step accomplishes × Worked-example dependence

Set up one fresh task in which the learner must practise ask what the step accomplishes while watching specifically for worked-example dependence. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. State the immediate purpose of the step in the solution or explanation.

The learner can explain visible steps but cannot choose a method on a blank problem. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Explain two possible approaches and why one is more efficient or appropriate. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 3: Ask why the step is valid × No transfer

Set up one fresh task in which the learner must practise ask why the step is valid while watching specifically for no transfer. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Name the rule, theorem, causal principle, grammatical relation or evidence standard that justifies it.

Self-explanation stays tied to one example instead of the underlying principle. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Solve one fresh question while verbalising method choice and key checks. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 4: Connect to prior knowledge × Explaining every trivial step

Set up one fresh task in which the learner must practise connect to prior knowledge while watching specifically for explaining every trivial step. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Identify the prerequisite idea that makes the step intelligible.

The process becomes exhausting and attention is wasted on actions that no longer require analysis. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Change one condition in the problem and explain which steps would need to change. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 5: Compare with an alternative × No fading

Set up one fresh task in which the learner must practise compare with an alternative while watching specifically for no fading. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Ask what other move might look plausible and why it would be weaker or wrong.

Fully worked solutions remain present indefinitely, preventing independent completion. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Pause at each important worked-example step and explain why it belongs. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 6: Check the explanation × Narration without reasoning

Set up one fresh task in which the learner must practise check the explanation while watching specifically for narration without reasoning. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Use a reliable source or teacher to correct misconceptions in the self-generated reasoning.

The student says what they are doing but not why it is valid. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Write the theorem, rule or concept beside the step that uses it. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 7: Fade the worked example × No accuracy check

Set up one fresh task in which the learner must practise fade the worked example while watching specifically for no accuracy check. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Hide or remove later steps so the learner must complete more of the process independently.

A misconception is explained confidently and repeatedly. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Explain two possible approaches and why one is more efficient or appropriate. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.

Workshop 8: Transfer to a fresh problem × No incorrect examples

Set up one fresh task in which the learner must practise transfer to a fresh problem while watching specifically for no incorrect examples. Begin without the answer visible. Ask the learner to state the goal, make the first decision and mark the first point of uncertainty. Use the same principle in a new case and self-explain the method choice before solving.

Students only explain correct work and never practise diagnosing why a plausible solution fails. Now change one consequential feature of the task—wording, numbers, representation, example, context or time pressure—and repeat the decision. Add this related drill: Solve one fresh question while verbalising method choice and key checks. Finish by scheduling a later version so the next session tests memory and transfer rather than immediate imitation.