eduKate Sengkang

Small Group Tutorials

Here to help students catch up, keep up, and move ahead. Book a consultation here.

Chat on WhatsApp

How to Improve Understanding | Meaning, Structure, Explanation and Transfer

Understanding improves when a learner can reconstruct meaning, explain relationships, distinguish similar ideas, use knowledge in a new situation, and recognise when a familiar rule no longer applies.

This article is part of the eduKateSengkang How to Improve series. It sits beside How Understanding Works in Learning | Meaning, Structure, Explanation, Transfer and Reconstruction. That article explains the mechanism. This article is about improvement: what to do when a learner can remember words, follow examples or recognise notes but still does not properly understand the idea.

Understanding is often confused with familiarity. A page looks familiar, a worked solution seems obvious, or an explanation feels clear while the teacher is speaking. Yet the learner may be unable to reconstruct the idea later, explain why it works, choose when it applies, or use it when the surface changes. That gap is the space this article addresses.

The Simple Answer

To improve understanding, train this loop:

Activate Prior Knowledge → Clarify Meaning → Build Structure → Explain Relationships → Test With Examples → Test With Counterexamples → Retrieve → Apply → Transfer → Revise the Model

The central idea is that understanding is not created by exposure alone. It develops when the learner actively builds and tests a usable internal model.

What Does Better Understanding Look Like?

A learner who understands something well can usually do more than repeat a definition.

  • Explain the idea in their own words.
  • Identify the important relationships.
  • Give an example.
  • Give a non-example.
  • Compare it with a similar concept.
  • Predict what changes when one condition changes.
  • Use the idea in an unfamiliar question.
  • Detect when the idea does not apply.
  • Connect the idea to prior knowledge.
  • Recognise and repair misconceptions.

This is why understanding is deeper than recognition and broader than recall. Recall can support understanding, but remembered words are not sufficient evidence that the underlying relationships are understood.

First Diagnose the Type of Understanding Failure

“I don’t understand” can describe very different states. Before adding another explanation, identify the failure.

The vocabulary is blocking the idea

The learner may understand the underlying relationship but not the technical language. Repair the vocabulary and map each term onto something already meaningful.

A prerequisite is missing

The new concept may depend on earlier knowledge that is unstable. Re-explaining the new idea repeatedly will remain difficult until the prerequisite is repaired.

The learner knows pieces but not relationships

The facts are present, but the structure is missing. Use diagrams, comparisons, causal chains, concept maps and explanation prompts.

The learner can follow but not generate

This is common after worked examples. Fade the model and require reconstruction.

The learner can explain but cannot apply

The knowledge may be verbal but not operational. Use varied problems and transfer tasks.

The learner applies a rule too broadly

The missing element is boundary knowledge. Use non-examples, edge cases and contrasting situations.

Diagnosis matters because another explanation is not always the right repair.

Activate Prior Knowledge

New understanding is built from what the learner already knows. Prior knowledge can help, but it can also mislead when it is incomplete or incorrect.

Before teaching or studying something new, ask:

  • What do I already know about this?
  • What does this remind me of?
  • Which earlier concept does it depend on?
  • What do I currently predict?
  • Which part feels familiar and which part is new?

Then compare the old model with the new one. If the prior model is wrong, make the change explicit.

See How Prior Knowledge Works in Learning.

Build Meaning Before Memorising Wording

Exact wording sometimes matters, especially for definitions and technical vocabulary. But memorising wording before understanding can create brittle knowledge.

Before memorising, identify:

  • what the concept refers to,
  • what changes,
  • what remains constant,
  • what causes what,
  • what the concept is used for,
  • what would count as an example,
  • what would not count.

Once the meaning is clear, precise terminology becomes easier to attach to the structure.

Organise Facts Into a Structure

Understanding improves when isolated facts become organised knowledge.

Useful structures include:

  • cause and effect,
  • sequence,
  • part and whole,
  • category and member,
  • input and output,
  • condition and consequence,
  • problem and method,
  • claim and evidence,
  • similarity and difference.

For example, memorising the stages of a Science process is weaker than understanding how one stage changes the conditions for the next. Memorising an algebraic method is weaker than understanding which mathematical relationship makes that method appropriate.

See How Schema Formation Works in Learning.

Use Multiple Representations

One representation can hide what another reveals.

Try translating the same idea across:

  • words,
  • diagrams,
  • tables,
  • graphs,
  • equations,
  • physical examples,
  • timelines,
  • concept maps.

Translation is powerful because it forces the learner to preserve the relationship while changing the surface form.

In Mathematics, this might mean moving from a word problem to a bar model to an equation. In Science, it might mean moving from a paragraph to a labelled mechanism diagram. In English, it might mean moving from a passage to an argument map.

Use Self-Explanation

Self-explanation asks the learner to explain why a step, fact or relationship makes sense.

Useful prompts include:

  • Why is this step necessary?
  • How does this connect to what came before?
  • Why does this formula fit?
  • What would happen if this condition changed?
  • Which evidence supports this conclusion?
  • What is the hidden assumption?

See How Self-Explanation Works in Learning.

Do Not Let Explanation Become Performance Theatre

A learner can produce smooth words without deep understanding. Memorised explanations can sound impressive while remaining fragile.

Test the explanation by changing the question:

  • Ask for a different example.
  • Remove one condition.
  • Reverse the direction of the problem.
  • Ask for a prediction.
  • Ask why a tempting alternative is wrong.
  • Ask the learner to draw the mechanism.

If the explanation survives variation, it is stronger evidence of understanding.

Use Examples to Ground Abstraction

Abstract ideas become more understandable when anchored to concrete cases.

But one example can accidentally teach irrelevant surface features. Use several examples that share the deep structure while varying the surface.

Then ask:

  • What stays the same across these examples?
  • What changes?
  • Which feature is essential?
  • Which feature is incidental?

This is how examples begin to support abstraction.

Use Non-Examples

A non-example shows what the concept is not.

Non-examples are especially useful when learners overgeneralise.

Ask:

  • Why does this case fail the definition?
  • Which condition is missing?
  • What looks similar but is actually different?

Boundary knowledge is part of understanding. Knowing when a rule does not apply can be as important as knowing when it does.

Use Contrast Pairs

Similar concepts are often confused because learners study them separately.

Put them side by side:

  • mass versus weight,
  • speed versus velocity,
  • correlation versus causation,
  • factor versus multiple,
  • inference versus direct statement,
  • accuracy versus precision.

Then identify the decisive difference. Contrast sharpens categories.

Find the Mechanism

Many learners can state that something happens without understanding why.

Ask for the mechanism:

  • What changes first?
  • What causes the next change?
  • What transmits the effect?
  • What evidence would show the mechanism is operating?
  • What would stop it?

Mechanistic understanding is especially important in Science, but the habit also applies in Mathematics, language and learning itself.

Ask “Why?” and “How?” at the Right Depth

Repeatedly asking “why?” can deepen understanding, but only if the learner has enough prior knowledge to answer productively.

If every answer introduces another unfamiliar idea, the explanation has moved beyond the learner’s current foundation.

Stop at the depth where the learner can connect the explanation to stable prior knowledge. Then build further later.

Manage Cognitive Load

Understanding can fail when too many new elements must be processed at once.

Reduce unnecessary load by:

  • breaking explanations into coherent stages,
  • placing related information together,
  • removing irrelevant decoration,
  • using diagrams when spatial relationships matter,
  • pre-teaching essential vocabulary,
  • externalising intermediate steps.

Then gradually increase complexity as the learner’s schema becomes stronger.

See How Cognitive Load Works in Learning and How Working Memory Works in Learning.

Break Complex Ideas Into Parts, Then Reassemble Them

Isolation helps when one component is blocking understanding. But isolated pieces must eventually be reconnected.

A strong sequence is:

  1. identify the components,
  2. understand each component,
  3. understand the relationship between components,
  4. reconstruct the whole system,
  5. test the whole system in a new case.

Understanding lives not only in the parts but in the relationships among them.

Use Worked Examples Properly

Worked examples reduce load and reveal expert structure. They are especially useful when the learner is new to a method.

But a learner can follow a worked example without owning the method.

Use this progression:

  1. Study the complete worked example.
  2. Explain why each step exists.
  3. Cover selected steps and reconstruct them.
  4. Complete a partially worked example.
  5. Solve a similar problem independently.
  6. Solve a mixed problem where the method is not announced.

This moves from comprehension to generation.

Fade Scaffolding

Support should make understanding possible and then become unnecessary.

Fade:

  • full diagrams into partially labelled diagrams,
  • worked examples into completion problems,
  • teacher prompts into self-prompts,
  • chapter labels into mixed practice,
  • answer choices into open responses.

See How Scaffolding Works in Learning.

Retrieve the Structure, Not Only the Facts

Retrieval practice is often used for factual recall. It can also strengthen understanding.

Ask the learner to retrieve:

  • the causal chain,
  • the comparison table,
  • the decision rule,
  • the reason a formula applies,
  • the conditions under which a concept fails,
  • the structure of an argument.

This turns understanding into something that can be reconstructed without the original explanation present.

Ask the Learner to Predict

Prediction is a strong test because it requires the internal model to generate an outcome.

Ask:

  • What happens if this variable increases?
  • What happens if this condition is removed?
  • Which method should work here?
  • How will the graph change?
  • What would this character likely do next?

Then compare the prediction with evidence. Incorrect predictions expose the model’s weak points.

Use Counterfactuals

Counterfactual questions ask what would happen if something were different.

Examples:

  • What if friction were absent?
  • What if this value were zero?
  • What if the writer removed this evidence?
  • What if this assumption were false?
  • What if the direction of the relationship were reversed?

Counterfactuals reveal whether the learner understands dependency rather than memorised sequence.

Explain Errors, Not Just Correct Answers

A wrong answer can be an unusually powerful teaching object.

Ask:

  • Why is this answer tempting?
  • At which step does the reasoning fail?
  • Which misconception would produce it?
  • What evidence disproves it?

This helps learners build not only the correct route but also warning signals for common wrong routes.

Use Misconception Checks

Misconceptions can survive ordinary teaching because correct examples do not always challenge them.

Design questions that distinguish the correct model from the misconception.

For example, if a student believes heavier objects always fall faster, choose a situation that directly tests that belief. If a student believes every increase in one variable causes an increase in another, use a counterexample.

The best misconception question is diagnostic: different answers reveal different internal models.

Teach the Boundary Conditions

Understanding is incomplete when the learner knows a rule but not its limits.

For every important rule, ask:

  • Under what conditions does it hold?
  • What assumptions are required?
  • What would make it fail?
  • What nearby rule takes over outside the boundary?

Boundary knowledge prevents overgeneralisation.

Compare Near Neighbours

Understanding becomes sharper when similar concepts are discriminated.

Use a table with columns such as:

  • definition,
  • key feature,
  • when it applies,
  • common example,
  • common confusion,
  • decisive difference.

This is particularly useful for grammar, mathematical methods, Science variables and conceptual categories.

Use Analogy, Then Break the Analogy

Analogies help by mapping a new idea onto a familiar structure.

But every analogy has limits.

After using an analogy, ask:

  • Which parts match?
  • Which parts do not?
  • Where would this analogy become misleading?

This protects the learner from turning a teaching aid into a false literal model.

Use Questions That Require Relationships

Questions shape what learners attend to.

Instead of asking only “What is X?”, ask:

  • How is X related to Y?
  • Why does X cause Y?
  • What changes if X is removed?
  • How is X different from Z?
  • Which evidence would show X is occurring?
  • When would X not apply?

These questions move the learner from labels toward structure.

Teach Learners to Generate Their Own Questions

Question generation reveals what the learner thinks is important.

Ask for:

  • one definition question,
  • one why question,
  • one comparison question,
  • one prediction question,
  • one transfer question,
  • one misconception question.

Then answer them later without notes.

Use Teaching as a Test of Understanding

Preparing to teach can expose gaps because the learner must organise knowledge for someone else.

A useful teaching test asks the learner to:

  • define the idea,
  • show an example,
  • explain the mechanism,
  • anticipate one misconception,
  • answer a follow-up question.

Then return to independent performance. See How Learning by Teaching Works.

Do Not Confuse Verbal Fluency With Understanding

Some learners speak confidently and rapidly. Others understand deeply but need more time to express the model.

Use multiple forms of evidence:

  • explanation,
  • diagram,
  • application,
  • prediction,
  • problem solving,
  • transfer.

Understanding should survive more than one performance format.

Do Not Confuse Correct Answers With Understanding

A learner can be correct for the wrong reason.

After selected correct answers, ask:

  • Why?
  • How did you know?
  • What alternative did you reject?
  • Would your answer change if this condition changed?

This is especially useful when guessing, pattern matching or memorised procedures can produce the right endpoint accidentally.

Do Not Confuse Memory Failure With Understanding Failure

A learner may understand a concept but fail to retrieve the terminology. Another may recall every term but misunderstand the mechanism.

Use separate tests:

  • Can the learner reconstruct the explanation with prompts?
  • Can the learner retrieve it without prompts?
  • Can the learner apply it to a new case?

The answers distinguish availability from meaning.

See How to Improve Memory.

Use Transfer as the Strong Test

Understanding becomes more convincing when it survives a change in surface form.

Change:

  • the numbers,
  • the context,
  • the diagram,
  • the order of information,
  • the response format,
  • which variable is unknown,
  • the amount of support.

Then ask what deep structure stayed the same.

See How Learning Transfer Works.

Move From Near Transfer to Farther Transfer

Transfer can be graduated.

  • Near transfer: same structure, slightly changed surface.
  • Moderate transfer: same principle in a new context.
  • Farther transfer: learner must recognise that an idea from one domain helps in another.

Do not jump immediately from worked example to extreme novelty. Build transfer progressively.

Interleave Related Concepts

Blocked practice improves execution because the method is known in advance. Interleaving improves discrimination because the learner must choose.

Once basic understanding is stable, mix neighbouring concepts and ask why one applies instead of another.

See How Interleaving Works in Learning.

Space Understanding Across Time

Understanding can feel deep immediately after an explanation because the explanatory structure is still active in working memory.

Return later and ask the learner to reconstruct the model. If the relationships can still be generated after a delay, the understanding is becoming more durable.

See How Spacing Works in Learning.

Use Delayed Explanation

One simple test is to explain the idea today and ask for the explanation again tomorrow without notes.

Then ask for:

  • one example,
  • one non-example,
  • one prediction,
  • one connection to another topic.

This small routine tests meaning, boundaries and transfer together.

Improve Understanding in Mathematics

Mathematical understanding includes knowing what quantities mean, how relationships are represented, why procedures work and when a method applies.

To improve it:

  • connect symbols to quantities,
  • translate between equations, diagrams and words,
  • explain each operation,
  • compare multiple solution methods,
  • test special cases,
  • use estimation,
  • solve mixed problems where the method is not labelled.

A student should know not only how to perform the algorithm but what makes the algorithm appropriate.

Improve Understanding in Science

Scientific understanding connects observations to mechanisms and models.

Ask:

  • What is directly observed?
  • What is inferred?
  • What causes the change?
  • What evidence supports the mechanism?
  • What alternative explanation is possible?
  • What experiment would distinguish the alternatives?

This prevents Science from becoming a collection of detached keywords.

See How Scientific Evidence Works.

Improve Understanding in English

English understanding involves building meaning across words, sentences, paragraphs, tone, purpose and evidence.

To improve comprehension:

  • paraphrase difficult sentences,
  • track reference words,
  • map claims and evidence,
  • identify tone from language choices,
  • distinguish explicit statement from inference,
  • compare plausible interpretations against the text.

For writing, understanding means knowing why a structure works rather than copying a template blindly.

Improve Understanding in Vocabulary

Knowing a word is more than recalling a dictionary definition.

Build:

  • meaning,
  • pronunciation,
  • grammatical role,
  • collocations,
  • tone,
  • contrast with related words,
  • examples and non-examples.

Then retrieve and use the word in varied sentences.

Improve Understanding in Studying

Students should understand not only subject content but also why their study methods work.

A learner who understands retrieval practice knows why closing the book matters. A learner who understands spacing knows why forgetting a little before returning can be useful. A learner who understands interleaving knows why mixed practice can feel harder while training method selection.

This understanding improves study decisions. See How to Improve Studying.

Improve Understanding in Problem Solving

Problem solving improves when learners recognise deep structure rather than matching surface patterns.

After solving, ask:

  • What was the decisive relationship?
  • Which information mattered?
  • Why did this strategy fit?
  • What other problem has the same structure?

See How to Improve Problem Solving.

Understanding and Critical Thinking

Understanding gives the learner a model. Critical thinking tests whether that model is justified.

Ask:

  • What evidence supports this explanation?
  • What assumption does it depend on?
  • What alternative model also fits?
  • What result would change my mind?

See How to Improve Critical Thinking.

Understanding and Motivation

Confusion can damage motivation because effort feels ineffective. Improved understanding can restore expectancy: the learner begins to see a route.

But motivation can also support understanding by sustaining the effort required to build and test models.

See How to Improve Motivation.

Understanding and Confidence

Confidence should follow performance evidence.

A learner may feel confident because an explanation seems familiar. Test confidence against:

  • closed-book explanation,
  • new examples,
  • counterexamples,
  • prediction,
  • transfer.

If confidence and performance disagree, recalibrate.

See How Confidence Works in Learning.

What to Do When an Explanation Does Not Work

Do not repeat the same explanation louder or longer.

Change something:

  • use a different representation,
  • repair vocabulary,
  • check prerequisites,
  • use a concrete example,
  • compare with a non-example,
  • reduce the number of simultaneous elements,
  • ask the learner to show exactly where the model stops making sense.

The failure of one explanation is information about the learner’s current representation.

What to Do When the Learner Says “I Get It” Too Quickly

Do not argue. Test gently.

  • Explain it without the notes.
  • Give a new example.
  • Tell me when the rule would fail.
  • Predict what happens if this variable changes.
  • Compare it with this similar concept.

If the learner succeeds, confidence is supported. If not, the gap becomes visible without turning the conversation into a contest.

What to Do When the Learner Can Explain but Cannot Solve

The missing step may be method selection or procedural fluency.

Move from verbal explanation to application:

  • solve one guided example,
  • solve one similar example independently,
  • solve one mixed problem,
  • explain why the method applies.

This connects conceptual knowledge to performance.

What to Do When the Learner Solves but Cannot Explain

Procedural knowledge may be ahead of conceptual understanding.

Ask the learner to annotate each step with its purpose. Compare two methods. Use a representation. Change one condition and ask whether the method still works.

The aim is not to slow every routine procedure forever. It is to expose enough structure that the learner can adapt when the familiar pattern changes.

What to Do When Understanding Disappears After a Week

Understanding and memory interact. If the learner cannot reconstruct the model after a delay, use spaced retrieval of the structure itself.

Do not simply reread the original explanation. First attempt reconstruction. Then repair the missing relationships.

What to Do When Transfer Fails

If the learner succeeds only on familiar forms, compare the original problem and the new one explicitly.

Ask:

  • What changed on the surface?
  • What stayed the same structurally?
  • Which cue did you rely on before?
  • What deeper cue should replace it?

Then create a sequence of variations rather than jumping directly to unrelated novelty.

A 15-Minute Understanding Drill

  1. 2 minutes: explain the concept from memory.
  2. 3 minutes: draw or map the structure.
  3. 3 minutes: give an example and a non-example.
  4. 3 minutes: answer one “what if?” question.
  5. 2 minutes: apply the concept to a new case.
  6. 2 minutes: check against a reliable source and repair the model.

A 30-Minute Understanding Rebuild

  1. Identify the exact point of confusion.
  2. Check the prerequisite.
  3. Study one clear explanation.
  4. Translate it into another representation.
  5. Explain the relationship in your own words.
  6. Test with an example.
  7. Test with a non-example.
  8. Apply to a new question.
  9. Schedule delayed reconstruction.

A Weekly Understanding Review

At the end of the week, choose the most important concepts and classify them.

  • Recognised: looks familiar but cannot be reconstructed.
  • Explained: can be described in own words.
  • Structured: relationships and conditions are clear.
  • Applied: can be used in standard questions.
  • Transferred: can be recognised and used in unfamiliar forms.

This gives understanding a progression rather than treating it as simply present or absent.

For Parents: Ask for Meaning, Not Recitation

Parents do not need to know every subject deeply to test understanding.

Ask:

  • Can you explain this without reading?
  • Can you give me an example?
  • What is it often confused with?
  • What changes if this condition changes?
  • How would you know when this idea applies?

The child is doing the intellectual work.

For Teachers: Make Invisible Structure Visible

Good explanations reveal relationships that novices cannot yet see.

Model:

  • what is important,
  • how parts connect,
  • why one method fits,
  • where misconceptions arise,
  • what changes under different conditions.

Then ask students to reconstruct the structure themselves. See How Explanation Works in Teaching.

Common Understanding Traps

  • Familiarity illusion: recognising the page is mistaken for understanding.
  • Definition-only learning: wording is memorised without relationships.
  • Single-example dependence: one surface form becomes the concept.
  • No non-examples: boundaries remain unclear.
  • Worked-example dependence: following is mistaken for generating.
  • Explanation theatre: fluent words hide weak application.
  • No mechanism: the learner knows that something happens but not why.
  • Overgeneralisation: a rule is applied outside its conditions.
  • Prerequisite blindness: new explanations are repeated while old foundations remain weak.
  • Transfer failure: knowledge works only in the original form.
  • No delayed reconstruction: immediate clarity is mistaken for durable understanding.

How to Know Understanding Has Improved

  • The learner explains more accurately in their own words.
  • Relationships become easier to represent.
  • Examples and non-examples are generated correctly.
  • Similar concepts are confused less often.
  • Predictions become more accurate.
  • The learner detects when a rule does not apply.
  • Fewer prompts are needed.
  • Transfer improves across changed surface forms.
  • Misconceptions recur less often.
  • The learner can reconstruct the model after a delay.

The Understanding Improvement Equation

Usable Understanding = Meaning × Structure × Explanation × Boundary Knowledge × Retrieval × Transfer

This is a conceptual model. It shows why understanding can remain fragile even when one element is strong. Meaning without structure stays disconnected. Explanation without boundary knowledge overgeneralises. Retrieval without transfer stays context-bound. Transfer without accurate meaning becomes guesswork.

The Deepest Test

When you think you understand something, change the situation.

Change the example. Change the representation. Remove a condition. Ask for a prediction. Ask for a non-example. Return after a delay.

If the model still works, the understanding is becoming real.

Continue the How to Improve Series

Final Principle

Understanding is not proved by how clear an explanation feels while it is in front of you.

It is proved by whether you can rebuild the meaning, preserve the relationships, recognise the boundaries and use the idea when the situation changes.