Chapter 10: Make Mistakes Useful

How feedback, correction and recovery turn imperfect performance into lasting learning

Full first draft | Part III: Building Knowledge and Competence

The red cross

A fifteen-year-old is working through a GCSE Biology question.

Explain how the structure of a red blood cell is adapted to its function.

She writes:

Red blood cells carry oxygen around the body. They have haemoglobin and no nucleus.

The answer is partly correct. She knows the principal function of red blood cells, understands the role of haemoglobin and remembers that mature human red blood cells lack a nucleus. But she has not explained how these features help the cell perform its function, and she has omitted the importance of its biconcave shape.

Her parent looks at the answer.

“You need to explain it properly. You’ve missed the adaptations.”

The teenager frowns.

“I did explain it.”

“You’ve just listed things. You need to say why.”

“I don’t know what you want.”

“We’ve done this before.”

“I know.”

“Well, then, why haven’t you done it?”

The teenager pushes the paper away.

“I’m rubbish at Biology.”

The conversation has taken less than a minute.

The original problem was small and educationally straightforward. The teenager possessed relevant knowledge but had not yet connected that knowledge to the demands of the question. A short explanation, a modelled sentence and another attempt might have been enough.

Instead, a correctable academic error has become a confrontation about competence, effort and memory.

The teenager is no closer to understanding how to answer the question. The parent has learned little about why the answer was incomplete. Both are frustrated.

Now imagine a different response.

“You’ve remembered three important facts. The missing part is explaining how those facts help the cell carry oxygen. Watch this.”

The parent writes:

Red blood cells contain haemoglobin, which binds to oxygen so it can be transported around the body.

“See the difference? You’ve connected the feature to its function. Can you make that connection for the missing nucleus?”

The teenager thinks.

“Without a nucleus, there’s more room for haemoglobin?”

“Exactly. Now try writing that as a complete explanation.”

She writes the sentence.

The parent then introduces the biconcave shape, explaining how it increases surface area relative to volume and helps oxygen exchange.

The teenager tries the question again.

This time her answer is better.

Not perfect. Not necessarily remembered next week. Not yet proof that she can independently answer every unfamiliar question about biological adaptations.

But something important has happened.

She has used feedback to improve a piece of academic work.

She has experienced an error as something that can be understood and corrected.

And she has begun learning a general strategy: when a question asks how a structure is adapted to its function, identify the feature, explain its effect and connect that effect to the function.

That is the purpose of feedback.

Feedback is not merely information about what was wrong. It is information the learner can use to do something better.

1. An error is information, not a verdict

Errors are unavoidable in serious learning.

A teenager studying GCSE Mathematics will occasionally choose the wrong operation. In History, she may confuse chronology with causation. In English Literature, she may identify a quotation correctly but struggle to explain its significance. In Psychology, she may know the definition of a concept but apply it incorrectly to an unfamiliar scenario.

None of these errors, by itself, tells us that the learner is incapable, unwilling or careless.

An error tells us that the performance did not meet a particular requirement.

The important question is what produced that result.

Consider a teenager who incorrectly answers:

\[ 3(x+4)=21 \]

She writes:

\[ 3x+4=21 \]

\[ 3x=17 \]

\[ x=\frac{17}{3} \]

The answer is wrong. But what exactly is wrong?

Perhaps she does not understand the distributive property. Perhaps she knows it but overlooked the brackets. Perhaps she has copied the example incorrectly. Perhaps she is rushing. Perhaps the page contains so many problems that she has stopped checking her work.

The same incorrect answer could arise through different mechanisms.

And different mechanisms require different responses.

If the learner does not understand distribution, she needs instruction.

If she understands it but failed to notice the brackets, she may need a checking strategy.

If she has forgotten a previously learned procedure, a cue or brief retrieval opportunity may be sufficient.

If she cannot concentrate because the environment is overwhelming, more explanation may achieve little until that barrier is addressed.

The correction should respond to the cause of the error, not merely its appearance.

This distinction is particularly important when supporting an autistic teenager.

Autistic learners are not a uniform group. Executive-function differences, sensory conditions, anxiety, communication differences and prior learning experiences may affect how knowledge is demonstrated. The 2024 meta-analysis of executive function across neurodevelopmental conditions found meaningful group-level differences, but also substantial variation. A diagnosis alone cannot establish why an individual answer is wrong.

2024 - Executive function in children with neurodevelopmental conditions- a systematic review and meta-analysis.pdf

The parent or teacher must investigate the actual difficulty.

Sometimes the difficulty is simply that the learner has not yet learned the material.

We should not turn every academic mistake into a complicated psychological explanation.

Nor should we assume that every mistake represents a failure of effort.

The practical discipline is to begin with the work itself.

What was the question asking?

What did the learner actually do?

Where did the response diverge from the requirement?

What would help her perform the next step correctly?

2. Why feedback sometimes fails

Educational research generally supports feedback as a potentially powerful influence on learning, but the effects are far from uniform.

The feedback literature reviewed for this book includes a large meta-analysis by Wisniewski, Zierer and Hattie, covering 435 studies and more than 61,000 participants. Its average effect was positive, but the results varied substantially. Some feedback interventions were associated with worse outcomes.

That matters because the word feedback covers very different activities.

A red cross is feedback.

So is a numerical mark.

So is a teacher saying “Excellent!”

So is a detailed explanation of a misconception.

And so is a carefully chosen hint that allows a student to solve a problem independently.

These are not educationally equivalent.

Consider four responses to an incorrect History answer.

Response A:

“Wrong.”

The learner knows the answer was not accepted. She does not know why.

Response B:

“Three out of six marks.”

The learner knows roughly how the answer was assessed. She still may not know how to improve it.

Response C:

“You need more detail.”

The learner has received a direction, but one that may be too vague to act upon.

Response D:

“You’ve identified the Bosnian Crisis as a source of tension. Now explain the causal link: how did Austria-Hungary’s annexation of Bosnia increase Serbian hostility? Add one sentence making that relationship explicit.”

The fourth response identifies a specific gap and supplies an actionable next step.

It is much more likely to support useful revision of the answer.

But even Response D is not automatically effective.

If the teenager is overwhelmed, does not understand the vocabulary or has no opportunity to revise the answer, the feedback may remain unused.

This gives us an important principle:

Feedback is not complete when the adult has delivered it. It is complete when the learner has had an opportunity to understand and use it.

Feedback should answer a useful question

The most productive feedback usually helps the learner establish one or more of the following:

• What was I trying to do?
• What did I do correctly?
• What is missing or incorrect?
• What should I change?
• How can I check whether the change worked?

These questions do not all need to be asked aloud.

A teenager who already understands an error may need only a small cue.

A novice who lacks the underlying knowledge may need a complete explanation.

A student who has produced a strong answer may need no correction at all, but may benefit from learning what made it effective.

The amount of feedback should depend on the learner’s current competence.

Too little information leaves the learner stuck.

Too much information can obscure the essential point, create unnecessary cognitive load or remove the opportunity to think.

The aim is not maximal feedback.

It is the minimum sufficient information that enables useful progress.

3. The Feedback-to-Learning Cycle

We can turn these principles into a practical routine.

The Feedback-to-Learning Cycle

1. Attempt the task

2. Identify the gap

3. Select useful feedback

4. Correct or reconstruct

5. Try again

6. Check independently

7. Revisit after a delay

Practical synthesis of general feedback, instructional and retrieval-practice research. This complete sequence has not been experimentally validated as a single intervention for autistic GCSE learners.

Step 1: Let the learner produce something

Feedback needs evidence.

The teenager must have an opportunity to demonstrate what she currently knows or can do.

That might mean answering a short question, solving a problem, explaining a concept, identifying a quotation or writing one paragraph.

The attempt should be appropriate to her current knowledge.

There is little educational value in repeatedly demanding an answer to something she has never been taught.

Equally, we should not always prevent mistakes by supplying every answer in advance.

If she has sufficient knowledge to attempt the task, an independent response provides useful information about her learning.

Step 2: Identify the most important gap

Do not correct everything indiscriminately.

Imagine a GCSE English Literature paragraph containing a weak argument, two spelling errors, an inaccurate quotation and poor punctuation.

The parent could mark every problem.

But if the learning objective is to explain how the writer presents a character, the most important issue may be the missing analysis.

Correcting spelling first could consume attention without improving the central skill.

A useful question is:

Which correction would most improve the learner’s ability to perform this kind of task?

Sometimes several errors share one underlying cause.

A student who repeatedly loses marks on six-mark Biology questions may not need six separate explanations. She may need to learn how to construct a chain of scientific reasoning.

Correct the underlying pattern where possible.

Step 3: Choose the appropriate feedback

There are several levels of assistance.

These are choices, not a rigid hierarchy.

A learner who is genuinely missing knowledge should not be forced through increasingly frustrating hints when a clear explanation would be more effective.

Step 4: Make the correction active

The learner should do something with the information.

If the feedback concerns a missing causal link in History, she writes the link.

If the feedback concerns algebraic distribution, she solves another example.

If the feedback concerns identifying the dependent variable in Psychology, she identifies it in a fresh scenario.

Listening to an explanation is not the same as being able to use it.

The correction should therefore produce a new attempt.

Step 5: Check whether the improvement is real

A corrected answer immediately after a model may show that the learner can follow the model.

That is useful performance, but it does not yet establish independent learning.

Ask a similar question without the same degree of assistance.

For example, after correcting:

\[ 3(x+4)=21 \]

try:

\[ 4(x+2)=28 \]

If the learner succeeds without prompting, there is stronger evidence that the procedure is becoming understood.

If she fails again, the explanation or level of assistance needs reconsideration.

Step 6: Return later

Immediate success is not the endpoint.

The next day, or later in the week, include a similar problem in a short retrieval session.

If the learner can solve it after a delay, there is better evidence of retention.

Later still, vary the problem so that she must recognise when the strategy applies.

For example:

\[ 5(2x-1)=35 \]

This requires more than reproducing the exact earlier example.

Eventually, she should encounter the procedure among other algebra problems and select it appropriately.

That is the difference between reproducing a practised response and developing flexible knowledge.

4. When correction feels like criticism

Feedback is an educational event, but it is also a social interaction.

For some teenagers, a correction is simply information.

For others, particularly those with histories of repeated academic difficulty, criticism or misunderstanding, it may carry additional meaning.

The teenager may hear:

“You’ve made an error.”

But interpret:

“You’ve failed again.”

Or:

“You should already know this.”

Or:

“The adult is disappointed in me.”

These interpretations are not inevitable consequences of autism. They may reflect temperament, experience, anxiety, relationship history or the particular circumstances of the interaction.

Nevertheless, autistic secondary-school students’ accounts repeatedly identify relationships, communication, wellbeing and being understood as important aspects of educational experience. The 2025 scoping review by Kasztelan Chapman and colleagues synthesised 13 studies foregrounding autistic students’ own perspectives.

2025 - What autistic students say about their learning and teaching experiences in mainstream secondary school- a scoping review.pdf

A parent supporting home education therefore has two jobs during correction.

The first is to communicate accurate academic information.

The second is to avoid turning that information into an unnecessary threat to the relationship or the learner’s sense of capability.

These goals are compatible.

Correct the work without judging the person

Compare:

“You always forget to explain your answers.”

with:

“This answer identifies the feature, but the explanation of its function is missing.”

The second statement is more precise.

It concerns the work, not the learner’s character.

It also suggests a concrete route to improvement.

Similarly:

“You’re being careless.”

is usually less useful than:

“You’ve multiplied the first term in the bracket but not the second.”

The second statement tells the learner what to inspect.

Precision makes feedback less personal and more actionable.

Do not manufacture praise

A common response to anxiety about correction is to praise everything.

“Brilliant!”

“Fantastic!”

“You’re amazing!”

But indiscriminate praise can become uninformative.

A teenager may know perfectly well that an answer is incomplete. Excessive reassurance can make the adult appear unreliable or can distract from the work.

More useful feedback identifies genuine success.

“You correctly identified haemoglobin and explained its role. Now we need to connect the other adaptations to oxygen transport.”

This communicates two things:

The learner has accomplished something real.

There is still something specific to learn.

Neither fact cancels the other.

Give the learner some control over feedback

A teenager may prefer to receive corrections in writing rather than through an immediate verbal discussion.

She may prefer one correction at a time.

She may want to finish an answer before anyone comments.

She may prefer to compare her work with a model answer before discussing it with a parent.

These preferences are worth investigating.

For example:

“Would you rather I mark the answer and leave you to improve it, or go through it together?”

This is meaningful autonomy within a structured learning process.

The academic standard remains.

The learner has some control over how she works towards it.

5. Different errors require different teaching

One of the most useful skills a parent or teacher can develop is recognising that errors are not interchangeable.

A knowledge error

A student cannot explain what homeostasis means.

She has never securely learned the concept.

The appropriate response is instruction.

Explain the idea clearly, give examples, check understanding and provide opportunities for retrieval.

Asking her repeatedly to guess the answer is unlikely to be productive.

A retrieval error

The student has previously demonstrated that she understands homeostasis but cannot recall the definition today.

A cue may be useful:

“Think about keeping internal conditions stable.”

If that enables recall, the problem may concern retrieval rather than complete absence of knowledge.

The information should still be revisited later.

A conceptual error

The student believes respiration and breathing are the same process.

She knows some relevant vocabulary but has constructed an incorrect relationship.

A correction should distinguish ventilation from cellular respiration, perhaps through a comparison of what happens in the lungs and what happens in cells.

Simply supplying the correct definition may not resolve the misconception.

A strategy error

The student knows the facts about the causes of the First World War but writes an essay that lists events without explaining relationships between them.

The missing skill is not necessarily historical knowledge.

It may be selecting and organising evidence to construct an argument.

A worked example showing how to connect an event to an increase in international tension may be more useful than further memorisation.

An execution error

The student understands a mathematical procedure but makes an arithmetic slip.

A brief prompt to check the calculation may be sufficient.

A complete reteaching of the procedure could be unnecessary and irritating.

An access or regulation problem

The student knows the answer but is struggling to read a dense page, organise multiple instructions or continue working in an overwhelming environment.

The immediate response may involve formatting, task organisation, movement, rest or sensory adjustment.

That is not a substitute for teaching.

It is a way of making teaching accessible.

A major practical error occurs when the adult uses the same response for every difficulty.

“Try harder.”

“Concentrate.”

“Think about it.”

“Read the question again.”

Sometimes these prompts help.

Often they fail because they do not address the actual obstacle.

The better question is:

What does the learner need in order to make the next useful attempt?

6. How much struggle is useful?

Parents and teachers sometimes face a difficult choice.

Should they intervene immediately when a teenager struggles?

Or should they allow the learner to persist?

Neither answer is always correct.

Difficulty can be productive when the learner has enough relevant knowledge, understands the goal and has a reasonable route towards progress.

But difficulty can become unproductive when the learner is searching blindly, repeating an ineffective strategy or becoming increasingly distressed without acquiring useful information.

General learning research supports both substantial guidance for novices and, under appropriate conditions, productive attempts before instruction.

The distinction is not simply between easy and difficult tasks.

It is between difficulty that contributes to learning and difficulty that consumes effort without useful progress.

Consider a teenager solving a GCSE Mathematics problem.

She understands the procedure but is taking time to work through it.

Intervening immediately may remove valuable independent practice.

Now consider a teenager who does not understand what simultaneous equations are and has been staring at the same question for ten minutes.

Insisting that she persist without instruction is unlikely to develop mathematical competence.

A practical intervention rule is to observe what the learner is doing.

If she is making plausible progress, allow room for thought.

If she is repeating the same error, offer a targeted cue.

If she lacks the necessary knowledge, teach it.

If the task has become inaccessible because of overload, address the barrier.

This is not an argument for removing challenge.

It is an argument for making challenge educationally worthwhile.

7. When the learner refuses feedback

Imagine a teenager has completed a GCSE Psychology worksheet.

Her parent begins reviewing the answers.

“The first three are right. On question four, you’ve confused the independent and dependent variables.”

“I don’t care.”

“Well, you need to know this.”

“I said I don’t care.”

“We have to correct it.”

“Leave me alone.”

At this point, the parent may feel compelled to continue because the correction matters.

But there are two separate questions.

First, does the work require correction?

Yes.

Second, must that correction happen immediately, in this interaction, using this method?

Not necessarily.

The distinction matters.

A parent can preserve the academic requirement while changing the route towards it.

For example:

“We do need to revisit question four. We can do it now, or I can mark the page and you can look at it after lunch.”

This is not an attempt to bribe the learner into compliance.

It separates the legitimate learning goal from an unnecessary struggle over timing and interaction.

However, repeated postponement may become a problem if the learner never returns to the work.

The parent therefore needs a reliable system for unfinished corrections.

A simple correction queue can help.

The queue makes correction a normal part of learning rather than a personal confrontation.

It also creates a visible record of improvement.

When refusal becomes a recurring pattern

If every correction produces conflict, changing the wording of feedback may not be enough.

The adult should investigate the pattern.

Does correction arrive unexpectedly?

Is the teenager receiving too much feedback at once?

Does she understand what success looks like?

Does she feel that her work is constantly being inspected?

Is the parent correcting work that the learner could check independently?

Are the tasks consistently beyond her current competence?

Does the interaction occur when she is exhausted?

Is there an established history of conflict around academic demands?

These questions are more useful than assuming the teenager simply dislikes being wrong.

They also avoid the opposite mistake of assuming that every refusal is caused by anxiety.

Sometimes a learner is avoiding an unpleasant but manageable task. Sometimes the task is genuinely inaccessible. Sometimes both processes are involved.

The appropriate response depends on the mechanism.

8. Teach the learner to use feedback independently

So far, the parent has been the person identifying errors and selecting corrections.

That may be necessary initially.

But the book’s long-term objective is not to produce a parent who becomes exceptionally skilled at correcting every piece of work.

It is to develop a learner who increasingly recognises and corrects her own errors.

This requires explicit teaching.

A teenager cannot be expected to develop effective self-correction merely because adults have corrected her work many times.

She needs opportunities to practise the process herself.

Stage 1: The adult identifies the error

The parent says:

“You’ve identified the feature but not explained its function.”

The teenager makes the correction.

At this stage, the adult is doing most of the diagnostic work.

Stage 2: The adult identifies the location

The parent says:

“There’s something missing in your second sentence. Compare it with the question.”

The teenager identifies the missing explanation.

The adult is supplying less information.

Stage 3: The adult prompts a check

The parent says:

“Check whether every feature you’ve named is linked to its function.”

The teenager inspects the answer and corrects it.

Now she is beginning to apply a checking strategy.

Stage 4: The learner uses a checklist

Before submitting an answer, the teenager checks:

• Have I answered the actual question?
• Have I included the required knowledge?
• Have I explained the relationship rather than merely naming facts?
• Have I checked for obvious mistakes?

The checklist externalises the checking process.

It may remain useful indefinitely.

Stage 5: The learner chooses when and how to check

The teenager recognises that an extended Biology explanation requires a causal-link check, whereas a simple recall question may require only factual verification.

She chooses the appropriate checking strategy.

This is a more advanced form of competence.

She is not merely following instructions.

She understands when the strategy is useful.

The progression is not necessarily linear.

A teenager may independently check Biology explanations while still needing substantial assistance with English Literature essays.

Support should respond to competence in the particular task.

And external checklists, calendars and other organisational tools should not be removed simply because the learner is becoming more independent.

The aim is increasing control of the checking process, not the elimination of every external aid.

9. Build a correction routine that does not dominate the day

Home education creates an unusual situation.

The same adult may be responsible for teaching, planning, marking, encouraging, organising and maintaining the parent-child relationship.

Without clear boundaries, correction can spread across the entire day.

A casual conversation about History becomes a quiz.

A discussion at lunch becomes an opportunity to revisit an error.

A teenager’s explanation of something interesting becomes another moment for improvement.

Eventually, the learner may feel that every interaction is being assessed.

The parent may feel equally trapped, constantly monitoring for missed opportunities to teach.

A better approach is to give correction a predictable place within learning.

For example, a twenty-minute study session might contain:

First five minutes: Retrieve a small amount of previously learned material.

Next ten minutes: Complete a focused task.

Final five minutes: Review one important error, correct it and record anything that needs revisiting.

The timings are illustrative, not a research-established optimum.

Some sessions will need longer instruction. Others may involve no correction at all.

The important feature is predictability.

The teenager knows when feedback is coming and what it is for.

The parent knows that not every mistake requires immediate intervention.

The one-priority rule

When a learner is easily overwhelmed by correction, begin by identifying the single most important improvement.

For an English Literature paragraph, that might be explaining the quotation.

For a History answer, it might be connecting evidence to the argument.

For Mathematics, it might be choosing the correct procedure.

For Biology, it might be making a causal explanation explicit.

Other errors can be addressed later.

This is a practical workload-management strategy, not a claim that research has established one correction as universally optimal.

Its value is that it keeps feedback focused and usable.

10. The difference between correction and learning

A teenager completes a worksheet containing ten questions.

She answers six correctly.

The parent explains the remaining four.

The teenager copies the correct answers.

The worksheet is now complete.

What has been achieved?

The written work has been corrected.

But we do not yet know whether the teenager has learned the missing material.

This distinction is essential.

A completed correction may demonstrate compliance or immediate understanding.

It may not demonstrate retention.

A successful answer with substantial prompting may not demonstrate independent performance.

And successful performance on a familiar problem may not demonstrate transfer to an unfamiliar one.

To evaluate feedback properly, we need to distinguish several outcomes.

These outcomes are related but not interchangeable.

Suppose the teenager corrects her Biology explanation today.

Tomorrow, she answers a similar question without help.

Next week, she correctly explains how a root hair cell is adapted to its function.

Later, she independently checks whether her explanation includes a link between structure and function.

Each success provides evidence of something increasingly valuable.

The original feedback has not merely repaired one answer.

It has contributed to a reusable academic skill.

That is the longer-term objective.

11. A worked example: improving a History answer

Consider this GCSE History question:

Explain why tensions between Austria-Hungary and Serbia increased between 1908 and 1914.

A teenager writes:

Austria-Hungary annexed Bosnia in 1908. Serbia was angry. Then Archduke Franz Ferdinand was assassinated in 1914.

The answer contains relevant facts.

But it does not sufficiently explain the relationships between them.

A parent could say:

“You need more detail.”

Instead, the parent identifies the missing reasoning.

“You’ve chosen two relevant events. The next step is to explain why each event increased tension.”

The parent models the first connection:

Austria-Hungary’s annexation of Bosnia angered Serbia because Serbian nationalists wanted closer ties with, or control over, territories containing South Slavic populations. The annexation therefore intensified hostility towards Austria-Hungary.

The parent then asks:

“Can you explain why the assassination made the relationship even more dangerous?”

The teenager tries:

“Austria-Hungary blamed Serbia, which made them more hostile.”

“That’s a useful start. What did Austria-Hungary do next?”

The teenager consults her notes.

“They issued an ultimatum.”

“And why did that matter?”

“Because Serbia didn’t accept all the demands, and Austria-Hungary declared war.”

The parent helps her combine these ideas into a causal explanation.

The teenager then rewrites the answer.

The next day, she receives a different question:

Explain why the Bosnian Crisis increased tensions in Europe.

This time, she must retrieve and organise the relevant knowledge herself.

The adult checks whether she can construct the causal relationship without being led through every step.

A week later, she encounters a broader question about the causes of the First World War.

Now she must decide which evidence is relevant and how to organise it.

The instructional support changes as the learner’s competence develops.

At first, the adult supplies the missing reasoning.

Later, the learner supplies it.

Eventually, she recognises independently that historical explanation requires more than a list of events.

This is how feedback can contribute to transfer.

12. A worked example: recovering from a failed Mathematics attempt

A teenager is learning to solve simultaneous equations.

She has studied worked examples and completed several guided problems.

Now she attempts:

\[ 2x+y=11 \]

\[ x-y=1 \]

She adds the equations incorrectly and obtains:

\[ 3x+2y=12 \]

The parent notices.

Instead of immediately supplying the answer, the parent asks:

“What happens to \(+y\) and \(-y\) when you add the equations?”

The teenager looks again.

“They cancel.”

“So what should the new equation be?”

\[ 3x=12 \]

She continues:

\[ x=4 \]

Substituting into the second equation:

\[ 4-y=1 \]

\[ y=3 \]

The parent asks her to check both original equations.

She verifies that the values work.

This is a successful correction.

But what should happen next?

The parent gives a similar problem.

If the teenager solves it correctly, the adult may reduce assistance.

If she makes the same error again, a more explicit explanation of elimination may be needed.

If she can solve similar problems but fails when subtraction is required instead of addition, she needs practice recognising which operation eliminates a variable.

The important point is that the adult should not simply celebrate the corrected answer and move on.

The correction has revealed a learning requirement.

That requirement should guide the next task.

13. A worked example: when feedback itself becomes the barrier

A teenager is writing an English Literature paragraph.

She has already spent considerable effort choosing a quotation and organising her thoughts.

The parent reads the paragraph and identifies several weaknesses.

The teenager becomes upset.

“I spent ages on that.”

The parent replies:

“I’m only trying to help.”

The teenager says:

“Nothing I do is ever good enough.”

The immediate temptation is to explain that the feedback is constructive.

But the teenager is no longer discussing the paragraph.

She is expressing something about her experience of the interaction.

The parent needs to distinguish the academic issue from the relational issue.

A possible response is:

“You’ve put a lot of work into this, and I can see that. I gave you too many corrections at once. The main thing we need to improve is how you explain the quotation. We can look at that together later.”

This response does not pretend the paragraph is already excellent.

It does not withdraw the academic expectation.

It acknowledges that the method of correction was poorly matched to the moment.

Later, the parent and teenager agree on a different feedback system.

The parent will identify one priority improvement.

The teenager will first compare her paragraph with a model.

She can request a verbal explanation if she needs one.

Over time, they review whether the system is helping.

The important measure is not whether the teenager appears pleased with every correction.

It is whether she can increasingly receive, understand and use accurate feedback without the process repeatedly breaking down.

14. The parent’s practical toolkit

The following routine can be used across GCSE subjects.

The Five-Minute Correction

An adaptable routine, not a fixed time requirement.

1. Identify one useful improvement.

“Your answer includes the facts, but the explanation is missing.”

2. Diagnose what is needed.

“Do you understand the relationship, or would an example help?”

3. Supply appropriate assistance.

Cue, explanation, model or worked step.

4. Ask the learner to act.

Rewrite, explain, calculate, compare or apply.

5. Plan a later check.

“We’ll try a similar question on Thursday without the example.”

A useful record might contain three short entries:

What was difficult? Explaining how an adaptation supports a function.

What helped? A model showing feature → effect → function.

What will we check later? A new question about a different specialised cell.

This record can eventually be maintained by the teenager.

The goal is not to build an elaborate tracking system that creates more work than it saves.

It is to preserve enough information to make future teaching more effective.

15. When to increase or reduce assistance

One danger in supportive teaching is providing so much help that the learner never needs to make an independent decision.

Another danger is withdrawing help before the learner has developed the necessary knowledge.

Neither approach serves independence.

A better principle is:

Adjust assistance to demonstrated competence.

Increase assistance when the learner repeatedly makes errors that reveal missing knowledge, cannot identify the next step or is relying on guesses.

Reduce instructional assistance when she consistently succeeds, understands why a method works and can apply it without the adult supplying every decision.

But distinguish instructional assistance from accommodations.

A teenager may no longer need a worked example for a familiar mathematical procedure while still benefiting from written instructions, a quiet environment or a visual timetable.

These supports serve different functions.

A checklist may remain useful even when the learner independently operates it.

An accommodation does not become unnecessary simply because academic competence improves.

Likewise, asking for help can be an advanced skill rather than a sign of dependency.

A teenager who says, “I understand the algebra, but I need you to check whether I’ve chosen the right elimination method,” is demonstrating a more precise understanding of her difficulty than one who simply says, “I can’t do it.”

The long-term objective is to transfer control of the learning process.

Not to require the learner to do everything alone.

16. What to do tomorrow

Choose one subject in which correction regularly becomes difficult.

Do not redesign the entire educational programme.

Begin with a single task.

Before the session, decide what the learner is supposed to learn or demonstrate.

During the task, allow an appropriate independent attempt.

When an error appears, identify the most important gap.

Choose one response that directly addresses it.

Ask the learner to use the feedback.

Then record a short follow-up question for a later session.

After several attempts, review what happened.

Did the teenager become more willing to attempt the work?

Did she produce better answers immediately?

Could she reproduce the improvement after a delay?

Could she apply the knowledge to a different question?

Did she begin identifying errors herself?

Did the feedback process remain manageable for both learner and adult?

These questions provide a much more useful picture than simply counting how many worksheets were completed.

They also help prevent a common mistake: treating a pleasant, conflict-free correction session as proof that learning has occurred.

A comfortable session is valuable.

But the educational objective is still to build knowledge and skill.

Equally, a technically correct explanation that repeatedly produces distress and disengagement may be ineffective in practice.

Good teaching must attend to both.

Conclusion: The learner who knows how to recover

The aim of education is not to produce a teenager who never makes mistakes.

That would be neither realistic nor desirable.

The aim is to produce a teenager who can encounter an error without treating it as the end of the task.

A learner who can say:

That answer isn’t right.

I need to work out why.

Perhaps I’ve misunderstood the question.

Perhaps I’ve forgotten something.

Perhaps I need a different strategy.

Perhaps I need an example or some help.

I can correct this and try again.

This is a substantial achievement.

It requires subject knowledge, useful strategies, experience of successful correction and an environment in which errors can be examined without unnecessary humiliation or conflict.

It also requires adults to resist two temptations.

The first is to rescue the learner from every difficulty.

The second is to insist that difficulty must always be overcome without help.

Neither develops the full range of capabilities we want.

The more useful approach is to make errors understandable, corrections actionable and improvement visible.

A teenager who has learned to recover from mistakes has acquired something more valuable than a perfectly marked worksheet.

She has begun to develop a reliable way of responding when learning becomes difficult.

And that ability will matter long after the GCSE examinations are over.

Evidence notes

The chapter draws on three overlapping bodies of evidence.

General educational evidence: Feedback can improve learning, but its effects depend on content, timing, context and whether the learner can act upon it. General research on explicit instruction, worked examples, expertise reversal and retrieval practice supports adapting assistance to prior knowledge, requiring active correction and checking later retention.

Autism-specific and near evidence: Research on autistic students’ educational experiences supports attention to relationships, communication, sensory conditions and wellbeing. Executive-function research establishes plausible barriers to organising and monitoring work, without determining the needs of any particular learner. The 2026 experiments by Cairney, Haebig and Lucas provide emerging evidence that autistic adults can benefit from retrieval practice and corrective post-retrieval feedback, although this does not establish effects on GCSE attainment.

Practical synthesis: The Feedback-to-Learning Cycle, Five-Minute Correction and staged transfer of self-correction are integrations of these findings. They have not been tested together as a complete intervention for autistic GCSE-age learners.

Key references

• Wisniewski, B., Zierer, K., & Hattie, J. (2020). The Power of Feedback Revisited: A Meta-Analysis of Educational Feedback Research. Frontiers in Psychology. DOI: 10.3389/fpsyg.2019.03087.
• Kasztelan Chapman, H., Southall, A., & Grieshaber, S. (2025). What autistic students say about their learning and teaching experiences in mainstream secondary school: a scoping review. The Australian Educational Researcher. DOI: 10.1007/s13384-025-00838-0.
• Cristescu, L., Papoudi, D., & Douglas, G. (2025). A Systematic Review of Interventions on Self-Determination for Autistic Individuals and Those with a Learning Disability: an Ecological Approach. DOI: 10.1007/s40489-025-00508-y.
• Executive function in children with neurodevelopmental conditions: a systematic review and meta-analysis. (2024). Nature Human Behaviour. DOI: 10.1038/s41562-024-02000-9.
• Education Endowment Foundation. Metacognition and Self-Regulated Learning: Guidance Report.

Evidence audit

Evidence debt

Before publication, the following points need further verification or development:

1. Verify the exact results and methodological limitations of the Maras, Gamble and Brosnan study on metacognitive monitoring and feedback in autistic secondary-school mathematics.
2. Incorporate the detailed findings of Cairney, Haebig and Lucas (2026), distinguishing benefits of retrieval attempts from benefits of post-retrieval feedback.
3. Examine autism-specific research on responses to corrective feedback, especially whether feedback modality, timing or perceived evaluation affects learning outcomes.
4. Check whether there are controlled secondary-school studies comparing learner-generated corrections with supplied correct answers in autistic populations.
5. Confirm the precise applicability of feedback meta-analyses to Key Stage 4 and avoid implying that average effects transfer directly to autistic GCSE learners.

Editorial self-review

The chapter deliberately focuses on what happens after an attempt, advancing beyond the earlier chapters on explicit instruction, worked examples and retrieval practice.

The recurring practical decision is whether the learner needs a cue, correction, explanation, model, accommodation or opportunity to try again.

The examples distinguish immediate improvement from retention, transfer and independent self-correction. The proposed developmental sequence is presented as a practical synthesis rather than a validated autism-specific programme.

The most important editorial principle for the final manuscript is to preserve the chapter’s central argument:

The purpose of correction is not to establish that the learner was wrong. It is to help the learner become more capable of getting it right.