There is a sentence every Mathematics teacher hears.
“I don’t understand.”
It can mean almost anything.
I do not understand the question.
I do not understand the topic.
I do not know which formula to use.
I know the formula but I cannot make the algebra work.
I know what to do but I cannot see why it is allowed.
I understand the worked example but I cannot start this one.
I was fine until this line.
I have two possible methods and I do not know which is better.
To a student, all of these can feel like the same experience.
Stuck.
But they are not the same educational problem.
And after many years of teaching Mathematics, I have become increasingly interested in what a student says immediately after:
“I don’t understand.”
Because there is a considerable difference between a student who can only say:
“I don’t know anything,”
and one who can say:
“I know these are simultaneous equations, and I know I need to eliminate one variable. I just cannot see which equation I should multiply first.”
Both students are stuck.
The second student is much less lost.
That matters.
The direct answer
One of the signs that a student is becoming mathematically independent is that she can locate her uncertainty more precisely.
She does not need to know the answer.
She does not even need to know the next step.
But she can increasingly distinguish:
what she understands,
what she has established,
what she is considering,
and exactly where confidence ends.
That ability makes teaching more efficient because help can reach the right level.
It makes revision more efficient because the student stops treating an entire chapter as weak when only one relationship is unstable.
And eventually it gives her something much more valuable:
a way of continuing to think when the teacher is not beside her.
A student who can say:
“This is the last part I know is true”
has somewhere to restart.
Consider a simultaneous-equations question
Suppose:
3x + 2y = 17
and:
5x − 2y = 7.
A student looks at the page and says:
“I don’t understand simultaneous equations.”
That is possible.
Perhaps the whole idea genuinely is weak.
But I ask:
“What do you recognise?”
She says:
“There are two equations and two unknowns.”
Good.
“What are we trying to find?”
“x and y.”
“Do you know either of the usual methods?”
“Elimination and substitution.”
“Anything interesting about these equations?”
She looks again.
“The y terms are +2y and −2y.”
Now something has changed.
Perhaps the problem was never:
“I do not understand simultaneous equations.”
Perhaps the student understood almost the entire architecture and simply had not noticed that adding the equations removes y:
8x = 24
so:
x = 3.
Then substitute:
3(3) + 2y = 17
giving:
2y = 8
and:
y = 4.
The amount of teaching required was tiny.
The original statement of difficulty was enormous.
That mismatch interests me.
Broad uncertainty creates broad intervention
If a student says:
“I am bad at algebra,”
the natural response is large.
More revision.
More worksheets.
Reteach the chapter.
Return to fundamentals.
Sometimes that is necessary.
But suppose the actual weakness is:
“I understand solving equations until a negative sign appears outside a bracket.”
That is a very different repair.
Take:
5 − 2(x − 3) = 11.
Perhaps the student writes:
5 − 2x − 6 = 11
instead of:
5 − 2x + 6 = 11.
Now we have something precise.
The issue is not “algebra”.
It is the distribution of a negative factor.
We can isolate it:
−2(x − 3)
means:
−2x + 6
because:
(−2)(−3) = +6.
Then test:
−3(x + 4)
−5(2x − 1)
4 − 2(3x − 7).
A large academic anxiety has become a small mathematical object.
That is almost always better.
Precision does not make the weakness disappear
I am not arguing for reassuring labels.
A student may identify precisely that she does not understand the cosine rule.
That is still a genuine gap.
But now we know what to teach.
Or she may say:
“I can use the cosine rule when you tell me to, but I cannot recognise when it belongs.”
That is different again.
The formula is present.
The selection conditions are not sufficiently organised.
The repair might therefore begin with comparing cases.
Three sides known, find angle.
Two sides and included angle known, find third side.
Then contrast with sine-rule situations.
The distinction affects the lesson.
Without it, we may simply practise the formula she already knows.
A-Math makes the value of precise uncertainty even clearer
Suppose the question says:
For the curve:
y = x³ − 3x² + 2,
find the equation of the tangent at x = 2.
A student says:
“I don’t know differentiation.”
Perhaps.
So I ask:
“Can you differentiate the function?”
She writes:
dy/dx = 3x² − 6x.
Correct.
“What does that give us?”
“The gradient.”
“At x = 2?”
She substitutes:
3(4) − 12 = 0.
“So what don’t you know?”
Now she says:
“I have the gradient, but I don’t know how to get the equation of the tangent.”
Excellent.
Not excellent that she is stuck.
Excellent that the problem now has an address.
She does not need differentiation reteaught.
She needs the connection between:
a point,
a gradient,
and the equation of a line.
Find the point on the original curve:
at x = 2,
y = 8 − 12 + 2 = −2.
So the tangent passes through:
(2, −2)
with gradient:
0.
Therefore:
y = −2.
One missing bridge.
Not an entire weak topic.
This is why I sometimes ask for the last secure line
When a student is stuck, I often find this more useful than:
“Which topic is this?”
I ask:
“What is the last thing on the page that you are sure is correct?”
Perhaps:
“I know the midpoint is (4, 7).”
Good.
Or:
“I know the derivative is 6x − 4.”
Good.
Or:
“I know these triangles are similar.”
Good.
Or:
“I know the total probability has to be 1.”
Good.
Then:
“What do you need next?”
The gap between those two statements is often the teaching problem.
This is much more precise than restarting the entire solution.
A secure line gives the student a foothold
Imagine a coordinate-geometry question.
The student has already found:
midpoint:
M(5, 9).
Gradient of AB:
2.
She needs the equation of the perpendicular bisector.
She says:
“I don’t know what to do.”
But she actually knows two important things.
The required line passes through M.
And its gradient must be perpendicular to 2.
So:
m = −1/2.
Then use point-gradient form:
y − 9 = −1/2(x − 5).
The student did not need a complete solution supplied.
She needed to see how already-secure information constrains the next move.
This is one reason I want students to treat their own working as evidence.
Earlier lines should remain available.
There is also a difference between not knowing and choosing between possibilities
Suppose a student is solving:
x² − 5x + 6 = 0.
She says:
“I don’t know what method to use.”
I ask:
“What methods could solve a quadratic?”
“Factorisation. Quadratic formula. Completing the square.”
That answer is important.
The student is not empty.
She has a choice problem.
Now inspect the coefficients.
Can we find two numbers whose product is 6 and sum is −5?
−2 and −3.
So:
(x − 2)(x − 3) = 0.
Factorisation is cheap.
The student’s uncertainty was not:
“What is Mathematics?”
It was:
“Which available route is appropriate here?”
This is a much more advanced kind of uncertainty.
Students should recognise that.
Parents can easily mistake hesitation for absence of knowledge
A child looks at a question for thirty seconds.
Nothing is written.
It can appear that she knows nothing.
But internally she may be doing something useful.
Is this sine rule or cosine rule?
Can I factorise this or should I use the formula?
Do I need the midpoint or gradient first?
Is this a percentage increase or percentage-point change?
Is the negative root allowed?
That hesitation may be mathematical judgement.
The question is whether the thinking is narrowing.
If after two minutes the student is still cycling through unrelated ideas, intervention may help.
If she can explain:
“I think there are two possible methods; I am deciding which uses the information more directly,”
I am much less worried.
Silence itself is not diagnostic.
This is also why speed can be misleading
A student who immediately launches into a familiar procedure can look confident.
Sometimes she is.
Sometimes she has simply recognised a surface pattern.
Another student pauses.
She checks the conditions.
Then chooses correctly.
The second student may take longer at first.
But the pause contains useful reasoning.
I therefore do not want parents interpreting every hesitation as weakness or every quick start as mastery.
What matters is what the student is doing with the uncertainty.
A useful question is: “What are the two things you are deciding between?”
This works particularly well when students are not completely lost.
Suppose a triangle question gives:
two sides,
and one angle.
Student hesitates.
Ask:
“What are your possible routes?”
“Sine rule or cosine rule.”
Good.
“What information does each need?”
Now she looks at the configuration.
If the known angle is included between the two known sides and the third side is required:
cosine rule is natural.
The student has made the decision herself.
That is preferable to simply hearing:
“Use cosine rule.”
The final calculation may be identical.
Ownership of the route is not.
Sometimes the student is confidently uncertain in the wrong place
There is a boundary here.
Students are not always accurate judges of their own difficulty.
A student may say:
“My algebra is fine. I just don’t understand the question.”
Then we inspect the working and discover that the interpretation was correct; the algebra contains repeated sign errors.
Or she says:
“I only made a careless mistake.”
But the same mistake appears five times.
So self-diagnosis is evidence.
It is not infallible.
We compare what the student thinks is wrong with what the working actually shows.
That comparison itself can teach something.
Language ability can also affect how precise uncertainty sounds
A quiet student may understand more than she can explain verbally.
A student working in a second language may struggle to name the mathematical difficulty despite seeing it.
Another student may be highly articulate and give a beautiful description of a concept she still cannot execute reliably.
So I would never grade mathematical understanding by eloquence.
The aim is not sophisticated vocabulary.
A simple statement is enough:
“I know up to here.”
“I don’t know why we can cancel this.”
“I can differentiate, but I don’t know what to do with the derivative.”
“I think it is similar triangles, but I cannot prove they are similar.”
Those sentences are educationally useful.
That is the standard.
The student’s pencil can sometimes speak when she cannot
If verbal explanation is difficult, I may ask her to mark the working.
Tick the last line she trusts.
Circle the part of the question she cannot interpret.
Put a question mark beside the first uncertain transformation.
Draw two arrows to the possible methods she is considering.
This often produces the same diagnostic value without requiring a long conversation.
Mathematics has the advantage of leaving a visible trail.
We can use it.
One of the most important distinctions is “I don’t know why” versus “I don’t know how”
Suppose:
2(x + 3) = 14.
Student expands:
2x + 6 = 14.
She can continue.
But asks:
“Why can we divide both sides by 2?”
That is a conceptual question about preserving equality.
Different from:
“I know we divide by 2, but I keep dividing only one term.”
That is execution.
And different again from:
“I understand the equation, but I don’t know whether expanding is the best first move.”
That is method selection.
The same written question can expose different weaknesses.
The student’s description helps us locate them.
“I can follow it when you do it” is another useful sentence
Parents often hear this.
It deserves careful interpretation.
Following a solution means the student can recognise the validity of steps once presented.
That is real understanding at one level.
Producing the route independently is another level.
Suppose I write:
(x − 4)² = 9
then:
x − 4 = ±3
so:
x = 7 or 1.
A student may look and say:
“Yes, I understand.”
Now remove the solution.
Give:
(x + 2)² = 16.
Can she independently produce:
x + 2 = ±4
and therefore:
x = 2 or −6?
If not, we have located something.
Recognition is present.
Independent retrieval is not yet stable.
That is a much better diagnosis than saying she “doesn’t understand”.
This precision makes tuition calmer
A student who thinks:
“I am weak at A-Math,”
is facing an enormous object.
Where should she begin?
Differentiation?
Quadratics?
Trigonometry?
Coordinate geometry?
A student who can say:
“I can differentiate basic functions, but I become lost when the derivative has to be used to find a tangent,”
has a much smaller problem.
Small does not mean trivial.
It means tractable.
Good diagnosis should make the next lesson more obvious.
It also makes parent decisions more rational
Suppose school marks fall.
The broad conclusion is:
“She needs more Mathematics tuition.”
Perhaps.
But useful questions come first.
Is she missing concepts?
Can she choose methods?
Is working slow but accurate?
Does she understand during lessons but fail on delayed mixed questions?
Are the same execution errors recurring?
Does she freeze on unfamiliar wording?
Can she solve once the first step is supplied?
Different patterns imply different forms of support.
The mark tells us that performance is below the desired level.
It does not by itself tell us why.
The quality of the student’s own uncertainty can add valuable information.
There is an emotional reason this matters during adolescence
Teenagers are very good at turning local academic difficulty into identity.
“I don’t understand this question”
becomes:
“I don’t understand algebra.”
Then:
“I am bad at Math.”
Then sometimes:
“I have never been a Math person.”
The statements grow faster than the evidence.
Precise uncertainty pushes in the opposite direction.
Not:
“I cannot do Mathematics.”
But:
“I keep losing the sign when the negative is outside a bracket.”
Not:
“I don’t understand trigonometry.”
But:
“I can calculate once I know the ratio, but I am still unreliable at identifying opposite and adjacent sides after the triangle is rotated.”
Not:
“I am terrible at A-Math.”
But:
“I understand differentiation, but I do not yet recognise when a stationary-point question requires me to solve dy/dx = 0.”
These statements are narrower.
They are also much more actionable.
Precision should not become self-criticism
There is another boundary.
I do not want students dissecting every mistake with anxiety.
“What exactly is wrong with me?”
That is not the purpose.
The object of diagnosis is the Mathematics, not the child’s worth.
We are asking:
What relationship is currently unstable?
What step is not automatic yet?
What cue is missing?
What condition has not been recognised?
The tone matters.
The goal is navigational clarity.
Teachers can accidentally prevent this development by rescuing too quickly
Student:
“I don’t know.”
Teacher:
“Use the sine rule.”
Problem solved.
Perhaps efficiently.
But we lost information.
What did the student know before the hint?
Maybe she knew it was trigonometry.
Maybe she had narrowed it to sine or cosine rule.
Maybe she had actually selected sine rule but forgotten how to rearrange it.
A large hint can erase the location of the original difficulty.
This is why small hints are diagnostically useful.
They tell us how far the student was from independent control.
I prefer the smallest hint that restores movement
If the student is stuck, perhaps ask:
“What are you trying to find?”
Still stuck.
“What information do you have?”
Then:
“Which relationship connects those quantities?”
Only if necessary:
“Could the cosine rule connect them?”
The size of the hint matters.
If one small question restarts the whole solution, the gap was narrow.
If even a large hint produces no movement, the concept may need rebuilding.
This gives us useful information while teaching.
The amount of help required can become a measure of progress
Suppose in Week 1:
“Use simultaneous equations.”
Week 2:
“How many unknowns do you have, and how many independent equations?”
Week 3:
“What are you looking for?”
Week 4:
Nothing.
The student identifies the structure herself.
Her final answer may have been correct in every week.
But the amount of external navigation has changed.
That is progress.
I think parents sometimes miss this because marks have not moved dramatically yet.
The child is becoming more independent before the result visibly catches up.
A powerful exercise is to stop before solving
Take five questions from different topics.
The student does not need to finish them.
For each, write only:
What I know.
What I need.
Where I am uncertain.
For example:
Question:
Find the equation of the perpendicular bisector of A(2,3) and B(8,15).
Student:
I know I need a line equation.
I know the perpendicular bisector goes through the midpoint.
I can find the midpoint.
I can find the gradient of AB.
I am unsure how to turn that gradient into the perpendicular gradient.
That is excellent diagnostic information.
Now the repair is one relationship.
For non-vertical perpendicular lines:
m₁m₂ = −1.
The rest may already be secure.
Then move the same uncertainty to a different surface
After repair, do not repeat the identical coordinate question only.
Change the coordinates.
Then use a line already given in gradient form.
Then perhaps embed perpendicular gradients inside another geometry problem.
If the student can locate and use the relationship after the surface changes, the repair is beginning to transfer.
If she only succeeds on the original corrected question, the learning may still be too local.
This is why diagnosis and transfer belong together.
The same method works with A-Math
Question:
Find the minimum point of:
y = x² − 6x + 11.
Student:
“I know a minimum relates to the turning point.”
Good.
“What methods could find it?”
“Complete the square, or differentiate.”
Good.
“Which can you do?”
“I can complete the square.”
Then:
y = (x − 3)² + 2.
Minimum point:
(3,2).
The student was not stuck in the concept.
She was temporarily uncertain about method selection.
Now ask another question where differentiation is more natural.
The uncertainty has to respond to structure.
Parents can use four quiet questions
When a child says:
“I don’t understand,”
there is no need to launch into teaching.
Ask:
What do you understand so far?
What is the last thing you are sure about?
What are you trying to find next?
What exactly are you unsure about?
Often, that is enough to reduce the problem.
If the child cannot answer any of them, that is also information.
The concept may need to be rebuilt earlier.
But if she can answer three and not the fourth, we have a much narrower starting point.
Do not insist on the correct diagnosis immediately
A student may first say:
“I don’t know what formula.”
Then after discussion realise:
“Actually I knew the formula. I didn’t understand what the diagram meant.”
Good.
Self-diagnosis itself is learnable.
The first description need not be perfect.
What matters is that the student becomes increasingly able to revise it based on evidence.
That is a sophisticated intellectual habit.
This is where correction becomes more powerful
After an error, do not ask only:
“What is the correct answer?”
Ask:
“What did you think the problem was?”
Then:
“Where did that understanding stop matching the Mathematics?”
Perhaps she selected the wrong theorem.
Perhaps the theorem was right and the algebra failed.
Perhaps the method was right but a condition made one answer invalid.
Now the correction belongs to a cause.
This is much more likely to help the next unfamiliar question.
Examination pressure tests this ability differently
During an exam, there is no tutor available to ask diagnostic questions.
The student has to do some of this internally.
I know what this question asks.
I know these relationships.
I am stuck converting the wording into an equation.
Or:
I have the equation; the algebra is the problem.
Or:
I know the method, but this arithmetic is becoming expensive.
Or:
I do not currently have a route; leave it and return later.
This is what independent examination judgement looks like in practice.
Not perfect certainty.
Useful localisation of uncertainty.
Strong students need this perhaps more than anyone expects
A high-performing student is accustomed to knowing.
When she does not know, the experience can feel unusually threatening.
She may respond by guessing quickly, hiding the difficulty, or refusing to articulate it.
But difficult Mathematics eventually guarantees uncertainty.
IP, IB, IGCSE and strong A-Math work are not improved by pretending otherwise.
The mature response is not:
“I should never be unsure.”
It is:
“I can be unsure precisely.”
That is a much stronger position.
There is a larger intellectual lesson here
Experts are not people who never say:
“I don’t know.”
Often they are unusually careful about what they do and do not know.
They distinguish:
known,
likely,
possible,
unsupported,
and unresolved.
Mathematics gives adolescents a clean place to practise that discipline.
This line is proved.
This one is a guess.
This method might work.
This quantity is known.
This relationship is missing.
This result needs checking.
We do not have to make the classroom philosophical.
The habit emerges naturally from good mathematical work.
The useful next route
Take one Mathematics question your child is currently stuck on.
Do not show the solution immediately.
Ask her to place a mark beside the last line or fact she trusts.
Then ask:
“What would you need to know next for the problem to move?”
If she can answer that, let her try to retrieve it.
If she cannot, offer one small cue.
Not the whole method.
Enough to narrow the uncertainty.
After the question is complete, ask:
“What were you actually stuck on?”
Write one sentence.
For example:
“I knew this was a quadratic, but I did not recognise that it was a difference of two squares.”
Or:
“I could find the gradient, but I forgot how to form the equation of a line through a point.”
Or:
“I understood the percentage calculation, but I treated a 20% increase as adding 20 rather than multiplying the original amount by 1.2.”
Then, several days later, use a new question containing the same demand.
See whether the student can identify the difficulty sooner—or no longer encounters it.
That is a useful measure.
What long teaching has made me notice
There is a temptation in education to think that confidence means saying:
“I know.”
Sometimes it does.
But I think there is another kind of confidence that becomes increasingly important as students grow older.
It can say:
“I know this part.”
“I do not know that part yet.”
“I have two possible methods.”
“I can justify this line but not the next one.”
“I think my answer is wrong, and this is why.”
“I need one missing relationship before I can continue.”
That is not weakness.
It is intellectual resolution.
The student’s uncertainty has acquired edges.
And once uncertainty has edges, it becomes easier to work with.
A teacher can help more precisely.
A parent can worry less broadly.
The student can revise more intelligently.
A difficult chapter stops being one dark object and becomes several smaller relationships, some secure and some still under construction.
I think that matters particularly during adolescence.
School becomes more demanding at exactly the time young people become increasingly conscious of what their performance says about them.
Mathematics can feel personal very quickly.
One difficult question becomes:
“I cannot do this.”
Then:
“I cannot do Mathematics.”
Good diagnosis interrupts that collapse.
Perhaps the student cannot do one thing.
Good.
Let us find the one thing.
Then teach it properly.
And eventually, I want the teacher’s diagnostic questions to become the student’s own.
She meets an unfamiliar problem.
She does not know the answer.
That no longer means she has no direction.
She can ask:
What is known?
What have I established?
Where exactly does certainty stop?
What would I need next?
What could I test?
That is a remarkably useful form of independence.
Not the absence of uncertainty.
The ability to navigate inside it.
So when a student tells me:
“I don’t understand,”
I rarely hear the conversation as finished.
I hear an opening.
And the next sentence matters enormously.
Because sometimes the distance between:
“I don’t understand any of this”
and:
“I know everything up to this exact line”
is not merely better language.
It is evidence that the student is beginning to know how to find her own way through Mathematics.