Sometimes a student is stuck on a question containing six or seven numbers.
She has underlined them.
Copied them carefully.
Perhaps entered two of them into the calculator.
The page is already becoming busy.
Then I cover part of the question with my hand and ask:
“Forget the numbers for a moment. What is happening?”
There is often a pause.
Then something changes.
“Oh. It is just a quadratic.”
Or:
“These two lines are parallel.”
Or:
“The same expression appears twice.”
Or:
“This is really a scale-factor question.”
The Mathematics was present from the beginning.
The numbers were not wrong.
They were not unnecessary.
But they had become so visually dominant that the student could no longer see the relationship they belonged to.
After many years of teaching Mathematics, I have become quite interested in this problem.
Students often assume that more numerical information makes a question more concrete and therefore easier.
Sometimes it does.
But sometimes numerical detail does the opposite.
It occupies attention.
It encourages premature calculation.
It makes two structurally identical questions look unrelated.
And it can persuade a student to start operating before she has decided what mathematical object she is looking at.
This is why I occasionally ask a student to do something that sounds strange:
Take the numbers out temporarily, so we can see the Mathematics they are sitting on.
The direct answer
When a student repeatedly knows the required methods but fails to recognise when to use them, the problem may not be lack of knowledge.
The mathematical structure may be hidden beneath surface detail.
Numbers, unfamiliar letters, contexts, units and diagrams can all change while the underlying relationship remains the same.
A stronger Mathematics student gradually learns to separate these two layers:
the surface information of the particular question,
and the structure that determines which mathematical ideas apply.
This does not mean ignoring numbers.
The numbers must eventually return.
They determine the actual answer.
They may create restrictions, signs, scales and special cases.
But method selection often becomes easier if the student first asks:
What relationship would still be here if these particular numbers changed?
That is one of the foundations of transfer.
A familiar quadratic can become invisible very quickly
Consider:
2x² − 5x + 2 = 0.
Most A-Math students eventually recognise this as a quadratic equation.
Factorisation gives:
(2x − 1)(x − 2) = 0.
So:
x = 1/2
or:
x = 2.
Now consider:
2sin²x − 5sin x + 2 = 0.
The student may suddenly say:
“I don’t know which trigonometry formula to use.”
That response is useful.
Because, at this stage, the main difficulty may not be trigonometry at all.
The expression has the same structure:
2(something)² − 5(something) + 2 = 0.
If we temporarily let u = sin x, the equation becomes:
2u² − 5u + 2 = 0.
Now the familiar quadratic reappears.
Factorise:
(2u − 1)(u − 2) = 0.
Therefore u = 1/2 or u = 2.
Restore the original meaning:
sin x = 1/2
or:
sin x = 2.
The second possibility is impossible for real angles because −1 ≤ sin x ≤ 1.
So the trigonometric work begins only after the quadratic structure has been recognised.
The substitution did not magically solve the question.
It removed a distracting surface long enough for the student to see an old relationship inside a new form.
This is what unfamiliar questions often do
Students sometimes leave an examination saying:
“They gave us questions we have never seen before.”
Occasionally that is fair.
A paper can contain genuinely novel combinations.
But sometimes the student has seen every piece of Mathematics involved.
What she has not seen is that particular surface arrangement.
For example:
3^(2x) − 10(3^x) + 9 = 0.
A student sees powers of 3 and immediately searches memory for an indices method.
But notice:
3^(2x) = (3^x)².
So if u = 3^x, then:
u² − 10u + 9 = 0.
Factorise:
(u − 1)(u − 9) = 0.
Therefore u = 1 or u = 9.
So:
3^x = 1
or:
3^x = 9.
Hence x = 0 or x = 2.
Again, the student did not need a completely new method.
She needed to recognise a familiar quadratic living inside an exponential expression.
That is an important distinction.
More practice does not automatically produce this recognition
Suppose the student practises twenty ordinary quadratics.
She may become excellent at factorisation.
That is useful.
But if every question visibly announces itself as a quadratic in x, practice can strengthen execution without strengthening recognition.
Then 2sin²x − 5sin x + 2 = 0 feels like a new chapter.
Or e^(2x) − 5e^x + 6 = 0 feels like a new chapter.
Or 2(log x)² − 5log x + 2 = 0 feels like a new chapter.
Yet structurally they may all ask the student to notice:
aU² + bU + c = 0.
The repeated object U changes.
The quadratic relationship does not.
This is why transfer practice has to change more than the numbers.
The eye needs to learn what to ignore temporarily
This is one of the quieter skills in Mathematics.
A beginner often treats every visible feature as equally important.
Numbers.
Letters.
Units.
Context.
Diagram shape.
Position on the page.
A stronger student becomes selective.
She asks:
Which parts determine the structure?
Which parts determine only the particular instance?
Consider:
(7 − 3)/(8 − 6).
The answer is 2.
But if this appears in coordinate geometry as:
m = (y₂ − y₁)/(x₂ − x₁),
I do not mainly want the student to see 7, 3, 8 and 6.
I want her to see:
vertical change divided by horizontal change.
The numerical values matter for the final gradient.
The relationship matters for choosing the method.
This is why I sometimes delay substitution
Students often want to insert numbers immediately.
It feels productive.
Suppose a question gives two points A(3,7) and B(11,23).
Find the gradient.
A student may immediately write (23 − 7)/(11 − 3).
Fine.
But if the student regularly confuses coordinate formulas, I may first ask her to write:
m = Δy/Δx.
Then:
m = (y₂ − y₁)/(x₂ − x₁).
Only then substitute:
m = (23 − 7)/(11 − 3) = 2.
The extra line is not there because the arithmetic is difficult.
It preserves the relationship before the numbers occupy the page.
Later, once that relationship is secure, the student can compress again.
Symbols can sometimes carry less cognitive load than numbers
This surprises students.
They assume letters are abstract and numbers are concrete.
Therefore numbers should always be easier.
But compare carrying Δy/Δx with repeatedly remembering “23 minus 7 over 11 minus 3”.
The symbolic version tells us what each operation means.
Or compare:
new value = old value × 1.12
with performing a 12% increase through several disconnected arithmetic steps.
The symbolic form can hold the relationship still while the actual values change.
That is one of the reasons algebra is useful.
It is not merely arithmetic made harder with letters.
It is a way of separating structure from instance.
A word problem often becomes easier when the quantities are named before they are numbered
Suppose:
The length of a rectangle is 3 cm more than its width. Its perimeter is 34 cm. Find its dimensions.
Some students immediately begin experimenting with numbers.
What two numbers differ by 3?
Which pair gives perimeter 34?
They may eventually succeed.
But name the width w.
Then the length is w + 3.
Perimeter gives:
2w + 2(w + 3) = 34.
Now the problem has a structure.
Simplify:
4w + 6 = 34.
So 4w = 28 and w = 7.
Length: 10.
The variables did not make the problem less concrete.
They prevented the student from carrying several verbal relationships mentally at once.
The page is now doing some of the remembering.
This is also why diagrams can become too numerical
A geometry diagram sometimes arrives covered with labels.
37°.
68°.
12 cm.
17 cm.
A radius of 5.
A tangent.
A chord.
A student begins calculating whatever looks available.
I may ask:
“Before using any number, what relationships are present?”
Parallel lines?
A right angle?
A radius perpendicular to a tangent?
Similar triangles?
An isosceles triangle?
Angles in the same segment?
The answer to that question determines the route.
The numbers then populate the route.
Without this pause, geometry becomes a search for arithmetic opportunities.
With it, the diagram becomes a network of constraints.
Strong students can be vulnerable because they calculate quickly
A student with good arithmetic and calculator fluency is rewarded for moving.
She sees numbers and starts.
That usually works on routine questions.
So the habit becomes reinforced.
But on a multi-step A-Math problem, premature calculation can be expensive.
She may compute three quantities that are all correct and all irrelevant.
The problem is not mathematical weakness in the ordinary sense.
She can calculate.
She may know every formula.
She has simply committed to local arithmetic before identifying the global structure.
This is why I sometimes tell a capable student:
“Don’t calculate yet.”
Not because speed is bad.
Because I want speed to begin after orientation.
Method selection often happens before the first number is used
Consider:
A quadratic graph touches the x-axis.
Before seeing any coefficients, I already know something important.
“Touches the x-axis” suggests one repeated root.
Therefore:
b² − 4ac = 0.
The numbers will determine the parameter.
But the wording determines the mathematical condition.
Or:
Two lines are perpendicular.
Before substituting coordinates, I know the gradients must satisfy:
m₁m₂ = −1
for non-vertical lines in the usual Cartesian setting.
Again, numbers matter later.
The structural statement arrives first.
Students who rush directly to arithmetic often miss these cues.
Removing numbers is not the same as ignoring context
There is an important boundary.
I do not want students to abstract so aggressively that the question loses meaning.
Suppose a calculated answer is x = −4.
If x represents an unrestricted algebraic variable, fine.
If x represents the number of students in a class, impossible.
If it represents time after an event, perhaps impossible depending on how the variable was defined.
If it represents a coordinate, perfectly possible.
The context must return.
Likewise, units matter.
Domains matter.
Ranges matter.
Measurements matter.
The purpose of temporarily stripping away numerical detail is to expose structure—not to pretend that the details are irrelevant.
The full problem owns the final answer.
Some numbers are themselves structural
This is another boundary.
Not every number is superficial.
Consider:
x² + 6x + 9.
The 9 matters structurally because:
x² + 6x + 9 = (x + 3)².
Or consider b² − 4ac.
Its sign determines the root structure of a quadratic.
Or a trigonometric angle such as 90°.
That may encode perpendicularity.
So “remove the numbers” is not a literal universal rule.
It is a teaching move.
The deeper question is:
Which details are carrying the mathematical relationship, and which are merely specifying this particular case?
That distinction requires judgement.
A useful temporary variable can reduce visual complexity
One of the most powerful small moves in A-Math is to name a repeated complicated expression.
Suppose:
2(x² + 1)² − 7(x² + 1) + 3 = 0.
A student may begin expanding (x² + 1)².
That is possible.
It may also create unnecessary algebra.
Instead, notice that x² + 1 appears repeatedly.
Let:
u = x² + 1.
Then:
2u² − 7u + 3 = 0.
Factorise:
(2u − 1)(u − 3) = 0.
So u = 1/2 or u = 3.
Return:
x² + 1 = 1/2
or:
x² + 1 = 3.
The first gives x² = −1/2, so no real x.
The second gives x² = 2, hence:
x = ±√2.
The substitution has reduced visual noise and exposed a simpler problem inside the larger one.
This is not a trick.
It is mathematical compression.
Good compression preserves the important structure
Students sometimes learn substitution as another procedure:
“When you see this type, let u = …”
I would rather they understand the purpose.
We are asking:
Is there a complicated object behaving like one unit?
Does naming it allow us to see a familiar relationship?
If yes, substitution may help.
This principle travels far beyond one question type.
In trigonometry.
Exponentials.
Logarithms.
Polynomial expressions.
Later calculus.
The specific substitution changes.
The cognitive move remains:
compress a repeated object so the relationship between objects becomes visible.
That is a powerful form of mathematical thinking.
Parents can notice this without teaching A-Math
When your child is stuck on a dense question, one useful question is:
“If the actual numbers were different, what kind of problem would this still be?”
Or:
“Is there one expression that keeps appearing?”
Or:
“Before calculating, what relationship do you already know?”
These questions do not supply the method.
They encourage the student to look one level above the arithmetic.
If she can say:
“It is quadratic in sin x,”
or:
“These triangles are similar regardless of the side lengths,”
or:
“The two lines must have the same gradient because they are parallel,”
the problem has already become more organised.
This can diagnose whether a topic is genuinely understood
Suppose a student says she understands quadratics.
Give 2x² − 5x + 2 = 0.
She solves it.
Good.
Now 2t² − 5t + 2 = 0.
Still easy.
Now 2(sin x)² − 5sin x + 2 = 0.
Now 2(3^x)² − 5(3^x) + 2 = 0.
The computational skeleton remains similar.
The surface becomes progressively less familiar.
Where does recognition fail?
That point tells me something much more precise than simply saying:
“She is weak at A-Math.”
Perhaps factorisation is secure.
Perhaps the quadratic concept is secure.
Perhaps the missing skill is recognising when an entire expression can temporarily behave like one variable.
That diagnosis leads to a much smaller repair.
Repair should therefore alter the surface, not merely repeat the answer
If a student fails to recognise the hidden quadratic, I do not want ten copies of the original question.
I might use a short sequence:
2u² − 5u + 2 = 0.
Then 2y² − 5y + 2 = 0.
Then 2(sin x)² − 5sin x + 2 = 0.
Then 2(3^x)² − 5(3^x) + 2 = 0.
Then perhaps 2(log x)² − 5log x + 2 = 0.
After each one, ask:
“What is playing the role of the quadratic variable?”
The aim is not to memorise five question types.
It is to make one structure portable.
Transfer is visible when the student stops caring about the costume
A question has a costume.
Trigonometry costume.
Coordinate-geometry costume.
Exponential costume.
Word-problem costume.
But beneath the costume, a familiar relationship may remain.
A student with weak transfer is highly sensitive to the costume.
Change the letter.
Change the context.
Change the diagram.
Change the order of information.
The question feels new.
A student with stronger transfer begins recognising invariants.
“This is still a quadratic.”
“This is still constant rate.”
“This is still similar triangles.”
“This is still a perpendicular-gradient relationship.”
That is one of the signs I look for before saying a concept is truly secure.
Examinations deliberately reward this kind of seeing
The harder questions in Secondary Mathematics do not always use harder arithmetic.
Often they remove the chapter label.
They combine ideas.
They hide a familiar object inside another.
They give unnecessary information.
They present the known relationship backwards.
They change the notation.
This is why students sometimes practise large quantities and still feel surprised in an examination.
The issue is not necessarily practice volume.
The practice may have been too visually consistent.
Every exercise told the student what kind of exercise it was.
The examination does not always offer that kindness.
This does not mean worksheets should become puzzles all the time
There is another boundary.
Students need routine work.
Fluency matters.
If factorisation itself is slow and inaccurate, hiding the quadratic inside trigonometry may simply overload the student.
First the component skill has to become reliable.
Then the surface can vary.
Development usually needs both.
Routine practice stabilises execution.
Variation tests recognition and transfer.
The error is not using routine questions.
The error is assuming success on routine questions automatically proves flexible understanding.
There is a sequence I like
First, make the Mathematics explicit.
“This is a quadratic.”
Teach the method properly.
Then practise until execution is sufficiently stable.
Then remove the chapter label.
Change the letter.
Change the context.
Hide the same structure inside another expression.
Mix it with nearby topics.
Finally, ask the student to explain what remained the same.
This final step matters.
If she can say:
“The numbers changed, but the relationship between the repeated expression and its square was still quadratic,”
she has extracted something reusable.
I also like asking what information can be ignored for the first thirty seconds
Not discarded permanently.
Ignored temporarily.
A long word problem may contain dates, prices, names, units and contextual sentences.
I ask:
“What do we need in order to decide the method?”
Perhaps only:
there is a fixed starting cost, plus a constant amount per unit.
That tells us the model is linear.
The exact dollar amounts can wait.
Or a geometry question contains four side lengths.
But before computing anything, the parallel lines establish similarity.
That relationship decides the route.
Students need to learn that reading Mathematics does not mean treating every piece of information with equal priority at the same moment.
Attention itself has to be organised.
This is one reason small-group teaching can reveal more than completed homework
When I watch a student begin a question, I can see which information captures her first.
Does she immediately start entering numbers?
Does she search for keywords?
Does she scan the whole structure?
Does she identify repeated forms?
Does she ask what is invariant?
The final answer does not reveal all of this.
A correct solution may have come from strong recognition.
Or from familiar surface cues.
Or from trial and error.
Or from a remembered template.
The first thirty seconds often tell me more about how transferable the learning really is.
The useful next route
If a student performs well on familiar exercises but struggles as soon as questions look different, I would not automatically increase the volume of practice.
I would first test whether structure recognition is the bottleneck.
Take one concept she already knows.
Keep the underlying Mathematics stable.
Change the numbers.
Then the letters.
Then the context.
Then place the same relationship inside another topic.
Ask her, before calculating:
“What has stayed the same?”
If she can identify the invariant but executes badly, repair execution.
If she executes familiar versions perfectly but cannot identify the invariant under a changed surface, work on recognition and representation.
If she recognises the structure and selects the method but fails only when several steps must be coordinated, then the problem lies elsewhere.
The point is to make the next piece of teaching smaller and more accurate.
What long teaching has made me notice
Students often think difficult Mathematics contains difficult numbers.
Sometimes it does.
But many of the questions that unsettle adolescents are difficult for another reason.
The familiar Mathematics has been placed inside unfamiliar clothing.
A quadratic becomes a trigonometric equation.
A rate becomes a gradient.
A scale relationship becomes a geometry problem.
A repeated object becomes algebraically complicated enough that the student stops treating it as one object.
The page becomes full.
And once the page becomes full, attention can become local.
Calculate this.
Simplify that.
Use this number.
Press this button.
The student may be working hard while moving further away from the organising idea.
That is why I sometimes remove detail before adding instruction.
Not because the detail does not matter.
Because I want the student to recover the hierarchy.
First:
What mathematical relationship is governing this problem?
Then:
Which quantities belong inside it?
Then:
What calculation follows?
Eventually, I want the student to perform that separation herself.
She reads a dense question and quietly sees through it.
Not instantly.
Not magically.
But with increasing confidence.
The numbers still matter.
The context still matters.
The units still matter.
Yet none of them prevents her from noticing the mathematical skeleton holding everything together.
And I think that is an important threshold in Secondary Mathematics.
A student begins by learning how to calculate with the information she is given.
Later, she learns something more powerful:
how to see which part of the information is actually carrying the Mathematics.