Worked solutions, then your turn — free
The practice portal pairs questions with step-by-step worked reasoning, so students can study a method then immediately try a fresh one — the example–problem pair, built in.
Open the practice portal →Common sense says you learn maths by doing problems, so the more a child struggles through them alone, the better. Cognitive science says: for a beginner, that's often the slowest, most demoralising route. Studying a fully worked solution first can teach more, faster — until the child improves, at which point the very same help starts to hold them back. Here's how to get the timing right.
Two tutors, same topic, same student. The first hands over a fresh problem and says, "Have a go — see what you can work out." Twenty minutes later the page is a graveyard of crossed-out attempts and the student is convinced they're hopeless. The second tutor puts a fully solved example on the table and says, "Let's read this together — tell me why each line follows from the one above." Ten minutes later the same student is confidently solving a fresh one. Nothing about the student changed. What changed was whether the method was made visible or left to be excavated under pressure.
Most of us instinctively admire the first approach — it looks like "real" learning, the heroic struggle. But decades of research on cognitive load suggest that, for genuine beginners, the second approach usually wins. And then, in a twist that catches many teachers out, the advantage reverses as the student gets better. Understanding both halves of that story is one of the most practical things a parent or teacher can do.
The choice shows up every evening at kitchen tables and every lesson in classrooms: when a child meets something new, do you show them how, or let them discover? The "let them discover" instinct is well-meant — we want children to think, not copy. But there's a difference between thinking about a method you can see and flailing for a method you've never met. The first builds understanding; the second often just builds frustration. The worked example is simply a way of letting a beginner think hard about the right things.
To see why, recall how the mind handles new information. Your working memory — the small mental workspace where thinking happens — holds only a few items at once. A beginner facing an unfamiliar problem has to juggle the goal, the givens, and a frantic search through half-remembered methods, all at the same time. That search devours the workspace, leaving almost nothing free to actually encode the method even on the rare occasion they stumble onto it.
Imagine being dropped in an unfamiliar city with no map and told to find a specific café by trial and error. You might eventually get there, exhausted, having learned almost nothing about the city — all your energy went into not getting lost. Now imagine being walked there once, with someone pointing out the turns. You arrive calm, and you could do it again tomorrow. A worked example is that guided walk. Unguided struggle is the exhausting wander — and for a beginner, the wander teaches surprisingly little.
So the paradox is that asking a beginner to "just try" a brand-new type of problem can be the least efficient way to teach it. They spend their limited mental resources on searching, not learning, and they come away with the lesson "I can't do this" rather than the method itself.
John Sweller's cognitive load research (1988 onward) repeatedly found that novices who study worked examples learn new procedures faster, with less effort and better transfer, than novices who solve the equivalent problems unaided. By removing the wasteful search for a method, worked examples free up the workspace to absorb the method itself. It's one of the most replicated findings in instructional psychology.
The most effective approach isn't worked examples forever, nor blank problems from day one — it's a gradual handover. "Completion problems" and "faded examples" show the first steps and ask the learner to finish, then progressively blank out more steps until they're solving independently. The scaffolding is removed one rung at a time, so the workspace is never overwhelmed and never under-challenged.
Here's where it gets genuinely surprising. Kalyuga, Ayres, Chandler and Sweller (2003) documented what they called the expertise-reversal effect: instructional support that helps a beginner can begin to hinder a more advanced learner. Once a student knows a method, being walked through a worked example becomes redundant — and processing all that now-unnecessary explanation actually gets in the way. For the more expert student, solving the problem themselves is the better workout.
It means there is no single "best" way to practise — it depends on the learner's current expertise. The same worked example that rescues a beginner can bore and slow an expert; the same blank problem that overwhelms a beginner is exactly what an expert needs. Good teaching isn't picking a side in "show vs struggle" — it's reading where the learner is and adjusting the support accordingly. Constant heavy scaffolding and constant deep-end struggle are both wrong for somebody.
None of this means struggle is bad — and here the research has an honest tension worth naming. Manu Kapur's work on "productive failure" shows that letting students grapple with a problem before being taught the method can deepen conceptual understanding. So which is it — show the method, or let them struggle first?
The reconciliation is about what you're trying to build. If the goal is to acquire a new procedure efficiently — learn how to solve quadratics, say — worked examples win for the novice. If the goal is to build deep conceptual understanding and problem-solving flexibility, a well-designed period of struggle on a carefully chosen problem, followed by good instruction, can pay off. The error is using the wrong tool: making beginners flounder through routine procedure (overload, no learning), or spoon-feeding every step so students never learn to think (no struggle, no flexibility). Productive struggle is designed and supported; the failure mode is struggle that's just abandonment.
High-performing Chinese classrooms are famous for carefully sequenced worked examples and "teaching with variation" — presenting a method, then systematically varying the problems around it to build robust understanding before increasing difficulty. It's close to a textbook application of the worked-example effect, paired with deliberate practice.
Australia is a centre of the "explicit instruction" movement, drawing directly on cognitive load theory (Sweller worked there). National bodies argue for fully-guided teaching of new material — worked examples, clear modelling — before independent practice, precisely to avoid overloading novices.
Japanese lessons often open with students grappling with a single rich problem before the method is formalised — the "productive struggle" tradition. It looks like the opposite of worked examples, but note the difference: the struggle is carefully chosen, time-limited and aimed at concepts, then resolved with instruction. Different goal, different tool.
England's "maths mastery" reforms lean on "I do, we do, you do" — the teacher models (worked example), the class practises together, then students work alone. That sequence is essentially the fade-the-support principle turned into a classroom routine.
The US "math wars" pitted discovery-oriented approaches against explicit teaching. The cognitive-load evidence sides with guidance for novices, but the expertise-reversal effect adds the crucial caveat both camps can accept: the right amount of guidance changes as students learn.
Lead new content with worked examples, then fade. Model the full method, move to completion problems (some steps given), then to independent solving. Build example–problem pairs into practice sets rather than pages of blank questions.
Watch for the reversal. Once students are fluent with a method, heavy worked-example support starts to bore and slow them — switch them to solving and problem-variation. The same lesson design can't serve a class spread across very different levels of expertise, which is a strong argument for adaptive or grouped practice.
Design struggle deliberately when you want concepts. If your aim is deep understanding rather than procedural fluency, a short, well-scaffolded period of grappling before instruction can be powerful — but choose the problem carefully and resolve it well. Struggle is a tool, not a default.
The practice portal pairs questions with step-by-step worked reasoning, so students can study a method then immediately try a fresh one — the example–problem pair, built in.
Open the practice portal →"Showing a child the method is spoon-feeding and stops them thinking."
For a beginner, a worked example focuses thinking on understanding the method, rather than wasting it searching blindly. It's a springboard to doing, not a substitute for it.
"More struggle always means more learning."
Productive struggle helps; unproductive flailing overloads and demoralises. Struggle must be designed and supported to pay off.
"Once worked examples help, keep using them."
The expertise-reversal effect means support that helps novices hinders experts. Fade it as the learner improves.
A problem shown with its complete step-by-step solution, for the learner to study rather than solve. It makes the method visible without forcing a beginner to discover it — usually a faster, less overwhelming route into a new topic.
For something new, study a worked example first; unguided struggle tends to overload a beginner. Once the method is secure, shift toward solving independently. The art is matching the support to the stage rather than always doing one or the other.
Instructional support that helps beginners can start to hinder more advanced learners — so good teaching fades the scaffolding as expertise grows, rather than keeping it constant.
Productive struggle can deepen understanding, especially for concepts and problem-solving. Unproductive struggle — no idea where to begin — mostly overloads and demoralises. Choose worked examples for acquiring procedures and reserve designed struggle for the right moments.
Pair them: study a fully worked solution together, talking through the "why" of each step, then immediately try a near-identical problem with the example covered. Gradually blank out more steps until your child is solving from scratch.
One-to-one means the support fades at exactly the right pace — modelled when it's new, handed over as they're ready. We find that level in the free assessment.
Book the free assessment