You are currently viewing Productive failure learning: why struggling first can help

Productive failure learning: why struggling first can help

Originally published on mistakeology.com.

By Dr Tom Hitchcock, founder of Mr Purple (LinkedIn)

Productive failure learning sounds like an educational contradiction. Why let someone struggle when you could simply show them the answer? Because the first attempt, even when it is gloriously wrong, can prepare the mind to notice what the later explanation actually means.

This is not a licence to abandon learners beside a difficult problem and call the confusion character building. It is a carefully designed sequence: attempt first, compare ideas, then receive clear instruction. The failure is useful because someone helps turn it into understanding.

What productive failure learning actually means

Productive failure is associated with the work of learning scientist Manu Kapur. The basic move is simple. Learners tackle a new and suitably challenging problem before being taught the formal method. They generate possible solutions, test assumptions and often discover that their first ideas do not quite work.

Instruction comes afterwards. The teacher or facilitator draws out the important features of the problem, compares the learners’ approaches and explains the established concepts. Earlier errors now have a job. They provide contrast, reveal gaps and give the explanation somewhere to land.

The order matters. In a conventional lesson, an expert explains the method and learners practise it. In productive failure, learners first experience the shape of the problem. Then the expert helps them make sense of that experience.

There is a useful Mistakeology distinction here. A mistake is not automatically productive. It becomes productive when it is examined, connected and used. Without that conversion, it is merely an awkward afternoon.

What the evidence says, without the confetti cannon

A major review by Tanmay Sinha and Manu Kapur examined 53 studies containing 166 comparisons. It compared problem solving before instruction with instruction before problem solving. The review found a moderate overall advantage for the problem first sequence, especially for conceptual understanding and transfer. You can read the published research review.

That finding is promising, but it needs honest handling. Productive failure is not proof that all struggle helps every learner. The review identified conditions that influenced the results. The nature of the task, the age of the learners, the knowledge being assessed and the quality of the lesson design all mattered.

The strongest case is not that direct instruction is bad. It is that direct instruction can become more meaningful after learners have tried to make sense of a problem. The attempt prepares questions. The instruction then answers questions the learner has genuinely encountered.

This also explains why copying the visible surface of the method can fail. Giving people an impossible task, offering no useful comparison and rushing the explanation is not productive failure. It is failure with a fashionable name badge.

Four ingredients that make struggle useful

Leaders, teachers and facilitators can use four practical checks.

  • A problem with several plausible approaches. The task should invite thinking, not merely test recall. People need enough prior knowledge to begin, even if they cannot yet finish elegantly.
  • Permission to produce imperfect ideas. If every wrong answer attracts embarrassment, learners protect themselves by staying quiet. Psychological room to explore is part of the design.
  • Visible comparison. Different attempts should be examined together. What did each approach notice? Where did it bend? What useful fragment did it contain?
  • Clear instruction after the attempt. The expert still teaches. Productive failure does not remove guidance. It changes when guidance arrives and what it can connect to.

This sequence resembles a good investigation. First gather clues. Then organise them. Finally bring in the expert model that explains the whole scene.

A practical version for teams and classrooms

You do not need to redesign an entire programme. Try one contained learning cycle.

  1. Choose a concept that people often understand superficially.
  2. Set a realistic problem that exposes the concept without naming the method.
  3. Give people time to attempt it alone or in small groups.
  4. Collect two or three different approaches, including incomplete ones.
  5. Ask what each attempt assumed, revealed and missed.
  6. Teach the recognised principle or method clearly.
  7. Use a fresh problem to see whether the learning transfers.

For a business team, the problem might be a service delay, a handover breakdown or a customer decision with competing priorities. For learners, it might be a mathematical relationship or a scientific explanation. The subject changes, but the rhythm remains: attempt, compare, explain, apply.

Keep the first attempt low stakes. The aim is to surface thinking, not to rank people. If a score, reputation or job feels immediately threatened, curiosity tends to pack its little suitcase and leave.

Where the idea has limits

Productive failure is not suitable for every task. When safety is involved, when a procedure must be performed correctly from the outset, or when learners have almost no relevant prior knowledge, explicit teaching may need to come first. Nobody needs an adventurous experiment with an emergency process.

It also asks more of the facilitator than simply withholding the answer. The problem must be chosen carefully. Learners need enough support to stay engaged. The later explanation must connect directly to what happened during the attempt.

Watch for cognitive overload, frustration and exclusion. A learner who cannot access the task is not receiving a useful challenge. They are being locked outside it. Adjust the problem, provide starting information or reduce unnecessary complexity while preserving the central puzzle.

For a wider set of research prompts, explore the Mistakeology research and thesis questions. The interesting question is rarely whether failure is good or bad. It is what turns a particular failure into useful data.

Make the wobble visible

Productive failure gives us a disciplined alternative to two unhelpful extremes. We do not need to prevent every mistake, and we do not need to romanticise struggle. We can design a safe attempt, study what it reveals and then teach with greater precision.

Try it on one modest problem. Notice where thinking wobbles before correcting it. Then use the Wobble Map to capture what happened, what it exposed and what you will change next.

Common questions

Is productive failure the same as learning by trial and error?

Not quite. Productive failure uses an intentionally designed problem, followed by comparison and clear instruction. Trial and error can happen without that structured sense making.

Does productive failure mean teachers should explain less?

No. The explanation remains essential, but it comes after learners have attempted the problem and developed a reason to notice the key ideas.

Can productive failure be used in workplace learning?

Yes, particularly with realistic problems that allow several plausible approaches. Keep the attempt low stakes, examine the reasoning, then connect it to a clear method or principle.

When should productive failure not be used?

Avoid it where an incorrect first attempt creates a safety risk, or where learners lack enough prior knowledge to begin. In those cases, teach the essential foundation first.

Leave a Reply