Desirable Difficulties: When Effort Helps Learning—and When It Does Not

Here, a desirable difficulty means a learning challenge designed to make a child do useful thinking now so that knowledge is more available later. The difficulty is not valuable by itself. It is productive only when the child has enough prior knowledge to make a meaningful attempt, the effort targets the idea being learnt, and feedback or teaching closes the gap afterwards.

That distinction matters at home. Recalling a method before reopening the book can be useful effort. Deciphering a vague question, working without necessary instruction or repeating errors without feedback is merely obstruction. “Harder” is not automatically “better”, and prolonged struggle is not proof that learning is happening.

Contents

What makes a difficulty desirable?

Why effort can help — and why effort can mislead

The calibration test

English example: explaining an inference

Maths example: choosing a fraction method

A parent action sequence

Common mistakes

Adjustment cues

Frequently asked questions

Next step

What makes a difficulty desirable?

Use desirable difficulty as a description of the relationship between a task and a learner, not as a permanent label attached to a worksheet. The same prompt can provoke useful thought for one child, effortless repetition for another and unproductive confusion for a third.

A productive challenge has four features:

The learning target is clear. The child knows whether they are recalling a fact, selecting a method, explaining evidence or applying a taught rule.

The necessary foundation exists. The material has been taught, or enough information is present to reason towards an answer.

The effort is relevant. Thinking goes into retrieving, discriminating, explaining or applying—not decoding awkward wording or guessing what the adult wants.

There is a route out. The child receives a cue, model, explanation or corrective feedback before confusion becomes rehearsal of an error.

This makes calibration part of teaching. Remove every challenge and practice can become a copy-and-complete exercise. Add challenge indiscriminately and the child may spend attention on everything except the intended idea.

Why effort can help — and why effort can mislead

Roediger and Butler's review treats retrieval as a learning event, not merely a way to measure what has already been learnt. It reports that bringing information to mind can support later retention more than repeated study, that feedback strengthens the benefits of testing, and that retrieval can support flexible use in different contexts.[1]

The review also makes the boundary useful for this article: retrieval conditions matter. Very easy immediate retrieval can do little for long-term retention, while effortful retrieval, repeated attempts with intervals and feedback can be more useful.[1] This does not mean that any difficult activity improves learning. The paper is a review of retrieval-practice research, not a trial of the calibration routine below, a study of individual 11+ pupils or evidence that frustration itself is beneficial.

Two misleading signals often confuse adults:

Fluency can look like learning. A child may complete an almost identical question while the worked answer is visible because the route is still in view.

Effort can look like learning. A child may work intensely because the wording is ambiguous, the prerequisite is missing or too many decisions have been introduced at once.

So do not judge the task by smoothness or sweat. Ask what thinking the child actually performed and what they can now explain or use.

The calibration test

Before preserving a difficulty, ask five questions.

1. What is the child meant to learn?

Name the target precisely: “justify an inference with two details” is more useful than “do comprehension”; “choose a common denominator” is more useful than “do fractions”.

2. Has the required knowledge been taught?

Retrieval is not a substitute for first instruction. If the child has never learnt what a denominator represents, concealing the explanation creates a knowledge gap, not useful retrieval.

3. Can the child make a meaningful start?

A productive attempt should contain some relevant move: identifying a clue, drawing a representation, stating a rule or choosing a plausible operation. Random guessing, silence and unrelated working suggest that the entry point is missing.

4. Is the difficulty in the target?

If the aim is inference, unfamiliar vocabulary that blocks the whole passage is incidental difficulty. If the aim is vocabulary in context, examining that unfamiliar word may be the target. Difficulty belongs in the thinking you want to strengthen.

5. What will happen after the attempt?

Plan feedback before setting the task. A difficult prompt without checking can strengthen a misconception or leave the child with no usable conclusion.

English example: explaining an inference

Suppose a child has been taught to combine a character's actions with details from the setting.

> Leila pressed the library book flat beneath her jumper. At the classroom door, she slowed and checked whether anyone was watching.

Ask: What can you infer about Leila, and which details support your answer?

A useful answer is that Leila probably does not want others to see the book. The reasoning comes from two connected details: she hides it beneath her jumper and checks whether anyone is watching. “Leila likes reading” is possible in life, but the passage does not establish it.

The desirable difficulty is producing and justifying an inference without choosing from supplied answers. It becomes poorly designed if the adult demands one secret phrase, withholds the meaning of infer when that term is new, or adds obscure vocabulary unrelated to the target.

Least adjustment: point to the two sentences and ask, “Which actions seem connected?” This narrows attention but preserves the inference.

Next adjustment: offer two interpretations—“She wants to protect the book from rain” or “She does not want it noticed”—and ask which the evidence supports. If the child still cannot explain, model the evidence-to-inference link, then use a fresh short passage. Do not keep repeating the same unanswered question.

Maths example: choosing a fraction method

Suppose the child has already learnt that fractions need equal-sized parts and has compared fractions with common denominators.

Ask: Which is greater, 3/4 or 5/8? Explain without using a calculator.

One route is to express both fractions in eighths:

3/4 = 6/8

6/8 is greater than 5/8

therefore 3/4 is greater than 5/8

The productive challenge is selecting a representation and carrying out a known equivalence. Giving the rule “double the numerator and denominator of 3/4” before the attempt would remove that decision. Asking for comparison before equivalence or fraction size has been taught would remove the foundation instead.

Least adjustment: ask, “Could both fractions be written with the same denominator?” The child still chooses and completes the conversion.

Next adjustment: draw two equal bars divided into eighths and ask the child to shade 5/8, then partition and shade 3/4. If that support reveals that equal parts are not understood, stop comparing and reteach the meaning of numerator and denominator. More comparison questions will not repair the missing concept.

A parent action sequence

1. Choose one useful challenge

Remove the worked answer, vary a familiar example, ask for an explanation or mix two already-taught methods. Change one source of difficulty at a time so you can see what caused the response.

2. State the target and lower the stakes

Try: “This is practice at choosing a method. Have a go, and we will use the attempt to decide what help comes next.” Avoid presenting productive difficulty as a test of intelligence or resilience.

3. Wait for evidence, not for distress

Give enough quiet time for the child to retrieve or reason. Watch for a relevant start. A pause can be thinking; repeated guessing or visible escalation calls for intervention.

4. Give the least help that restarts useful thought

Move through a short ladder:

restate the goal;

direct attention to a relevant clue;

ask a choice question;

show one analogous step;

model and explain the method.

“Least help” does not mean withholding necessary teaching. It means preserving thinking when a small cue is enough and supplying instruction when it is not.

5. Close the loop

Check the answer, explain the correction and ask for one fresh application. Success immediately after copying proves little; the fresh item shows whether the explanation has become usable.

6. Set the next challenge from the evidence

If the child succeeded unaided, add modest variation or return after a gap. If one cue was enough, keep the same target and try a parallel example. If the method was absent, reteach before another independent attempt.

For more detail on recall followed by checking and feedback, use the retrieval-practice guide.

Common mistakes

Treating confusion as productive

Confusion can signal a missing definition, unclear instruction or overloaded task. Diagnose it; do not celebrate it automatically.

Adding several difficulties together

New content, unfamiliar layout, a timer and tricky wording create an uninterpretable result. You cannot tell which element blocked the child.

Helping too soon

Completing the first step for the child every time can remove the exact decision they need to practise. Begin with a cue rather than a takeover when the foundation is secure.

Helping too late

A long deadlock does not become educational through duration. When attempts stop being relevant, change the support.

Using correctness as the only signal

A guessed correct option can hide weak understanding; a small arithmetic slip can hide sound reasoning. Ask for the clue, rule or method.

Turning effort into praise for suffering

Praise the useful action—reconsidering evidence, retrieving a rule, correcting a method—not discomfort itself. Children should not learn that asking for necessary instruction is a failure.

Adjustment cues

Keep the challenge when the child:

makes a relevant start;

can explain the target even if execution is imperfect;

uses feedback to correct the attempt;

succeeds on a fresh example with similar structure.

Reduce or restructure it when the child:

guesses without reference to the text or quantities;

cannot name or recognise the required idea;

repeats the same error after feedback;

is blocked by wording, layout or an irrelevant demand;

can only proceed while copying the model.

Increase it slightly when the child:

answers rapidly and explains accurately;

recognises the method from superficial cues alone;

succeeds only on questions presented in one familiar format.

A sensible next increase is one change: remove a prompt, vary the context, ask for justification or revisit later. It is not “make everything harder”.

Frequently asked questions

Should learning always feel difficult?

No. Explanation, modelling, successful practice and consolidation can feel clear. Introduce effort where it serves a known learning target; do not manufacture friction throughout the session.

How long should I let my child struggle?

Use the quality of the attempt rather than a fixed number of minutes. Continue while relevant thinking is developing. Intervene when the child is guessing, repeating an error or lacks an entry point.

Does giving a hint spoil the learning?

Not necessarily. A narrow hint can restart retrieval or reasoning without supplying the answer. If the child lacks the underlying knowledge, a fuller explanation is better than a trail of cryptic hints.

Is a harder worksheet a desirable difficulty?

Only if the added challenge targets taught knowledge and remains interpretable. Extra reading load, ambiguous wording or untaught content can make a worksheet harder without making it more useful.

Can this replace clear teaching?

No. Productive challenge depends on knowledge to work with. Teach and model new or insecure ideas, then calibrate what the child does independently.

Next step

Choose one already-taught English or Maths idea. Ask one question that requires recall, selection or explanation, then prepare a least-help cue and a next-step explanation before the child begins. See the science.

Sources

[1] https://doi.org/10.1016/j.tics.2010.09.003