Desirable Difficulty: Why Making Learning Harder Helps Kids Remember More

It feels wrong to watch a child struggle. When your 9-year-old can’t remember how to do long division three days after you explained it perfectly, the instinct is to explain it again, more clearly, with more examples, in a calmer environment, until it finally clicks and they can do it smoothly. That instinct is almost always counterproductive.

Robert Bjork, a cognitive psychologist at UCLA and one of the most influential learning scientists of the last 50 years, coined the term “desirable difficulties” to describe a set of conditions that appear to make learning harder in the short term but produce dramatically better retention over time. The key insight is that the mechanisms that signal fluency — ease of processing, smooth retrieval, confident performance during practice — are unreliable predictors of long-term retention. The mechanisms that actually produce durable memory are often uncomfortable and require visible effort.

Key Takeaways

• Bjork’s desirable difficulties include spacing (distributing practice over time), interleaving (mixing topics), retrieval practice (recalling rather than re-reading), and the generation effect (producing answers before seeing them).
• Making learning feel easier in the moment usually makes it stick less — the correlation between perceived difficulty and long-term retention is often negative.
• Interleaving different problem types in practice is harder and less fluid than blocked practice, but produces better performance on novel problems 1–2 weeks later.
• The generation effect is one of the most reliable findings in memory research: generating an answer — even an incorrect one — produces better retention than reading the correct answer.
• Calibrating difficulty for children requires age-sensitivity: the level of desirable difficulty that helps an 11-year-old may be genuinely overwhelming for a 6-year-old.

What Makes a Difficulty “Desirable”

Not all difficulties improve learning. Bjork and Bjork (2011) draw an explicit distinction: difficulties are desirable when they:

1. Trigger deeper cognitive processing during learning
2. Force the learner to reconstruct what they know (retrieval) rather than recognize what’s in front of them
3. Create variability in practice conditions that forces generalization rather than pattern-matching
4. Produce interference that requires the learner to actively suppress competing information

Difficulties are undesirable when they stem from poor instruction, inadequate prerequisite knowledge, or stress levels that suppress hippocampal encoding. The diagnosis matters: making material confusing because you haven’t explained it clearly is not the same as making retrieval effortful because the child has to reconstruct what they understood.

The Four Main Desirable Difficulties

1. Spacing (Distributed Practice)

The spacing effect is arguably the most replicated finding in memory research: distributing practice over time produces dramatically better long-term retention than concentrating the same amount of practice in a single session.

Cepeda and colleagues (2006) analyzed 317 experiments on spacing and found an effect size that dwarfs most educational interventions. The spacing advantage is not about studying more — it’s about when you study. An hour of homework spread across three 20-minute sessions on different days consistently outperforms a single 60-minute block, even when total study time is identical.

For parents, this reframes evening homework. A child who reviews Tuesday’s spelling for 5 minutes on Wednesday and again on Friday is practicing spacing — and will retain dramatically more than a child who does one thorough Sunday review session. For a full framework on implementing spacing at home, see our article on spaced repetition and the spacing effect for children.

2. Interleaving (Mixed Practice)

Blocked practice means drilling one skill type completely before moving to the next (all addition problems, then all subtraction problems). Interleaved practice mixes skill types in random or varied order (addition, subtraction, multiplication, in rotation).

Research by Rohrer, Dedrick, and Stershic (2015), published in Journal of Educational Psychology, randomly assigned middle school students to either blocked or interleaved math practice. During practice, blocked students performed significantly better. On a test one month later, the interleaved group outperformed blocked practice by a striking margin.

The mechanism: interleaving forces students to identify which procedure applies to each problem, not just execute a procedure they’ve just been primed with. This extra discrimination step is effortful but produces the kind of flexible knowledge that transfers to novel problems — the goal of real learning.

3. Retrieval Practice (The Testing Effect)

Having students practice retrieving information from memory — rather than re-reading, reviewing notes, or re-watching explanations — is one of the most robust desirable difficulties. Roediger and Karpicke (2006), in Psychological Science, showed that a single retrieval practice session produced better week-later retention than a full additional study session, even when the subjects thought the study session had been more productive.

The mechanism isn’t simply “practice makes perfect” on recall. Each retrieval attempt modifies the memory trace — it updates and strengthens the connections required to retrieve that information again. This is why children who can recall information feel more certain they know it, while children who have only re-read it often suffer retrieval failures on tests despite feeling confident during studying.

For a deeper exploration of retrieval practice with practical strategies by age group, see our article on retrieval practice and the testing effect for kids.

4. The Generation Effect (Producing Beats Reading)

Slamecka and Graf (1978) published the original generation effect finding: subjects who generated words to fill in blanks (e.g., “Cold and hot: C___ and ___”) retained them significantly better than subjects who read word pairs fully (e.g., “Cold and hot”). This effect has been replicated across decades and across a wide range of content types.

The mechanism involves the depth-of-processing principle: generating a response requires constructing an answer from your own knowledge networks, which activates more memory associations and produces a more distinctive, retrievable trace. Reading a provided answer requires only recognition.

For children, the practical implication is significant: having a child try to answer a question before looking at the answer is more valuable than presenting the answer and then confirming understanding. Getting the answer wrong and then seeing the correct answer still produces better retention than just reading the correct answer from the start.

Comparison: Which Study Strategies Actually Work

Study strategy	Short-term performance	Long-term retention	Evidence quality
Re-reading notes	High (familiar)	Low	Strong (consistently ineffective)
Highlighting / underlining	Moderate	Low	Strong (consistently ineffective)
Summarizing	Moderate	Low–moderate	Moderate
Elaborative interrogation (“why?”)	Moderate	Moderate–high	Moderate
Practice testing / retrieval	Low–moderate	High	Very strong
Interleaved practice	Low (feels harder)	High	Strong
Spaced practice	Low–moderate	Very high	Very strong (most replicated)
Generation (fill-in, produce first)	Low (struggle)	High	Strong

This table summarizes the findings from Dunlosky and colleagues (2013), a landmark review published in Psychological Science in the Public Interest that evaluated 10 common study strategies across hundreds of studies. The counterintuitive finding: the strategies students perceive as most effective (re-reading, highlighting) have among the weakest evidence bases, while strategies that feel harder and less fluid perform best over time.

Calibrating Difficulty by Age — What “Desirable” Looks Like at 6 vs. 12

This is where the framework requires parent judgment. Desirable difficulties are calibrated to a learner’s current knowledge state. A difficulty that’s desirable at one level of expertise becomes undesirable beyond it.

Ages 5–7: Spacing and generation effects are strong at this age, but the timescale must be compressed. A spacing interval of 1–2 days is more appropriate than a week-long gap because younger children’s forgetting curves are steeper. Interleaving should be modest — alternating between two skill types, not five. Generation works best with partial cues (fill-in-the-blank with a first letter hint) rather than open-ended recall.

Ages 8–11: Standard spacing intervals (2–5 days) become productive. Interleaving of 3–4 related topic types works. Retrieval practice can be more open-ended. The generation effect is particularly strong at this age if children have enough foundation knowledge to generate reasonable attempts.

Ages 12–17: All four desirable difficulties apply at adult-comparable timescales. Interleaving across subjects (not just within a subject) begins to produce benefits. The challenge at this age is motivational, not developmental — teenagers often resist strategies that feel harder despite the evidence for their effectiveness.

Applying Desirable Difficulties at Home — Concrete Tactics

Replace “review the chapter” with “close the book and tell me”

After any reading or learning session, ask your child to close their materials and summarize or explain what they just learned. This converts passive re-exposure into active retrieval. It’s harder and initially produces more errors — that’s what makes it work.

Make homework slightly interleaved

If your child has 20 math problems of the same type, suggest they do 5, then switch to a different task (reading, spelling), then come back for 5 more. This creates the interleaving benefit even within a single subject. The interruption feels disruptive, but the evidence says it helps.

Use “guess first” before looking up answers

When your child encounters an unfamiliar word, fact, or concept, ask them to guess before they look it up. The guess doesn’t need to be correct — the act of generating a guess before seeing the answer consistently improves retention of the correct answer, even compared to the condition where the correct answer is just presented.

Space out the subjects over the week

Rather than doing all math homework on Monday and all history on Tuesday, distribute topics across the week. Each return to a subject is itself a desirable retrieval event.

What to Watch For Over the Next 3 Months

Month 1: Try one desirable difficulty consistently — choose spacing. Set a 3-day review calendar for whatever your child is currently learning in school. After 4 weeks, compare how much of that material they can recall versus a comparably-studied topic that didn’t get spaced review.

Month 2: Introduce the “guess first” habit before homework lookups. Track whether your child is resisting it (normal) and whether, over time, their initial guesses become more accurate. Improving first-guess accuracy is a sign that the generation effect is building stronger memory networks.

Month 3: Assess long-term retention on material from month 1. Can your child recall and explain month-1 content without reviewing it? If yes, you’ve produced the first measurable evidence that desirable difficulties are working in your specific household. If not, examine whether the spacing intervals were too long or whether the initial encoding was too shallow.

Red flag: If introducing more effortful study strategies produces significant emotional distress (tears, avoidance, anxiety), back off and examine whether the difficulty level is genuinely desirable or whether there’s an underlying skill gap that’s making the difficulty unresolvable. Desirable difficulty requires having the prerequisite knowledge to generate reasonable attempts.

Frequently Asked Questions

Why do kids (and parents) keep choosing ineffective study strategies if better ones exist?

Because short-term performance is a compelling but misleading signal. After highlighting a chapter, a student feels prepared — recognition during re-reading is high. After a retrieval practice session, performance is lower and the experience is more uncomfortable. The feedback loop misleads learners into choosing familiar-feeling strategies over effective ones. Knowing the research helps, but changing the habit requires deliberate effort.

Is the generation effect real for young children who might generate the wrong answer and reinforce it?

Generating wrong answers is fine — the effect holds even when the initial generation is incorrect, as long as the correct answer is provided immediately after. What appears to happen is that the error creates a distinctive memory event, and the correction is then encoded against that distinctive event. The key is that feedback must follow promptly.

How do I convince my child to study this way when their school still uses re-reading and highlighting?

Frame it as a test — not of your child, but of the strategy. “Let’s try the harder way for one month and see if your quiz scores go up.” Children who see their own results are often more convinced than children who are told research supports something. Track scores and let the data do the persuading.

Does interleaving work for subjects outside math?

Yes. Interleaving has been studied in foreign language vocabulary, music practice (mixing scales, pieces, and technique drills), and athletic skill acquisition. The general principle — that variable practice requiring discrimination among types produces better transfer than blocked practice — holds across domains.

---

About the author Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Bjork, E. L., & Bjork, R. A. (2011). “Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning.” In M. A. Gernsbacher et al. (Eds.), Psychology and the Real World (pp. 56–64). Worth Publishers.
2. Roediger, H. L., & Karpicke, J. D. (2006). “Test-enhanced learning: Taking memory tests improves long-term retention.” Psychological Science, 17(3), 249–255. https://doi.org/10.1111/j.1467-9280.2006.01693.x
3. Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). “Distributed practice in verbal recall tasks.” Psychological Bulletin, 132(3), 354–380. https://doi.org/10.1037/0033-2909.132.3.354
4. Rohrer, D., Dedrick, R. F., & Stershic, S. (2015). “Interleaved practice improves mathematics learning.” Journal of Educational Psychology, 107(3), 900–908. https://doi.org/10.1037/edu0000001
5. Slamecka, N. J., & Graf, P. (1978). “The generation effect: Delineation of a phenomenon.” Journal of Experimental Psychology: Human Learning and Memory, 4(6), 592–604. https://doi.org/10.1037/0278-7393.4.6.592
6. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). “Improving students’ learning with effective study techniques.” Psychological Science in the Public Interest, 14(1), 4–58. https://doi.org/10.1177/1529100612453266
7. Kornell, N., & Bjork, R. A. (2008). “Learning concepts and categories: Is spacing the ‘enemy of induction’?” Psychological Science, 19(6), 585–592. https://doi.org/10.1111/j.1467-9280.2008.02127.x