Desirable Difficulties Explained: Why Struggle Produces Better Learning

In 1994, the cognitive psychologist Robert Bjork published a paper called Memory and Metamemory Considerations in the Training of Human Beings in which he proposed one of the most counterintuitive ideas in learning science. The strategies that feel most effective during practice — easy repetition, immediate feedback, smooth progress — produce worse long-term learning than strategies that introduce specific kinds of difficulty into the practice itself. Bjork called these productive struggles desirable difficulties, and the framework has shaped subsequent research on study techniques, sports training, and skill acquisition.

This article explains what desirable difficulties actually are, the four main techniques that fall under the framework, why they work at the cognitive level, what the research shows about their effect sizes, and how to apply them in practice without making learning needlessly miserable.

What Desirable Difficulties Are

Bjork's framework distinguishes two kinds of difficulty in learning.

Undesirable difficulties are obstacles that prevent learning entirely — confusing instructions, irrelevant distractions, badly designed materials, content beyond the learner's current level. These produce frustration without producing the cognitive engagement that leads to durable knowledge. Reducing undesirable difficulty is straightforward and almost always helpful.

Desirable difficulties are obstacles that slow down or complicate the learning process in ways that feel counterproductive but actually produce stronger and more transferable learning. Examples include spacing study sessions over time (rather than massing them), interleaving topics within a session (rather than blocking them), testing yourself instead of re-reading, and varying the conditions of practice rather than keeping them constant.

The central insight is that learning and felt-fluency during practice are not the same thing. Practice that produces high in-session performance often produces poor long-term retention. Practice that produces low in-session performance often produces excellent long-term retention. The discrepancy is what makes desirable difficulties counterintuitive — they require trusting research evidence over the immediate experience of practice.

The Four Main Desirable Difficulties

The framework includes four specific techniques, each backed by extensive research.

Spacing

Distributing practice across time rather than concentrating it. Five 10-minute sessions across a week produce dramatically better retention than one 50-minute session at the same total time. The brain consolidates memory during the gaps between sessions; concentrated practice bypasses consolidation.

The size of the effect is substantial. Carpenter et al.'s 2018 meta-analysis found average retention gains of approximately 20% from spaced practice over massed practice across hundreds of studies. The effect holds across age groups, content types, and time scales. See our forgetting curve explained and science of spaced repetition guides for the underlying memory science.

The difficulty is that spacing feels less productive. A single 50-minute session feels like real progress; five 10-minute sessions feel scattered. The fluency illusion that develops in a concentrated session does not develop in spaced sessions, so learners often rate their concentrated practice as more effective even when delayed tests show the opposite.

Interleaving

Mixing topics or problem types within a study session rather than blocking them. Practising quadratics, geometry, and trigonometry rotated across a single session produces dramatically better learning than the same total problems done in three sequential blocks.

The Rohrer and Taylor (2007) study is the canonical reference: blocked practice produced 89% accuracy during the session compared to 60% for interleaved practice, but one week later the interleaved group performed at 63% accuracy while the blocked group dropped to 20%. The within-session pattern reversed completely on the delayed test. Our interleaving vs blocked practice guide covers the research in depth.

The difficulty is that interleaving forces you to identify which technique applies to which problem on every problem — the constant context-switching feels harder than the smooth-flowing blocked practice. The harder identification work is exactly what produces the transfer to novel contexts that blocked practice fails to deliver.

Generation (Retrieval Practice)

Producing information from memory rather than re-reading it. Closing the book and trying to recall what you just read, even when you only remember some of it, produces dramatically stronger memory than reading the same passage a second time.

Roediger and Karpicke's (2006) foundational study found students using retrieval practice retained 50% more material at a one-week delay compared to students re-reading. The 2017 meta-analysis by Adesope, Trevisan, and Sundararajan of 118 testing-effect studies confirmed an average effect size of approximately 0.50. Our retrieval practice explained and testing effect explained guides cover the underlying research.

The difficulty is that retrieval feels effortful and surfaces failure. Re-reading produces the comfortable feeling of familiarity; retrieval produces the uncomfortable feeling of "I should know this but I do not quite." The discomfort is the technique working — failed retrieval followed by feedback produces stronger memory than successful re-reading.

Variation

Varying the conditions of practice rather than keeping them constant. Practising a tennis serve from different positions, in different weather, with slightly different ball weights produces better generalisation than perfect repetition of the same serve under identical conditions. The variation forces the learner to extract the underlying principle rather than memorising a single instance.

The research on motor skill learning is particularly strong here. Hall, Domingues, and Cavazos (1994) studied baseball batters who interleaved different pitch types in practice versus batters who blocked single pitch types. The interleaved batters hit novel pitches in games measurably better than the blocked batters, despite worse in-practice performance.

Variation applies to cognitive learning too. Studying material in different rooms, at different times of day, in different orderings, with different cue contexts produces stronger and more flexible memory than perfect-repetition study. The "associative interference" that variation seems to produce in the moment is actually building cross-context retrieval ability.

The difficulty is that variation feels like inefficiency. Repetition under identical conditions feels like building skill; variation feels like making practice unnecessarily complicated. The research consistently shows the opposite is true for long-term outcomes.

Why Desirable Difficulties Work

Three cognitive mechanisms explain why the framework holds up so consistently across research.

Storage Strength vs Retrieval Strength

Bjork's foundational distinction is between storage strength (how well-encoded a memory is in the long term) and retrieval strength (how readily accessible it is at a given moment). Easy practice builds retrieval strength quickly — the material is fresh, retrieval is fluent. But it does not build storage strength as efficiently. Difficult practice forces effortful retrieval, which builds storage strength even if retrieval strength temporarily lags.

The implication is that the felt-difficulty of practice partly tracks how much storage strength is being built. Practice that feels too easy is producing retrieval strength that will decay; practice that feels difficult is often building storage strength that will last.

Effortful Processing

The cognitive psychology principle of effortful processing predicts that information processed with more cognitive engagement produces stronger and more durable memory. Reading a passage at normal speed produces less effort than reading it slowly with comprehension checks; rereading produces less effort than testing yourself on the material. Desirable difficulties impose effort that the learner would otherwise not engage in voluntarily.

This is related to the broader levels of processing framework — material processed at deeper, more meaningful levels is retained better than material processed at shallower levels. Desirable difficulties push learners toward deeper processing, often against their stated preferences for easier practice.

Discrimination and Transfer

Several of the desirable difficulties (particularly interleaving and variation) specifically target the transfer problem — applying learned material to novel contexts. Blocked practice teaches you to solve the problems you have already seen; interleaved practice teaches you to identify which kind of problem you are looking at and choose the right approach. The first skill is what tests reward in the short term; the second is what real-world application requires.

This is why desirable difficulties matter particularly for skills that need to transfer outside the practice context. A maths student who can solve every problem in the textbook chapter but cannot identify which approach to use on a novel problem has built retrieval strength without transfer ability. Desirable difficulties help close that gap.

What the Research Shows

The desirable difficulties framework has accumulated substantial empirical support over three decades. A few headline findings:

• Spaced over massed practice produces approximately 20% better retention at delayed tests across hundreds of studies (Carpenter et al., 2018).
• Interleaved over blocked practice produced 43-percentage-point retention advantages in the Rohrer and Taylor 2007 study, with effect sizes around 0.60 in subsequent meta-analyses.
• Retrieval over re-reading produces effect sizes around 0.50 across 118 studies (Adesope, Trevisan, and Sundararajan, 2017).
• Variation over repetition produces stronger transfer to novel contexts across motor-skill, cognitive, and language-learning studies.

In each case, the effect is larger at delayed tests than at immediate tests, and opposite in sign at immediate tests in many studies — the easier practice produces better immediate performance and worse long-term retention.

This is the most important practical implication: studies should not be evaluated by how they feel during practice, nor by immediate test performance, but by delayed test performance. The literature is full of cases where the practice condition that won the immediate test lost dramatically on the one-week or four-week follow-up.

How to Apply Desirable Difficulties in Practice

The framework translates into specific study habits that produce measurable benefits.

Replace Re-Reading With Testing

The single highest-leverage change for most adult learners. After every reading session, close the source and write down everything you remember before checking. The discomfort of the technique is the storage strength being built. Use spaced-repetition apps like Anki, RemNote, or Chunks to automate the spacing. See our best spaced repetition apps 2026 guide for the practical infrastructure.

Spread Practice Sessions

Take whatever total study time you have planned for a topic and divide it across more sessions of shorter length. Four one-hour sessions across four days produces better retention than one four-hour session in a single day. For most adult learners, the constraint is finding the daily slot — see our how to study effectively for adults guide for the 20-minute daily routine.

Mix Topics Within Sessions

For any subject that involves multiple problem types or topics, deliberately interleave them within each session rather than blocking them. A maths student should mix problem types within a worksheet; a language student should mix grammar, vocabulary, and listening within each session; a history student should switch between periods rather than studying one period exclusively per session.

Vary the Conditions

Study in different rooms, at different times of day, with different background contexts. The variation feels disruptive but builds the cross-context retrieval ability that makes knowledge accessible outside the original learning environment. For exam preparation specifically, varying conditions during practice produces measurable benefits at the actual exam where conditions cannot be controlled.

Use Discomfort as a Signal

The most important meta-skill is to learn to read your own discomfort during practice as a positive signal rather than a negative one. The fluency illusion that builds during easy practice is misleading; the discomfort that arises during effortful practice often tracks real learning. Practices that feel suspiciously easy are often building retrieval strength that will not last; practices that feel difficult are often building storage strength that will.

Common Mistakes With Desirable Difficulties

Making Practice Too Hard

The framework is specifically about desirable difficulties, not all difficulties. Practice that exceeds the learner's current capacity produces frustration without learning. The right difficulty is at the edge of current ability — challenging enough to require effort, achievable enough to allow at least partial success. Material far above the learner's level is undesirable difficulty.

Confusing Difficulty With Quantity

More problems, more reading, or more practice does not by itself constitute desirable difficulty. The framework is about the structure of practice rather than its volume. A learner who does twice as many problems but blocks them by type has not introduced desirable difficulty; a learner who does the same number of problems but interleaves them has.

Giving Up When It Feels Hard

The most common failure mode. Desirable difficulties feel uncomfortable in the moment, which produces the natural impulse to abandon them in favour of easier practice. Practices like retrieval and interleaving need to be applied consistently over weeks before the gains become visible in delayed tests. Most learners who try desirable difficulties once and stop never see the benefits the research promises.

Skipping Initial Skill Building

The framework applies to practice after initial skill acquisition. A novice who is interleaved across topics from the first lesson produces confusion rather than learning. The optimal pattern is 10-15 minutes of focused initial exposure on a single new topic — essentially blocked practice — followed by interleaved practice across topics the learner has at least basic familiarity with. Trying to apply desirable difficulties before the foundation exists is counterproductive.

Treating It as Optional

The framework is consistently effective only when applied as a sustained study practice rather than as occasional adjustments. A learner who interleaves once a month while blocking most of the time is not actually applying interleaving; a learner who tests themselves once a week while re-reading the rest of the time is not actually applying retrieval practice. The benefits come from consistent application across the entire study workflow.

Frequently Asked Questions

What are desirable difficulties?

Desirable difficulties are study practices that introduce specific kinds of difficulty into the learning process in ways that feel counterproductive but actually produce stronger and more transferable learning. The concept was developed by cognitive psychologist Robert Bjork in 1994. The four main techniques are spaced practice (rather than massed), interleaving topics (rather than blocking), retrieval practice (rather than re-reading), and varying conditions of practice (rather than keeping them constant).

Who developed the desirable difficulties framework?

Robert Bjork, a cognitive psychologist at UCLA, developed the desirable difficulties framework in his 1994 paper Memory and Metamemory Considerations in the Training of Human Beings. His broader work on memory, learning, and the relationship between practice and retention has been one of the most influential research programmes in cognitive psychology of the past four decades.

Why are desirable difficulties effective?

Three cognitive mechanisms account for the effect. Bjork's distinction between storage strength (long-term encoding) and retrieval strength (immediate accessibility) shows that easy practice builds retrieval strength without efficiently building storage strength. Effortful processing produces deeper, more durable memory than passive exposure. Discrimination and transfer practices like interleaving build the cross-context retrieval ability that real-world application requires. Together, these mechanisms explain why practice that feels harder often produces better long-term learning.

What are the four desirable difficulties?

The four main desirable difficulties are spacing (distributing practice across time rather than concentrating it), interleaving (mixing topics or problem types within a session rather than blocking them), generation or retrieval practice (producing information from memory rather than re-reading), and variation (varying the conditions of practice rather than keeping them constant). All four are backed by extensive research and consistently produce better long-term retention than easier alternatives at the same total practice time.

Is harder practice always better?

No. The framework is specifically about desirable difficulties, not all difficulties. Practice that exceeds the learner's current capacity produces frustration without learning. The right difficulty is at the edge of current ability — challenging enough to require effort, achievable enough to allow at least partial success. Material far above the learner's level is undesirable difficulty and should be avoided until the foundation supports it.

What is the difference between storage strength and retrieval strength?

Bjork's distinction is between how well-encoded a memory is in the long term (storage strength) and how readily accessible it is at a given moment (retrieval strength). Easy practice quickly builds retrieval strength — the material is fresh in mind — but does not efficiently build storage strength. Difficult practice forces effortful retrieval, which builds storage strength even when retrieval strength temporarily lags. The implication is that practice should be evaluated by delayed performance (storage strength) rather than immediate performance (retrieval strength).

Should children study using desirable difficulties?

Yes, with the caveat that initial skill acquisition still benefits from focused blocked practice. The 2015 Rohrer, Dedrick, and Stershic study with seventh-graders found interleaved maths practice produced 25% better delayed-test scores than blocked practice. The framework applies across age groups, and most school curricula default to heavy blocking rather than the more effective interleaved alternative. Adding interleaved review sets at the end of each unit produces measurable gains for students of any age.

Summary

Desirable difficulties, a framework developed by cognitive psychologist Robert Bjork in 1994, are study practices that introduce specific kinds of difficulty into learning in ways that feel counterproductive but produce dramatically better long-term retention. The four main techniques are spaced practice (distributing study over time), interleaving (mixing topics within sessions), retrieval practice (testing yourself rather than re-reading), and varying the conditions of practice. Each is backed by extensive research showing substantial effect sizes — typically 0.50 standard deviations or larger in cognitive research terms. The mechanism is that easy practice builds retrieval strength quickly but storage strength inefficiently, while effortful practice builds the durable memory that lasts. The central practical implication is to evaluate study practices by delayed test performance rather than by immediate performance or felt-fluency during practice. The strategies that feel most effective are not the strategies that are most effective; overriding the fluency illusion is what unlocks the gains. Apply the framework consistently — replace re-reading with testing, spread practice across more sessions of shorter length, mix topics within sessions rather than blocking them, and vary the conditions of practice deliberately. The compound effect across months of consistent application is larger than almost any other change to how you study.