Desirable Difficulties: Why Easy Studying Fails

Desirable difficulties are study conditions that make learning feel harder and slower in the moment, yet produce stronger memory later. The psychologist Robert Bjork named them in 1994, and the list is short and specific: space your sessions, mix your topics, test yourself instead of rereading, and vary the conditions you practise under.

The idea is uncomfortable because it inverts the signal most students steer by. A session that goes smoothly, where the material flows and the answers arrive quickly, is the session that tends to leave the least behind. Understanding why that happens turns a frustrating study session from evidence that you are struggling into evidence that the method is working.

What are desirable difficulties?

They are training conditions that appear to impede performance while you are practising, but that yield greater long-term benefits than their easier counterparts. That phrasing comes from the UCLA Bjork Learning and Forgetting Lab, which describes how Robert Bjork dubbed these difficult but effective conditions desirable difficulties in 1994.

The fuller argument appears in a 2011 chapter by Elizabeth Ligon Bjork and Robert Bjork with an unusually honest title, Making Things Hard on Yourself, But in a Good Way. They name four: varying the conditions of learning rather than keeping them constant and predictable, interleaving instruction on separate topics rather than grouping it by topic, spacing rather than massing study sessions, and using tests rather than presentations as study events.

Notice what is absent from that list. There is nothing about studying for longer, nothing about suffering, and nothing about making material harder for the sake of it. Every item is a change to the arrangement of practice, not to its quantity. Two students can do the same three hours of work and get very different results from how those hours are shaped.

Why does easy studying feel like it is working?

Because you are reading a signal that measures the wrong thing. The Bjorks build their case on a distinction between learning and performance: performance is what can be observed during a study session, while learning is the more permanent change you are actually after, and current performance can be a highly unreliable index of whether learning has occurred.

Underneath that sits a sharper theoretical claim. The Bjorks separate storage strength, which reflects how entrenched a memory is with related knowledge and skills, from retrieval strength, which reflects how accessible that memory is right now and is heavily influenced by recency of study. Easy conditions raise retrieval strength fast. A chapter you have just read is maximally accessible, so every idea in it feels available. Storage strength, the thing that determines whether it is still there in three weeks, moves far more slowly and gives off no sensation at all.

The consequence is the whole problem in one sentence: learners who interpret current retrieval strength as storage strength become susceptible to preferring poorer conditions of learning to better ones. This is the same machinery behind the illusion of competence, where a reread chapter registers as mastered because the answer never left the page. Fluency is real. It is simply a fact about the present moment rather than a forecast.

What are the four desirable difficulties?

Each one has its own evidence base, and each fails in a different way when applied carelessly.

• Spacing rather than massing. Distributing study sessions on a topic instead of stacking them produces both good exam performance and good long-term retention, where cramming produces only the first. The Bjorks call the spacing effect one of the most general and robust effects across the entire history of experimental research on learning and memory. The practical layer is a calendar, which our guide to building a spaced repetition schedule lays out in detail.
• Testing rather than restudying. Using tests as study events rather than as measurement is the difficulty with the strongest evidence behind it, and the one students most often skip because producing an answer is slower than rereading one. The pairing with spacing is covered in our comparison of active recall and spaced repetition.
• Interleaving rather than blocking. Mixing related problem types forces you to identify which method applies before you apply it, which blocked practice removes entirely. In a classic 1979 comparison by Shea and Morgan, participants who practised movement patterns in a blocked order improved faster during training, then performed far worse than the interleaved group when everyone returned ten days later. Our guide to the interleaving study method covers how to build a mixed pool and where the technique underdelivers.
• Varying the conditions of practice. When instruction happens under constrained and predictable conditions, learning becomes contextualised: easily retrieved in that context, weaker anywhere else. Studying the same material in two different rooms rather than twice in the same room increased later recall in a 1978 study by Smith, Glenberg and Bjork, a result that runs directly against the standard advice to find one quiet spot and do all your studying there.

These are related rather than independent. Interleaving a mixed pool automatically introduces spacing between repetitions of any single topic, and running that pool as self-testing rather than review stacks a third difficulty on top. You do not need four separate habits.

When is a difficulty not desirable?

This is the part that gets lost when the concept is compressed into a slogan about embracing struggle. The Bjorks stop mid-chapter to emphasise the importance of the word desirable, and state plainly that many difficulties are undesirable during instruction and forever after.

Their criterion is functional. Desirable difficulties are desirable because they trigger the encoding and retrieval processes that support learning, comprehension, and remembering. If the learner lacks the background knowledge or skills to respond to them successfully, they become undesirable difficulties. The difficulty has to be one you can meet.

For a student, that translates into a few concrete failure modes. Shuffling problem types before you understand any of them individually produces confusion rather than discrimination. Self-testing on a chapter you have never read produces a blank page and a wasted evening. Spacing your reviews so far apart that every session starts from zero is not spacing, it is forgetting on a schedule. A difficulty is working when you are effortful and mostly correct. When you are effortful and mostly wrong, the answer is to go back and build the base first.

Bad handwriting, a noisy room, and a badly written textbook are also difficulties. None of them are desirable, because none of them ask memory to do anything useful.

How do you build desirable difficulties into a study session?

The changes are structural, and none of them require more hours.

1. Learn it once, properly, the easy way. First exposure is the wrong place for difficulty. Read the chapter, follow the worked example, sit through the lecture. The difficulties apply to practice, and practice comes after understanding.
2. Close the book before you judge yourself. Any assessment made while the material is visible is measuring retrieval strength you borrowed from the page. Commit to an answer from the cue alone, then check.
3. Break one long session into three short ones. The same total time, distributed across days, is the cheapest upgrade available. Nothing needs to be added.
4. Shuffle related topics into one pool. Keep the mixing inside a family of genuinely confusable material so that each question requires a decision about which method applies.
5. Change something about the setting. Different room, different order, different phrasing of the same question. Varied practice keeps the memory from binding to a single context that the exam hall will not reproduce.
6. Judge the method by a delayed test, never by the session. This is the discipline the whole idea depends on. A session that felt bad and a test a week later that went well is the outcome you are optimising for.

Making the harder version the default with GeniusPal

The honest obstacle with desirable difficulties is setup cost. Rereading requires nothing. Self-testing requires questions, interleaving requires questions from several topics, and spacing requires those questions to still exist next week. Each difficulty is easy to endorse and slightly annoying to arrange, which is why the easy version keeps winning at eleven at night.

That arrangement gap is where GeniusPal fits, and the fit is worth stating precisely rather than overselling. GeniusPal was not designed from this research, and it does not schedule anything for you: spacing still lives on your own calendar, the same way it does with a stack of paper flashcards. What lines up, honestly, is testing and interleaving. A generated quiz is a test used as a study event rather than as measurement, and the recall drill holds the answer back until you have committed to one before showing you whether you were right, which is retrieval practice rather than review. And because a study set is generated across everything in the file you uploaded, working through it mixes topics instead of blocking them, which is interleaving by default rather than by planning.

Upload the chapter, the lecture slides, or the notes you were about to read for a third time. GeniusPal accepts PDF, Word, PowerPoint, plain text, Markdown, and CSV files up to 10 MB, and the free plan covers two generations in total as a lifetime allowance rather than a monthly one.

None of this is required. A blank sheet of paper, a shuffled pile of past-paper questions, and the willingness to feel slow will deliver the same difficulties for free. What software removes is the setup cost, and with it the most common reason students who know all of this still end up rereading the chapter.