Desirable difficulties

What are desirable difficulties?

Desirable difficulties are study conditions that slow down visible progress during learning but improve long-term memory and the ability to use what you learned in new situations.^[1][2] The examples given most often are spacing study sessions instead of cramming, mixing topics instead of studying one at a time, testing yourself instead of rereading, and varying the conditions of practice.^[2]

The idea runs against common study habits, which aim to make studying smooth: rereading until a text feels familiar, practising one type of problem until it goes well, always working in the same quiet place. Desirable difficulties do the opposite on purpose. They rely on a gap between how well you do while studying and how much you actually learn.

Why can making learning harder help?

Making learning harder can help because performance during study and long-term learning are different things, and the conditions that boost the first often do little for the second.^[2][3] Performance is what you can see during a session, such as answering correctly. Learning is the lasting change that shows days or weeks later, and it can only be inferred.^[3]

The theory behind the idea, called the new theory of disuse (1992), gives every memory two strengths.^[4] Storage strength is how well something is learned and connected to what you already know. Retrieval strength is how easily you can bring it to mind right now; it fades with time and changes of context, while storage strength, according to the theory, is never lost. The key assumption is that the easier something is to recall when you review it, the less that review adds to storage strength.^[4] So reviewing right after studying adds little, while reviewing after some forgetting does more work. This is why spacing and testing feel hard: they make you retrieve memories that are less accessible.

What are examples of desirable difficulties?

The best-studied desirable difficulties are spacing, retrieval practice, interleaving, varying the conditions of study and generating answers yourself.^[2] Each one has its own body of research; the table links to the pages that cover them in depth.

Two of these are less known. Varying conditions: in one experiment, people who studied the same material in two different rooms recalled more than people who studied it twice in the same room.^[2] Generation: words that people produced themselves, for example from a cue, were remembered better than the same words simply read.^[5] Spacing also connects to the forgetting curve: the gap between sessions lets some forgetting happen, which makes the next review more useful.

What makes a difficulty desirable rather than harmful?

A difficulty is desirable only when it triggers the kind of processing that builds memory, such as retrieving, comparing or reconstructing, and when you have the background knowledge to respond to it successfully.^[2][4] Otherwise it becomes an undesirable difficulty. Generating an answer helps, for instance, only if earlier learning lets you produce it, or at least bring related knowledge to mind.^[4] The right level of difficulty therefore depends on what you already know.

Not every way of making study harder counts. A 2011 study reported that students remembered more from material printed in hard-to-read fonts.^[6] Later studies often failed to find this, and a 2018 meta-analysis of studies on learning from text found no reliable benefit of such fonts for recall or for applying the material, although learners felt they had learned less.^[7] A harder font does not seem to add the kind of effort that helps.

Why do desirable difficulties feel less effective?

Desirable difficulties feel less effective because people judge their learning by how easily things come to mind during study, which reflects retrieval strength, not lasting learning.^[2][8] Rereading, for example, makes a text feel familiar, and that familiarity is easy to mistake for understanding.^[2]

In one study, people learned the styles of several painters from examples. They identified new paintings better when paintings by different artists had been mixed, yet they rated studying one artist at a time as more effective, even after their own test results had shown the opposite.^[9] The lesson is that a smooth session is weak evidence of learning: the conditions that make study feel slower can be the ones that make learning last.^[3]

How do you use desirable difficulties when studying?

Build difficulties into your routine at a level you can succeed at: test yourself, space your sessions and mix topics once you know the basics.^[2]

• Test yourself instead of rereading. Close the book and recall, or use flashcards, then check your answer. See active recall.
• Space your sessions. Come back to a topic after a day or more instead of going over it again straight away. Spaced repetition apps such as Anki schedule this for you.
• Mix problem types once each one is familiar, so you have to work out which method applies.
• Try before you look. Attempt the answer, the next step or a summary before you read it.
• Vary the setting and examples. Study in different places, and practise with different versions of a problem.
• Lower the difficulty if you keep failing. If you get almost nothing right, go back to the material first: a difficulty you cannot overcome does not help.^[4]
• Judge by delayed tests. Check what you remember a few days later, not at the end of the session.