Desirable difficulty: the study methods that work feel like the ones that are not working

That gap is worth understanding before you choose how to spend an evening.

The trap, stated plainly

Rereading a chapter feels smooth. Closing the book and trying to reconstruct it feels rough, slow, and faintly humiliating. Most students conclude that the smooth one is working. The evidence points the other way, and the reason the mistake is so durable is that the feeling arrives immediately while the result arrives weeks later.

What the study actually measured

Kirk-Johnson, Galla and Fraundorf in 2019 had learners try two contrasting strategies, rate them, then pick one for future use. The mediation result is specific: “participants who perceived a strategy as more effortful rated it as less effective for learning and, in turn, were less likely to choose it for future study.” Effort was read as evidence of failure.

And what happened to retention

The same paper reports the other half: “choosing the more effortful strategy was associated with better long-term retention (Study 3), contrary to participants' judgments.” So the strategy that felt worse produced more learning, and the students who felt it was worse were less likely to use it again. The judgment and the outcome moved in opposite directions.

Why fluency is a bad instrument

The feeling of ease during study is mostly a report on how available the material is right now, not on how available it will be in May. Familiar text reads quickly, and quick reading feels like knowing. This is why the check has to be external: close the book. Retrieval either happens or it does not, and that is not a feeling.

Which difficulties are the desirable ones

Not all difficulty helps. Bad handwriting and a noisy room make studying harder and worse. The useful ones are structural: retrieving instead of reviewing, spacing instead of massing, mixing problem types instead of blocking them, and producing an answer instead of recognising one. Marsh, Edelman and Bower found the production advantage extends even to remembering the context an item was learned in.

What to do with this on a Tuesday night

Pick the version of the task that makes you generate rather than recognise. Cover the answer. Say the definition before you read it. Do problems from mixed chapters rather than one. Each of these makes the session feel less productive and the exam go better, and knowing why makes the feeling easier to ignore.

How to tell difficulty from confusion

There is a line. Desirable difficulty is struggling to retrieve something you have understood. Confusion is struggling with something you never understood. The first is the work; the second means going back to the explanation. If you cannot produce even a rough version of the idea, testing yourself on it is not the right step yet.

The measurement that settles it

Run your own comparison. Take two similar topics, study one by rereading and one by closed-book recall, then test both a week later. Dunlosky and colleagues found the two techniques with the strongest generalisable evidence were practice testing and distributed practice, and a week is long enough for the difference to show up in your own numbers.

By Mert Erarslan (Exam prep and subject routes)

Last updated: 2026-08-21 · Written with AI assistance and reviewed before publishing.