Harder now, stronger later

The gist. Testing yourself on new material straight after learning it feels harder and produces worse recall minutes later. But in a classic memory study, the self-testers remembered 61% of a passage a week on, against 40% for learners who had just reread it four times, even though they had read it far fewer times.

What the study found

In a study of 180 university learners, one group read a short passage once and then tested themselves on it three times from memory, writing down everything they could recall with no notes and no feedback. A second group reread the same passage four times instead of testing. Five minutes after learning, the rereaders did better: 83% recalled against 71% for the self-testers.

A week later the pattern flipped. The self-testers recalled 61% of the passage, against 40% for the rereaders, even though the self-testers had read the passage far fewer times on average across the study (3.4 times against 14.2). A third group who studied three times then tested once landed in between at both points, recalling 78% after five minutes and 56% a week later.

What it might mean for your classroom

This is a memory-lab study of university learners recalling short factual passages, not a UK classroom trial. The moves below reflect what the self-testers actually did, worth trying with one class before building a whole scheme around it.

• Write it, don't just reread it. Learners close the book and try to recall the whole passage from memory in writing, rather than scanning the text again.
• Skip the feedback and let the attempt do the work. The practice tests in the study had no correction or answer key. The effortful attempt itself produced the later advantage.
• Fewer rereads, more attempts to recall. The self-testers read the passage 3.4 times on average against 14.2 for the rereaders, and remembered more a week on. Building this alongside routine retrieval practice may help it stick as a habit rather than a one-off task.

Treat it as an experiment, not a recipe: this is university learners recalling short passages under timed lab conditions, not school-age learners doing normal classwork, so try the write-it-from-memory step with one class for two weeks and watch what actually changes before widening it. It sits inside Bjork's desirable difficulties framework, the idea that some of the hardest-feeling study methods produce the strongest long-term learning.

How solid is this?

Secure. Two controlled experiments in the same paper show the same reversal, with large, precisely reported effect sizes, and the underlying testing effect is one of the most extensively replicated findings in memory research. Three limits are worth holding onto: the study population was university learners in the US, not UK school-age classes; the comparison is specifically free recall against rereading, not any task that merely feels harder; and retention was measured only out to one week, so this doesn't show whether the advantage lasts a whole term.

The source

Roediger, H.L. & Karpicke, J.D. (2006). ‘Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention.’ Psychological Science, 17(3), 249-255.

Read the original study