Rereading feels productive, but does it stick? Retrieval practice, pulling an answer out of memory instead of looking at it again, tends to build more durable learning than rereading, so close the notes and quiz yourself.
This method is one of the study techniques that work, and it combines well with the others.
| Item | Value |
|---|---|
| Earliest report of the effect | Over 100 years ago |
| Word lists in a classic forward-effect experiment | Five word lists |
| Recall by the tested groups on the final list | About twice as many list 5 items |
| Transfer effect size reported for test-enhanced learning | d=0.58 (medium-large) |
The figures come from a Frontiers in Psychology review and a set of working notes on the testing effect. Research findings can change, so check the sources before you rely on them.
What retrieval practice actually means
When you try to recall a detail from memory, that act strengthens the memory, and when it is used on purpose as a study activity it is called retrieval practice. Researchers also call it the testing effect, and the working notes describe it in just these terms (source).
The idea is not new. The Frontiers review says the effect was reported over 100 years ago (Abbott, 1909), although much of the detailed research came later (source).
It also changes what a test is for. In typical classrooms, teachers and students imagine that learning happens during lectures or while reading, yet the notes argue that the tests themselves are an important part of the learning process (source).
Why does testing yourself beat rereading?
The Frontiers review summarises numerous studies showing that retrieving studied information can raise long-term retention more than repeated study or elaborative encoding does (source). In plain terms, recalling beats re-reading when the goal is remembering weeks later.
The catch is that it feels worse. The working notes say retrieval practice leads to more durable long-term memory than simply re-reading, even though students are less successful during the practice itself (source). A struggle that produces mistakes can still be doing the work, and the notes list Bjork’s idea of desirable difficulties as a related concept.
Two more details from the notes are useful. The effect has been shown even when the correct answer is not provided, though it may be less effective without feedback, and it has also been shown when the test-taker has open-book access to find the answer (source).

The forward effect: testing helps what you study next
Most talk about testing concerns remembering what you already studied. The review also describes a forward effect, where recall testing of studied material can increase long-term retention of new information studied afterwards (source).
One experiment by Szpunar and colleagues (2008) shows the pattern, as the review describes it (source):
- Participants studied five word lists before a final cumulative recall test.
- After each earlier list, groups did maths problems, restudied the words, or tried immediate free recall.
- Everyone was tested straight away on the last list, called list 5.
The groups that had been tested recalled about twice as many list 5 items as the two non-tested groups, and they made far fewer intrusions from earlier lists. Because restudy did not produce the same result, the authors read it as a retrieval-specific effect.
The review adds that the effect reaches beyond word lists, to complex text, narratives, pictures, videos, and faces and names. These are laboratory findings, so treat them as a reason to test yourself between topics, not as a promise of a result.
How to practise it in a study session
A simple routine follows from the research. Try it on one topic first.
- Read or work through the material once.
- Close the notes and write down everything you can remember, or answer questions you wrote earlier.
- Open the notes and check your answers, because feedback matters when the correct answer is not supplied.
- Mark the gaps, then test those gaps again later.
Format matters too. The notes say the testing effect is generally measured to be more pronounced for production tests, such as short answer and essay, than for discrimination tests such as multiple choice or true-false (source). Writing answers from scratch is therefore a sensible default.
Retrieval can be spread over several days, and the guide on how to use spaced repetition covers that timing. If you want to mix topics within a session, the page on interleaving study explains how.
Limits worth knowing
Retrieval practice is not a cure-all. The notes say the testing effect may be diminished or inverted for immediate tests, so quizzing yourself a minute after reading may show little (source).
It may also depend on the goal. The same notes list a 2015 study by van Gog and colleagues whose title reports that testing after worked-example study did not enhance delayed problem-solving performance compared with restudy.
On the encouraging side, the notes mention Pan and Rickard (2018), who found a medium-large effect (d=0.58) for transfer across initial and final test formats (source). None of this guarantees grades, so use it as one tool among several.
Frequently asked questions
Is retrieval practice the same as taking a test?
Show answer
Not exactly. A graded exam measures what you know, while retrieval practice is any attempt to recall that you do for learning, such as flashcards, a blank-page summary or self-written questions.
Do I need an answer key?
Show answer
Having one is wise. The research notes report benefits even without the correct answer, but they also say practice may work less well without feedback, so check your attempts afterwards.
Which question type should I choose?
Show answer
Questions that make you produce an answer, like short answers, tend to show a larger effect than recognising an option from a list, according to the notes. Multiple choice can still help when nothing else is at hand.
Can I rely on it alone?
Show answer
There is no basis for that in these sources. Findings come largely from laboratory work, and results for problem solving were mixed, so combine it with other methods and your teacher’s guidance.



