You have reread your notes three times. The pages feel familiar, the headings ring a bell, you feel ready. Then the exam starts and the answer will not come. Psychology has a name for that gap, and a remedy: test yourself. The testing effect is one of the most robust findings in learning research, and it changes how you should study. To apply it without writing every question yourself, Mindlet’s MCQ generator builds questions from your course material.
What is active recall?
We usually think of a test as a measurement: it tells you what you know. The testing effect shows it does more. Retrieving information from memory strengthens it, more than rereading it for the same amount of time. The test is a learning event in its own right.
Researchers also call this retrieval practice. Active recall is the everyday name for the same idea: it is the effort of retrieving that consolidates the memory.
What do the experiments show?
Roediger and Karpicke, 2006: reread or test?
Henry Roediger and Jeffrey Karpicke had students study short science-style passages (Roediger and Karpicke, 2006, Psychological Science 17(3), 249-255). In their second experiment, three groups went through four study periods, each spent either reading the passage or writing down everything they could remember:
| Group | What they did | Recall after 5 minutes | Recall after one week |
|---|---|---|---|
| Read × 4 | Read the passage four times | 83% | 40% |
| Read × 3, test × 1 | Read three times, then wrote what they remembered | 78% | 56% |
| Read × 1, test × 3 | Read once, then tested themselves three times | 71% | 61% |
The result flips over time. Right after studying, rereading looks better. A week later, the students who had tested themselves three times remembered far more, even though they had read the passage 3.4 times on average, against 14.2 times for the rereading group.
The most telling detail came from a questionnaire. Students who had reread four times were more confident that they would remember the passage a week later than those who had tested themselves. They were wrong.
Karpicke and Roediger, 2008: don’t drop what you know
Two years later, the same researchers had students learn 40 Swahili words (Karpicke and Roediger, 2008, Science 319(5865), 966-968). Once a word was recalled, it was, depending on the group, dropped from study, dropped from testing, or dropped from both.
A week later, the groups that kept being tested on every word recalled about 80 percent. The groups whose known words were removed from testing recalled only 36 and 33 percent, even when they kept restudying them. And every group predicted roughly the same score, around half the words: nobody saw the difference coming.
What the reviews say
These experiments are not isolated. Christopher Rowland’s meta-analysis brings together dozens of studies comparing testing with restudying (Rowland, 2014, Psychological Bulletin 140(6), 1432-1463). It confirms the effect and adds that tests where you produce the answer give a larger benefit than tests where you recognize it. The review by Dunlosky et al. (2013) rates practice testing as one of only two high-utility study techniques, alongside spacing.
Why does rereading feel like knowing?
When you reread, the text is in front of you. Every sentence seems obvious because you recognize it. That fluency is easy to mistake for knowledge. Psychologists call it an illusion of competence.
A test removes the text. It puts you in the exam situation: nothing in front of you, a question and your memory. If the answer does not come, you find out now rather than on the day. It is uncomfortable, and that discomfort is exactly what makes it useful.
How do you grade yourself without fooling yourself?
Self-assessment, done badly, is the best way to feel reassured for no reason. A few simple rules:
- Answer before you look. In writing, out loud, or at least fully formed in your head. “It’s on the tip of my tongue” is not an answer.
- Compare point by point. The expected answer has three parts? Count how many you had.
- Treat hesitation as a partial miss. An answer dragged out with effort will soon be forgotten. It should come back early.
- Test again later. Getting a question right five minutes after reading it proves very little. The test that counts is the next day, then the next week.
- Correct your mistakes. An uncorrected error can stick. With multiple choice especially, feedback matters: it stops you from keeping a wrong option as the right answer (Butler and Roediger, 2008, Memory & Cognition 36(3), 604-616).
Which kinds of test should you use?
| Format | Effort required | When to use it |
|---|---|---|
| Blank page | Write everything you remember about a chapter | Right after a lecture, and again before the exam |
| Flashcards | Retrieve a short answer | Definitions, dates, formulas, vocabulary |
| Open questions | Write an explanation | Reasoning, cause and effect |
| Multiple choice | Pick from options | Covering a lot of ground fast, practicing the exam format |
| Past papers and practice exams | Work through a full, timed paper | In the last few weeks |
Combining them works best. The blank page trains free recall, the hardest kind. Flashcards keep facts alive over time. Multiple choice covers a lot of ground quickly. A practice exam prepares you for real conditions. Our guide on how to make good flashcards covers the card side in detail.
What does active recall look like on a real chapter?
Say you have a biology chapter on cell division. Here is how to work through it with active recall instead of rereading.
- Read once, to understand. Clear up anything confusing now. This is the only reading pass.
- Close the book and write. On a blank page, write everything you remember: the phases, what happens in each, the key terms. Do not look anything up.
- Check and mark. Open the chapter and compare. Circle what you missed or got wrong. These gaps are your to-do list.
- Turn the gaps into questions. “What happens to the nuclear envelope during prophase?” “How does meiosis differ from mitosis in the number of divisions?”
- Answer them tomorrow. Without looking. Then in three days, then in a week, alongside questions from other chapters.
The whole process takes little longer than two careful rereads, and it tells you exactly what you know. Rereading cannot do that: everything looks familiar when it is in front of you.
How Mindlet uses the testing effect
Mindlet is built on this principle: you study by answering, never by rereading. Depending on the subject, your material becomes a dozen exercise formats: flashcards, multiple choice, true or false, cloze deletions, open questions graded by a language model, and more. Flashcards ask you to grade yourself; the other formats are checked automatically. The app interface is in French, while your questions can be generated in nine languages, including English.
When an exam gets close, the practice exam gives you a full paper built from your material, graded out of 20 (the French scale), with no corrections during the test and a breakdown of your mistakes at the end.
The testing effect pairs naturally with spaced repetition: testing yourself at the right moment is the most reliable study method research knows.
FAQ
Frequently asked questions
What is the testing effect?
It is the finding that retrieving information from memory makes you retain it better over time than rereading it for the same amount of time. A test is not only a measure of what you know; it is a way of learning. Roediger and Karpicke demonstrated it clearly in 2006.
Why isn't rereading enough?
Rereading creates a feeling of familiarity: the text looks known, so you assume you know it. But recognizing is not recalling. In testing effect experiments, students who reread felt more confident than those who tested themselves, yet remembered less a week later.
Are multiple-choice questions as good as open questions?
Both help, but tests that make you produce the answer give a larger benefit on average than tests where you recognize the right option, according to Rowland's 2014 meta-analysis. A well-built multiple-choice quiz, with plausible wrong options and feedback, is still a useful tool.
How do I grade myself honestly?
Answer before you look at the solution, ideally in writing or out loud. Then compare point by point. If the answer came with hesitation or was incomplete, treat it as not yet learned. And test yourself again a few days later: only delayed recall tells you whether you have retained it.
