The Testing Effect

At a Glance

1. Quick Summary

The Testing Effect is the finding that pulling information out of memory strengthens it more than putting the same information in front of your eyes again. A student who reads a passage once and then writes down everything she remembers will usually beat a student who reads it twice, as long as the final test comes days later rather than minutes later. The first student feels worse while studying, because recall is effortful and exposes gaps, while rereading feels smooth and familiar. Learners trust that feeling, so they overwhelmingly choose rereading and highlighting over self-quizzing. Meta-analyses of lab and classroom work put the benefit at about half a standard deviation, but it depends on conditions: feedback matters, and a practice test on which you remember almost nothing helps little unless you see the answers afterward.

Simply About the Bias

The testing effect means you remember things better when you try to pull them out of your head than when you just read them again. Trying to remember is exercise for your memory. Reading again feels easier, but it does less.

Imagine a spelling test on Friday. On Monday you read the word list five times. Your friend reads it once, covers it, and tries to write each word from memory, then checks the ones she got wrong. Reading feels great, because every word looks familiar. Your friend feels a bit stuck. On Friday she spells more words right than you do. The words looked familiar to you, but you never practiced getting them out of your own head.

Here is a simple trick. After you read something you need to learn, close the book. Say or write what you remember. Then open the book and check what you missed. Do it again a day or two later. It will feel harder than rereading, and that is a sign it is working.

2. The Science Behind It

2.1. Discovery and History

People noticed this long before psychologists measured it. Francis Bacon wrote in 1620 that reading a text twenty times teaches it less well than reading it ten times while trying to recite it. William James made the same point in 1890 about "active" versus "passive" repetition.

The first systematic study came in 1917, when Arthur Gates at Teachers College, Columbia, published Recitation as a Factor in Memorizing. Schoolchildren studied nonsense syllables and short biographies. Some spent all their time reading; others spent part of it "reciting," looking away and trying to recall. Children who gave a large share of their time to recitation, often more than half, remembered more.

In 1939 Herbert Spitzer tested 3,605 sixth graders in 91 Iowa schools. They read short factual articles and took tests on different schedules over the following two months. Children tested soon after reading forgot much less over the following weeks than children whose first test came later. Spitzer concluded that tests are learning devices in their own right.

Education mostly ignored this. Tests stayed at the end of units as measuring instruments. The modern revival came from Henry Roediger and Jeffrey Karpicke (2006), who used prose passages and showed that rereading wins on a test five minutes later while testing wins clearly at two days and one week. Hundreds of studies followed, including classroom trials led by Pooja Agarwal, Mark McDaniel and Kathleen McDermott.

The meta-analyses also mapped the limits. Christopher Rowland (2014) found the effect reliable but larger when the practice test demanded recall rather than recognition, and larger with feedback. Without feedback, items the learner fails to retrieve get no boost, so a hard test with no answers given helps little. Tamara van Gog and John Sweller (2015) argued that the effect fades for highly complex, interconnected material; Karpicke and Althea Aue (2015) replied that the evidence for complex materials is stronger than that. The question is still open.

The "bias" is the learner's side of the story. Roediger and Karpicke's repeated readers were the most confident they would remember, and a 2009 survey by Karpicke, Andrew Butler and Roediger found that most students reread and very few practice recall. That neglect is closely tied to the Illusion of Competence (Fluency Illusion).

Key milestones:

• 1620: Francis Bacon notes in Novum Organum that reading combined with recitation beats reading alone
• 1917: Arthur Gates publishes Recitation as a Factor in Memorizing
• 1939: Herbert Spitzer shows in 3,605 Iowa sixth graders that an early test slows forgetting
• 1994: Robert Bjork introduces "desirable difficulties," which include testing
• 2006: Roediger and Karpicke publish "Test-Enhanced Learning" in Psychological Science
• 2011: Karpicke and Blunt report in Science that retrieval practice beat concept mapping
• 2014: Rowland's meta-analysis in Psychological Bulletin confirms the effect
• 2021: Yang, Luo, Vadillo, Yu and Shanks pool 222 classroom studies with 48,478 students

2.2. Key Researchers

2.3. Landmark Studies

"Test-Enhanced Learning" (Roediger & Karpicke, 2006)

College students read short science passages. Each passage was either read twice or read once and followed by a free-recall test with no feedback. A final test came five minutes, two days or one week later. At five minutes the rereaders did better. One week out, students who had taken the practice test recalled about 56 percent of the ideas against about 42 percent for those who had reread. A second experiment found the same reversal with three tests against three extra readings, and the repeated readers were the most confident.

"Metacognitive Strategies in Student Learning" (Karpicke, Butler & Roediger, 2009)

Of 177 surveyed college students, 84 percent listed rereading as a strategy and 55 percent named it their top one. Only 11 percent described practicing recall, and 1 percent (2 students) ranked it first. Asked what they would do after reading a chapter, 57 percent chose to reread and 18 percent to test themselves; the share choosing self-testing rose to 42 percent when they could reread afterward. A 2019 reexamination by Christof Kuhbandner and Katharina Emmerdinger qualified this: in their more detailed surveys, few students preferred rereading late in learning, and much rereading targeted passages they had not understood.

"Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping" (Karpicke & Blunt, 2011)

Students read science texts and then studied once, studied repeatedly, built a concept map with the text in view, or practiced free recall. A week later, retrieval practice beat concept mapping on verbatim questions and on inference questions that required connecting ideas, and it still won when the final test asked students to draw a concept map. Students had predicted the reverse. Concept-mapping researchers led by Joel Mintzes challenged the methods in a technical comment in Science, and Karpicke and Blunt replied in the same issue.

"Unsuccessful Retrieval Attempts Enhance Subsequent Learning" (Kornell, Hays & Bjork, 2009)

Participants answered questions they could not know, such as fictional trivia or guessing a weak associate of a cue word ("whale" for "mammal"). In one condition they guessed first and then saw the answer; in the other they saw question and answer together for the whole time. Even with correct guesses thrown out, guessing first produced better later memory. A failed attempt is not wasted if the answer follows quickly.

The Meta-Analyses (Rowland, 2014; Adesope, Trevisan & Sundararajan, 2017; Yang et al., 2021)

Rowland found a reliable advantage for testing over restudy, larger when the practice test required recall, with the data favoring effortful-retrieval accounts over semantic elaboration. Adesope and colleagues pooled 272 effect sizes and found practice tests beat restudying and every other comparison condition. Yang and colleagues kept to classrooms: 222 studies and 48,478 students gave an average effect of g = 0.499, larger with corrective feedback, with repeated quizzes, and when quiz and exam matched in format.

2.4. Neurological Basis

The neuroscience is thin. Most imaging studies are small, use word lists, and compare retrieval with restudy, which makes cause hard to separate from correlation.

In an fMRI study, Erik Wing, Elizabeth Marsh and Roberto Cabeza (2013) found that initial retrieval leading to later remembering engaged the anterior hippocampus, lateral temporal cortex and medial prefrontal cortex more than restudy did, and strengthened hippocampal connections with prefrontal regions. James Antony, Catarina Ferreira, Kenneth Norman and Maria Wimber (2017) proposed that retrieval works as a fast route to consolidation: reactivating a memory while awake may replay hippocampal-neocortical patterns much as sleep is thought to. That remains a hypothesis.

The metacognitive half needs no special machinery. People judge their learning partly by how easily material is processed in the moment. Rereading raises that ease and with it confidence; the struggle of recall lowers both.

3. Evolutionary Origins

Since the testing effect is a property of memory, the useful question is why memory would work this way.

Retrieval as a signal of need: John Anderson and Lael Schooler (1991) showed that human forgetting mirrors how often information recurs in the environment. Retrieving a memory is the brain's best evidence that it is useful, so it makes sense for retrieval to strengthen it more than passive re-exposure, which can happen by accident.

Storage versus access: Robert and Elizabeth Bjork's "new theory of disuse" (1992) separates how well something is stored from how easily it can be reached right now. Effortful retrieval, when access has weakened, produces bigger gains in storage.

Why rereading feels right: For most of history, familiarity was a decent guide to knowledge. A familiar face or plant had probably been met before. Text that can be reread at will, and exams that demand recall weeks later, are recent, and in that setting familiarity misleads.

These are plausible framings rather than tested evolutionary claims.

4. How This Bias Manifests

4.1. In Everyday Life

• Rereading a recipe card every time you cook the dish, and still needing it after the tenth time
• Highlighting a nonfiction book heavily and remembering almost nothing of it a month later
• Relying on GPS for a route you drive weekly and never learning it
• Recognizing words when reading a new language, then freezing when you have to speak
• Flipping flashcards over before really trying to answer

4.2. In the Workplace

• Onboarding built on slide decks and reading packets with no recall exercises
• Compliance quizzes whose answers sit on the previous screen
• Treating "I read the report" as equivalent to "I can explain the report"
• Training evaluations that ask how confident people feel, which rewards passive formats

4.3. In Business and Marketing

• Flashcard and language apps are built on retrieval, but only work if users answer before revealing the card
• Online courses sell hours of video because watching feels like learning and packages easily
• Advertising repetition builds recognition, enough to pick a brand off a shelf, but not unprompted recall of details

4.4. In Politics and Media

• Readers who scroll many headlines feel informed but cannot recall the main facts of a story a day later
• Debates over standardized testing treat tests only as accountability tools and ignore the learning value of low-stakes quizzes
• Retrieval can strengthen false claims too: repeatedly recalling a claim that was later corrected may help it stick, so corrections need to be practiced, not just stated

4.5. In Healthcare

• Residents forget much of what conferences teach; repeated testing with feedback gave markedly better six-month retention than review sheets (Larsen, Butler and Roediger, 2009)
• Medical students who self-test heavily with flashcards tend to score higher on licensing exams (Deng, Gluckstein and Larsen, 2015), though this correlation may partly reflect stronger students using the tools more
• Patients read discharge instructions once; the "teach-back" method, in which patients explain them in their own words, adds a retrieval step

4.6. In Finance and Investing

• Investors who read about fees and diversification, feel educated, and cannot explain either when deciding
• One-off financial-literacy seminars; a meta-analysis by Daniel Fernandes, John Lynch and Richard Netemeyer (2014) found such interventions had very small effects on behavior that decayed over time, and the absence of repeated retrieval is one plausible contributor
• Certification candidates who reread manuals instead of working practice questions

5. Real-World Case Studies

Case Study 1: Quizzing in Sixth-Grade Social Studies

• Context: Researchers from Washington University in St. Louis, among them Pooja Agarwal, Henry Roediger, Mark McDaniel and Kathleen McDermott, worked with teachers at a suburban middle school to see whether low-stakes quizzes would help students learn regular course content.
• What happened: Sixth graders took short multiple-choice quizzes on some material from each unit; other material was not quizzed or was simply reviewed again. The outcomes were the real chapter and semester exams the students were graded on (Roediger, Agarwal, McDaniel and McDermott, 2011).
• The bias at work: Review had meant going over material again. Teachers and students assumed more exposure was the way to prepare, and saw quizzes as assessment.
• Consequences: Students did better on quizzed material than on non-quizzed or merely reviewed material, on chapter exams and weeks later on semester exams, including exams that required written answers.
• Lessons learned: Short, frequent, ungraded retrieval inside ordinary lessons improves learning. The work became a model for later classroom programs, including Agarwal's book Powerful Teaching.

Case Study 2: What Residents Remembered Six Months Later

• Context: Douglas Larsen, Andrew Butler and Henry Roediger (2009) studied pediatric and emergency medicine residents at a teaching conference on status epilepticus and myasthenia gravis.
• What happened: Residents were randomly assigned to take short-answer tests, with feedback, on one topic and study a review sheet on the other. The sheets held exactly the information on the test answer keys. This happened after the session and twice more about two weeks apart. Around six months later, 40 residents took a final test on both topics.
• The bias at work: Medical education leans on lectures and handouts. Review sheets feel efficient and complete, and here they contained the same information as the tests.
• Consequences: Tested material scored 39 percent against 26 percent for studied material (d = 0.91). Both numbers are low, a sobering reminder of how much residents forget either way.
• Lessons learned: For knowledge clinicians need months later, a few spaced tests with feedback beat repeated review of identical content.

Historical Example: Herbert Spitzer's Iowa Study (1939)

Herbert Spitzer wanted to know how fast schoolchildren forget what they read and whether a test could slow it. He recruited 3,605 sixth graders across 91 Iowa schools. Each group read short factual articles and took tests on its own schedule, some immediately and others after a day, a week or longer.

Forgetting was fast. An early test changed the curve: groups tested soon after reading kept more on later tests than groups first tested late. Spitzer argued that immediate recall should be used as a teaching method.

Schools did not follow him. Through most of the twentieth century tests stayed at the end of units and review meant rereading. His study was rediscovered after 2006. The seven-decade gap between his conclusion and classroom practice is the neglect this article describes, at the scale of a school system.

6. The Cost of This Bias

6.1. Personal Costs

• Hours of rereading and highlighting that leave little lasting memory
• Exam results below the confidence built during study, often misread as low ability
• Relearning the same material every term because it was never consolidated

6.2. Professional Costs

• Training budgets spent on content delivery that leaves little behind
• Repeatedly looking up the same procedures, which slows work and raises error risk under pressure
• Failing certification exams after many hours of passive review

6.3. Societal Costs

• School review time spent on re-exposure instead of quizzing that could deliver measurable gains
• Gaps between students who learn effective strategies at home and those who default to rereading
• An evidence base drawn mostly from Western settings: only 6 percent of applied classroom experiments in one review were run in non-WEIRD countries (Agarwal, Nunes and Blunt, 2021)

6.4. Statistical Impact

• Classroom quizzing raised achievement by g = 0.499 across 222 studies and 48,478 students (Yang et al., 2021)
• 57 percent of effect sizes in a review of 50 classroom experiments (5,374 students) showed medium or large benefits (Agarwal, Nunes and Blunt, 2021)
• One week after study, testing beat rereading by about 56 to 42 percent of ideas recalled, while rereading was ahead at five minutes (Roediger and Karpicke, 2006)
• 84 percent of students reread as a strategy and 55 percent ranked it first; 1 percent ranked self-testing first (Karpicke, Butler and Roediger, 2009; N = 177)
• Of ten study techniques, only practice testing and distributed practice were rated "high utility"; rereading and highlighting were "low utility" (Dunlosky et al., 2013)

7. The Hidden Benefits

The pull toward rereading has some basis.

• Rereading helps comprehension: If you did not understand a passage, reading it again is right. Retrieval strengthens what you encoded; it cannot strengthen what you never grasped.

• Fluency is accurate in the short run: For a test five minutes away, rereading does produce higher scores. The feeling misleads only about the longer term.

• Rereading shows only correct answers: Practice tests without feedback, especially multiple-choice tests with plausible lures, can leave people remembering wrong answers (Roediger and Marsh, 2005).

• Testing too narrowly has a cost: Practicing some items can temporarily impair memory for related items that were not practiced (retrieval-induced forgetting).

The practical answer is a mix: read until you understand, then switch to recall with feedback.

8. Self-Assessment: Do You Have This Bias?

8.1. Warning Signs Checklist

• When studying, you mostly reread notes, textbooks or slides
• You highlight or underline as a main way of learning
• You feel ready for an exam because the material "looks familiar"
• You flip flashcards over before seriously trying to answer
• You save practice questions until you feel you "know it all," then run out of time
• You see quizzes as a way to check knowledge, not to build it
• Your exam scores are often lower than you expected
• You rewatch lectures instead of trying to explain them from memory
• Closing the book and trying to recall feels uncomfortable, so you rarely do it
• When you miss a practice question, you move on without checking the right answer

Scoring:

• 0-2 checked: Low susceptibility
• 3-5 checked: Moderate susceptibility
• 6-8 checked: High susceptibility
• 9-10 checked: Very high susceptibility

8.2. Self-Reflection Questions

1. In the last thing you studied seriously, what share of time went to reading and what share to recalling without looking?
2. When you feel you know something, is that feeling based on producing it from memory or on recognizing it?
3. What do you recall of a course or training you completed a year ago?
4. How do you react when you try to recall something and fail?
5. If you had to teach a topic you recently "learned" tomorrow with no notes, how much could you explain?

8.3. Quick Diagnostic Scenario

Scenario: You have a certification exam in three weeks, a 300-page study guide, a bank of practice questions, and about an hour a day.

How would you spend your study time?

• A) Read the guide carefully twice, highlighting key points, and save the practice questions for the last few days → High susceptibility
• B) Read the guide once, then mix rereading of hard chapters with practice questions once you feel confident → Moderate susceptibility
• C) Read each chapter once for understanding, then answer questions from memory, check answers, and return to earlier chapters every few days → Low susceptibility

9. Identifying This Bias in Others

9.1. Behavioral Indicators

• Heavily highlighted textbooks, many hours logged, disappointing grades
• Preparing for presentations by rereading slides rather than rehearsing without them
• Asking for the handout "to review later" instead of trying the exercises
• Recognizing the right answer in a list but unable to produce it unprompted
• Surprise when asked to explain something they "went through" last week

9.2. Conversational Red Flags

Phrases people say when under this bias:

• "I read it three times, so I should be fine."
• "I'll do practice questions once I've finished reviewing."
• "I know it, I just can't explain it right now."
• "It all looks familiar."

Types of arguments they make:

• Hours spent with the material count as proof of learning
• Tests are only for measuring, so testing before you are "ready" is pointless

Questions they avoid asking:

• "Could I write this down from memory right now?"
• "What did I get wrong last time, and why?"

9.3. Situational Triggers

• Easy-to-read, well-formatted material that boosts fluency
• Short delays between study and use, which hide how fast familiarity fades
• Learning settings with no built-in quizzes
• Performance anxiety that makes testing feel threatening
• Answers that are always visible (open tabs, searchable notes, AI assistants)

10. Cognitive Debiasing Strategies

10.1. Immediate Techniques

• Close and recall: After a section, close the material, write what you remember, then compare
• Answer first, then look: Commit to an answer, even a guess, before checking
• Always check: Look up every miss right away; retrieval without feedback helps only items you get right
• Predict, then test: Write the score you expect before a practice test and compare

10.2. Long-Term Strategies

• Space your retrieval: Return to the material after a day, a few days and a week, combining the testing effect with the Spacing Effect
• Practice cumulatively: Include earlier topics in each session
• Expect the struggle: Self-testing feels worse than rereading; per Bjork's "desirable difficulties," that discomfort often comes with better retention
• Prefer recall formats: Free recall and short answers tend to help more than recognition

10.3. Environmental Design

• Build in quizzes: Open lessons and training sessions with a two-minute recall prompt on earlier material
• Use spaced-repetition tools that require an answer before revealing it
• Keep stakes low: Ungraded or lightly graded quizzes give the benefit without heavy pressure
• Give feedback fast: Follow every quiz promptly with correct answers

10.4. When to Seek External Input

• When results keep falling below your confidence after studying
• When preparing for high-stakes certification and you cannot judge your readiness
• When anxiety stops you from testing yourself at all

Whom to ask: A tutor or study partner who controls the answers so you cannot peek. Ask them to quiz you without notice and log what you miss, rather than asking "Do I seem ready?"

11. Practical Exercises

Exercise 1: The Blank Page Test

• Objective: Replace passive review with free recall
• Time required: 15 minutes per session
• Materials needed: Study material, blank paper, a pen of a second color
• Difficulty level: Beginner
• Instructions:
  1. Read a section once, carefully, for understanding
  2. Put the material out of sight
  3. Write everything you remember: ideas, terms, examples, connections
  4. Compare with the source and mark misses in the second color
  5. Two days later, repeat before rereading anything
• Reflection questions:
  • How did your recall compare with how well you felt you knew the material?
  • Which parts did you miss both times?
• Frequency: After every new chapter or major topic

Exercise 2: The Prediction Experiment

• Objective: Test the fluency illusion on yourself
• Time required: 20 minutes on day one, 10 minutes a week later
• Materials needed: Two comparable sets of material (for example, two lists of 20 vocabulary words), a timer
• Difficulty level: Intermediate
• Instructions:
  1. Study set A by reading it four times
  2. Study set B by reading it once and testing yourself three times, checking after each attempt
  3. Spend equal time on each set
  4. Write how many items you expect to remember from each in a week
  5. A week later, test both sets without review and compare with your predictions
• Reflection questions:
  • Which set did you predict would win, and which did?
  • What will you change in how you study?
• Frequency: Once, then again with new material if the result surprised you

Exercise 3: Teach It Back

• Objective: Use explanation from memory as retrieval practice
• Time required: 20-30 minutes
• Materials needed: A listener or a voice recorder
• Difficulty level: Intermediate
• Instructions:
  1. Pick a topic you studied in the past week
  2. Without notes, explain it as if teaching a beginner
  3. Note every hesitation or skipped point
  4. Check the source only for those points
  5. Explain it again a few days later
• Reflection questions:
  • Where did your explanation break down?
  • What gaps did explaining reveal that rereading had hidden?
• Frequency: Weekly for important topics

Daily Practice

The Two-Minute Recall

At the end of each day, recall without notes three things you learned that day and write them down.

• Suggested duration: 2 minutes
• Best time of day: Evening
• How to track progress: Once a week, try to recall each logged item before reading it

Weekly Challenge

Swap One Review Session for a Self-Test

Each week, replace one rereading or rewatching session with practice questions, flashcards answered before flipping, or blank page recall, followed by checking answers.

• Expected outcomes after 4 weeks: Better recall of earlier material; a more accurate sense of what you know; less cramming
• Journaling prompts for reflection:
  • How did the self-test session feel compared with rereading?
  • What did it reveal that rereading would have hidden?

12. For Specific Audiences

For Leaders and Managers

• Training Design: Ask how much of a program requires recall without looking, alongside how much content it covers
• Meetings: Open recurring meetings with a quick recall of the last meeting's decisions
• Evaluation: Judge training by knowledge checks weeks later; satisfaction surveys on the day reward passive formats

For Parents and Educators

• Age-Appropriate Explanation: "Remembering without looking is like walking a path through the grass. Each time you walk it, the path gets clearer. Looking at the answer again is like reading a map without walking."
• Low-Stakes Quizzing: Use short ungraded quizzes, brain dumps or exit tickets regularly
• Feedback: Always follow quizzes with the correct answers, especially when many items are missed
• Teach Study Skills: Let students run Exercise 2 on themselves

For Healthcare Professionals

• Continuing Education: Prefer formats with question banks, case-based recall and spaced follow-up quizzes
• Training: Repeated short-answer tests with feedback beat review sheets for six-month retention in Larsen and colleagues' trial
• Patient Education: Use teach-back to check understanding and strengthen memory of instructions

For Financial Professionals

• Exam Preparation: Practice questions under exam conditions beat rereading the curriculum
• Client Education: Ask clients to summarize a plan in their own words before they leave
• Know the Limits: Recalling a rule is different from applying it well under market stress

13. Interactions with Other Biases

Biases That Amplify This One

Biases That Counteract This One

Common Bias Chains

Mere Exposure Effect → Illusion of Competence (Fluency Illusion) → Neglect of self-testing → Overconfidence Effect → Poor exam or job performance

Rereading makes material familiar. Familiarity is taken as knowledge. Feeling ready, the learner skips self-testing, so the gaps stay hidden and confidence stays high until recall is required.

Interrupting the chain: Break it at the second step. One closed-book recall attempt replaces the fluency signal with real data about what can be retrieved.

14. Cultural Perspectives

Recitation with correction runs through many older educational traditions. Quranic memorization (hifz) has students recite from memory to a teacher who corrects errors. Vedic oral traditions preserved texts for centuries through structured recitation. Jewish chavruta study pairs learners who read and explain texts to each other, and imperial examinations in China rewarded exact recall of the classics. None of these was designed around memory research, but they build in what the research supports: repeated attempts to produce material from memory, with feedback.

Much modern schooling moved to a lecture-and-reading model in which tests come at the end and serve to grade. The research suggests the trouble with testing is high stakes and infrequency, not retrieval itself. The evidence base is narrow, though: Agarwal, Nunes and Blunt (2021) found only 6 percent of applied classroom experiments came from non-WEIRD countries, and there is little direct research on whether the effect or the preference for rereading differs across cultures.

15. Myths and Misconceptions

16. Expert Insights

Arthur Gates found that schoolchildren who spent part of their study time reciting material from memory remembered more than children who spent all of it re-reading. — Summary of Arthur I. Gates, "Recitation as a Factor in Memorizing," Archives of Psychology, No. 40 (1917)

"A curious peculiarity of our memory is, that things are impressed better by active than by passive repetition." — William James, The Principles of Psychology (1890), vol. 1, ch. 16, "Memory"

Roediger and Karpicke review decades of research showing that taking a memory test does more than measure learning: it improves later retention, often more than additional study of the same material. — Summary of Roediger & Karpicke, "The Power of Testing Memory: Basic Research and Implications for Educational Practice," Perspectives on Psychological Science 1(3) (2006), pp. 181-210 https://doi.org/10.1111/j.1745-6916.2006.00012.x

Conditions that make learning feel slower and harder, such as testing yourself instead of rereading, often produce better long-term retention, while conditions that make learning feel easy can create an illusion of mastery. — Robert A. Bjork (paraphrasing his work on "desirable difficulties")

17. Key Takeaways

1. Retrieval strengthens memory: Trying to recall information improves long-term retention more than studying it again. Classroom meta-analyses find a medium-sized benefit.

2. Timing changes the winner: Rereading wins on a test minutes later. Testing wins at delays of days or weeks, which is when learning is usually needed.

3. The bias is in the choice: Most students reread and few self-test, because rereading feels fluent and recall feels effortful. This ties directly to the Illusion of Competence (Fluency Illusion).

4. Feedback matters: With feedback even failed attempts help. Without it, the benefit depends on how much you manage to recall, and very hard tests with no answers help least.

5. Low stakes are enough: Quizzes need not be graded to work. Short, frequent recall followed by the right answers is the core recipe.

6. Combine with spacing: Retrieval spread over days and weeks beats retrieval in one session, linking this effect with the Spacing Effect.

7. Know the limits: It is a memory and learning effect, strongest for material you already understand. It may shrink for highly complex material and can strengthen errors that are never corrected.

18. Further Resources

Academic Papers

• Gates, A. I. (1917). Recitation as a factor in memorizing. Archives of Psychology, 6(40).
• Spitzer, H. F. (1939). Studies in retention. Journal of Educational Psychology, 30(9), 641-656. https://doi.org/10.1037/h0063404
• Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255. https://doi.org/10.1111/j.1467-9280.2006.01693.x
• Karpicke, J. D., Butler, A. C., & Roediger, H. L. (2009). Metacognitive strategies in student learning: Do students practise retrieval when they study on their own? Memory, 17(4), 471-479. https://doi.org/10.1080/09658210802647009
• Kornell, N., Hays, M. J., & Bjork, R. A. (2009). Unsuccessful retrieval attempts enhance subsequent learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(4), 989-998. https://doi.org/10.1037/a0015729
• Larsen, D. P., Butler, A. C., & Roediger, H. L. (2009). Repeated testing improves long-term retention relative to repeated study: A randomised controlled trial. Medical Education, 43(12), 1174-1181. https://doi.org/10.1111/j.1365-2923.2009.03518.x
• Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772-775. https://doi.org/10.1126/science.1199327
• Roediger, H. L., Agarwal, P. K., McDaniel, M. A., & McDermott, K. B. (2011). Test-enhanced learning in the classroom: Long-term improvements from quizzing. Journal of Experimental Psychology: Applied, 17(4), 382-395. https://doi.org/10.1037/a0026252
• Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4-58. https://doi.org/10.1177/1529100612453266
• Rowland, C. A. (2014). The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin, 140(6), 1432-1463. https://doi.org/10.1037/a0037559
• Adesope, O. O., Trevisan, D. A., & Sundararajan, N. (2017). Rethinking the use of tests: A meta-analysis of practice testing. Review of Educational Research, 87(3), 659-701. https://doi.org/10.3102/0034654316689306
• Yang, C., Luo, L., Vadillo, M. A., Yu, R., & Shanks, D. R. (2021). Testing (quizzing) boosts classroom learning: A systematic and meta-analytic review. Psychological Bulletin, 147(4), 399-435. https://doi.org/10.1037/bul0000309

Books

• Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make It Stick: The Science of Successful Learning. Belknap Press of Harvard University Press. https://doi.org/10.4159/9780674419377
• Agarwal, P. K., & Bain, P. M. (2019). Powerful Teaching: Unleash the Science of Learning. Jossey-Bass. https://doi.org/10.1002/9781119549031

Book Chapters

• Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. P. Shimamura (Eds.), Metacognition: Knowing about Knowing (pp. 185-205). MIT Press. https://doi.org/10.7551/mitpress/4561.003.0011
• Roediger, H. L., Putnam, A. L., & Smith, M. A. (2011). Ten benefits of testing and their applications to educational practice. In J. P. Mestre & B. H. Ross (Eds.), Psychology of Learning and Motivation (Vol. 55, pp. 1-36). Academic Press. https://doi.org/10.1016/b978-0-12-387691-1.00001-6

19. Summary Card

20. Glossary of Terms Used

21. Discussion Questions

For book clubs, classrooms, or self-reflection:

1. Why does rereading feel more productive than self-testing, even for people who know the research?

2. If quizzes improve learning, why do most schools and workplaces use tests only to grade and certify?

3. With search engines and AI assistants giving instant answers, does retrieval practice matter more or less? What should we still keep in our own memory?

4. The effect may be weaker for highly complex material. How would you decide whether a subject is better learned through recall practice, worked examples, or a mix?

5. Traditions from Quranic recitation to Vedic study relied on reciting from memory with correction. What can modern education take from them?

Laas opgedateer 4 November 2026