← All articles
Paivio blog

The Testing Effect: Why Self-Quizzing Beats Rereading

The Testing Effect: Why Self-Quizzing Beats Rereading

Student writing remembered material beside a closed textbook

The testing effect is the finding that trying to recall information from memory helps you remember it later better than studying it again does. The test is not just a way to measure what you know. It is one of the most reliable ways to learn it. Researchers also call it retrieval practice, practice testing, or test-enhanced learning, and it shows up across ages, materials and classrooms. Below you will find what the effect is, why it works, where it stops working, and how to build it into your study week.

Paiviopaivio.aiMake Retrieval Practice Part of Your RoutinePaivio turns your Anki decks, notes, or topics into personalized spaced reviews that help you practise recalling what you study.Explore Paivio

TL;DR

  • The testing effect is better long-term memory after you spend part of your study time retrieving information instead of rereading it.
  • A large meta-analysis found practice tests more beneficial than restudying and other comparison conditions, across many kinds of learners.
  • It works best when recall takes real effort, succeeds most of the time, and is followed by corrective feedback.
  • It weakens when your first attempts mostly fail with no feedback, or when material has many tightly interacting parts.
  • Spacing your retrieval sessions and tying them to your exam date turns a good technique into a dependable system.

Table of Contents

What the testing effect actually is

In the research literature, the testing effect is improved later performance after adding test or retrieval trials in place of additional study trials. Picture two students with the same chapter and the same 30 minutes. One reads it twice. The other reads it once, closes the book, and writes down everything she can remember, then checks the text. Weeks later, the second student typically remembers more, even though she spent less time looking at the material.

That result surprises people because rereading feels productive. The text looks familiar, so you assume you know it. Familiarity is not recall, though, and exams ask for recall. Retrieval practice trains the exact skill the exam measures.

A few terms get used interchangeably, so it helps to separate them:

  • Testing effect: the benefit itself, measured by comparing retrieval against restudy.
  • Retrieval practice: the activity you do, such as flashcards, free recall or practice questions.
  • Active recall: the everyday study-advice name for the same idea.

None of these require a formal, graded test. A blank sheet of paper, a flashcard or a question you ask yourself all count, as long as you try to produce the answer before you look at it.

Why retrieval strengthens memory

The short version: pulling something out of memory changes the memory. It does not just read it.

Karpicke’s 2025 review of retrieval-based learning explains that retrieving requires learners to retrieve, reinstate and update context representations. Each successful recall rebuilds the memory in the context of what you know now, which makes it easier to find next time.

Direct and mediated effects

The same author’s earlier review separates two routes. Direct effects are the strengthening that comes from the act of recall itself. Mediated effects happen when testing changes what you do next: it shows you what you do not know, guides where you spend study time, and can raise motivation.

Both matter for a student. A failed recall attempt on Monday is a free diagnostic. It tells you which topic to reopen on Tuesday, which rereading never does.

Testing can help what you learn next

There is also a forward benefit. A 2018 meta-analysis describes how testing can potentiate, or enhance, the learning of new material presented afterward, and it found considerable support for an integration account of that effect. An earlier paper on the forward testing effect notes that it has been observed for both accurate information and misinformation, so the benefit concerns how well you take in new content, not whether the content is true. In practice: quiz yourself on what you studied this morning before starting the next topic.

What the research shows

The testing effect is one of the better-replicated findings in learning science. A review in the literature describes it as distinguished by its robust effect size and near-ubiquitous occurrence across diverse learning situations and a broad range of parameters.

The most useful single study for a student is the 2017 meta-analysis by Adesope and colleagues. It found practice tests more beneficial for learning than restudying and other comparison conditions. The average effect varied with test features, participant and study characteristics, outcomes, and methodology. The same paper reported effects across education levels: secondary students showed g = 0.83, primary students g = 0.64, and postsecondary students g = 0.60. Even the smallest of those is a substantial effect by the usual standards for educational interventions, and college students sit right in it.

One point that matters for serious coursework: retrieval practice is not limited to word lists. Karpicke’s 2025 review states that it has been generalized to improve learning of complex educational materials, not just simple facts. McDermott’s 2021 Annual Review reaches a similar conclusion: practicing retrieval of newly learned information makes it more likely you can retrieve it again later, across materials and learner ages.

What the research shows

What makes the effect bigger or smaller

Not all retrieval is equal. Five factors decide how much you get out of it.

1. Effort

A 2014 meta-analytic review by Rowland suggests that effortful processing during testing is a key contributor to the effect, and that initial recall tests often produce larger benefits than recognition tests. Writing or saying an answer from scratch beats picking it from a list. If you can only recognize the right answer when you see it, you have not yet practiced what an exam with short answers will demand.

2. Success

Karpicke’s review notes that retrieval-based learning works best when learners can successfully retrieve knowledge with relatively less retrieval support. The sweet spot is a question hard enough to make you work and easy enough that you usually get it. If every card feels impossible, you are testing too early or too much at once. Go back, study the material, then retrieve.

3. Feedback

When retrieval fails and nothing corrects it, errors can persist. The practical rule is to check every answer against your source right after you try. A review of the field points out that effortful recall with appropriate corrective feedback is generally the safer approach. Low-stakes quizzes are useful partly because they give feedback on strengths and weaknesses.

4. Format match

Benefits can shrink when the final test uses a different format from the practice. If your exam is clinical vignettes and you practice only isolated definitions, you are covering half the job. Mix formats as the exam date approaches.

5. Spacing

The studies above compare retrieval against restudy, and retrieval spread over days and weeks tends to hold up better than retrieval crammed into one night. We cover interval choices in our guide to a spaced repetition schedule for exams.

Where the testing effect falls short

A good study technique has boundaries, and knowing them keeps you from misapplying it.

When your starting knowledge is very low. If you retrieve unsuccessfully and get no feedback, errors may persist. When initial performance is very low, practice tests can even produce a negative testing effect. Learn the material first, then quiz.

When material is highly complex. Van Gog and Sweller’s 2015 article in Educational Psychology Review identifies material complexity as a boundary condition: the testing effect decreases and may disappear as the number of interacting information elements increases, because restudying can better support relational or organizational processing. Think of a physiology pathway where every step depends on five others. Isolated flashcards on that pathway may help less than rereading a well-organized explanation or drawing the whole system.

The lesson is not to drop retrieval. It is to match the retrieval to the structure of the content. For interconnected topics, retrieve at the level of relationships: “Explain how a drop in cardiac output triggers the renin-angiotensin response” instead of “Define renin.” Then compare your explanation with the source.

When the practice is recognition only. Multiple-choice questions have a place, particularly late in preparation, but they should not be your only mode.

The field’s own summary is that the effect is robust but not universal: test format, retrieval success, feedback and material complexity determine whether practice testing beats restudy.

Where the testing effect falls short

How to use the testing effect: a weekly routine

The architecture that works is simple. Karpicke’s review describes it as initial study, an activity that requires retrieving the information, and a final assessment. Here is how to turn that into habits.

  1. Study once, with attention. Read or watch the lecture, and mark the ideas that would make good questions.
  2. Close the source and recall. Write a free-recall summary, or answer prompts from memory. Do it before you feel ready; mild difficulty is the point.
  3. Check and correct. Compare against your notes. Mark what you missed or got wrong.
  4. Turn misses into prompts. Every gap becomes a question on a card, in your own words, so you can retrieve it again.
  5. Return on a schedule. Revisit cards at growing intervals. Cards you recall easily go further out. Cards you miss come back sooner.
  6. Mix formats near the exam. Add practice questions and written explanations to match what the exam will ask.
  7. Tie it to the calendar. Work backward from your exam date so the number of daily reviews stays manageable. Our exam study schedule guide walks through that.

Your own cards or premade ones?

The act of writing a question is itself a form of processing, and cards in your own wording tend to match how your course frames the material. Premade decks save time but can drift from your syllabus. We compare the trade-offs in custom flashcards vs premade.

A note on AI

AI tools can draft questions quickly, but if a tool hands you the answers you will skip the effortful part that makes the effect work. Use it to generate prompts, then attempt them yourself first. More on that in using AI for studying without weakening learning.

A worked example for a biomedical course

Suppose you have a week on heart failure in pathophysiology.

  • Day 1: Read the lecture and your textbook section once. Write 15 to 20 prompts, a mix of definitions and “why” questions (why does preload rise in left-sided failure?).
  • Day 1, evening: Answer them from memory, check, and mark misses.
  • Day 2: Retrieve the misses first, then the rest. Draw the compensation cascade on a blank page from memory and compare it with the lecture slide.
  • Day 4: Retrieve again. Add three or four case-style questions that link heart failure to renal and endocrine responses.
  • Day 8: Retrieve once more, mixing in older topics you covered the previous week.

Total daily effort is small, and each topic is touched four or five times. For a longer semester plan, see our 12-week pathophysiology study plan.

Our perspective

We built Paivio around one belief: the best study material is your own, and the best way to learn it is to keep retrieving it. Most students already have notes, lecture slides or Anki decks. The problem is rarely a lack of material. It is turning that material into a loop of studying, recalling, rating and returning, and keeping the loop going for months.

The research above is why we care about the details. Retrieval should take effort, so we ask you to rate how well you recalled each item, and that explicit rating drives when it comes back. Retrieval should come with feedback, so you check your answer against your source. And because complexity can blunt the effect, we added concept mapping, so you can retrieve relationships between ideas and not only isolated facts.

We would also add a caution that no app can replace. A tool can schedule your reviews, but it cannot make you try before you peek. The testing effect comes from the attempt. If you flip cards without guessing, you are rereading with extra steps. Guess first, every time, even when you think you will be wrong.

— Jackson

Put the testing effect to work with Paivio

Paivio lets you import your own Anki decks, notes or topics and builds a review system around them, with spaced repetition scheduled by your recall ratings, daily review queues, exam-date pacing and concept maps that show how ideas connect. You can bring in Anki decks with their review history, so you do not start over.

Paivio

The Free plan costs $0 per month, and the beta currently gives unlimited cards. Plus is $20 per month or $200 per year, and Pro is $40 per month or $400 per year, with higher credit and AI usage limits. Credit top-ups are $10 one-off. See plans and pricing, or look at how Paivio works before you decide.

FAQ

What is the testing effect in simple terms?

It is the finding that you remember information better after trying to recall it than after studying it again. The benefit holds over the long term, which is why it matters for cumulative exams. It is also called retrieval practice or test-enhanced learning.

Is the retrieval practice effect the same as the testing effect?

For practical purposes, yes. Researchers use both terms, along with practice testing, for the same benefit: long-term memory improves when part of the learning period is spent retrieving. “Testing effect” names the result, while “retrieval practice” usually names the study method.

Does the testing effect work for complex subjects like medicine or nursing?

Often yes, with care. A 2025 review reports that retrieval practice has been generalized to complex educational materials. But research on material complexity suggests the effect can shrink when many elements interact. For those topics, retrieve explanations and relationships, and check them against a well-organized source.

How often should I test myself?

Spread it out rather than doing it all at once. A common pattern is a first attempt the same day, then returns after a day or two, then a week, then longer gaps, adjusting by how easily you recall. Our spaced repetition schedule guide gives specific intervals for exam prep.

Can testing ever hurt learning?

It can when initial performance is very low and you get no feedback, since mistakes may stick. Study first, test after, and always check your answers. That keeps retrieval on the productive side.

Sources

Recommended

Your next chapter

Bring one deck.
Start one session.

Create an account and make your next review a place to begin.

Start studying free

Free during beta · No credit card · For students 18+