Pretesting Effect: Why Testing Before You Study Works

The pretesting effect is the finding that trying to answer questions before you study something improves how well you remember it, even when your guesses are wrong. In one set of experiments, students who spent two minutes on five questions about a passage they hadn't read, then read it, beat students who spent the whole time reading.

It sounds backwards. You can't know the answers yet, so why ask? Because a failed attempt changes how you read: you notice the answer when it appears, and it sticks.

This post explains what the research shows, where pretesting works and where it doesn't, why it feels like a waste of time, and a 10-minute routine for using it on your next chapter or lecture.

What is the pretesting effect?

Researchers use several names for the same idea: pretesting, prequestions, and errorful generation. All of them mean being tested on information before you've had a chance to learn it.

Two classic studies show how it works:

• Text passages. In a 2009 study by Richland, Kornell, and Kao, participants read a two-page essay about a vision disorder. Some first spent two minutes on five questions such as "What is total color blindness caused by brain damage called?", then read for eight minutes. Others read for the full ten. Most pretested participants got three-quarters or more of the questions wrong. They still did better on a later test, in all five experiments, even counting only the questions they had missed.
• Facts. In a 2009 study by Kornell, Hays, and Bjork, participants were asked made-up trivia questions like "What peace treaty ended the Calumet War?", which nobody could answer. Trying to answer first, then seeing the answer, led to better memory than reading the question and answer together.

One detail from the Richland study matters for you: in one experiment, simply showing participants the questions before reading, without asking them to answer, did not produce the same benefit. Committing to a guess is part of what makes it work.

What does the research on prequestions show?

The effect has been found with text passages, trivia, vocabulary, and video lectures, and across test formats. A few studies give a sense of its range:

That last row is the most important practical finding. Pooling the research, the 2024 meta-analysis found a clear effect for material the prequestions asked about and essentially no effect on other material in the same lesson.

So write prequestions about the ideas you most need to learn. A pretest on trivia from the chapter will make you remember the trivia.

Why does guessing wrong help you learn?

Researchers are still testing explanations, but three ideas have the most support:

1. It directs your attention. Once you've been asked a question, you read looking for its answer. The 2024 meta-analysis found support for this attention-based account, which also explains why the benefit stays with the questioned material.
2. It gives you a reality check. A pretest shows you, in under a minute, what you don't know, which sends you looking for the answers. Carpenter and Toftness called this a metacognitive "reality check". In eye-tracking studies, pretested readers spent more time on the sentences that answered their questions.
3. It activates related knowledge. With word pairs, guessing helps mainly when the cue and answer are related, such as guessing "mammal" for "whale". Your guess wakes up nearby knowledge that the correct answer then attaches to. For unrelated pairs, the benefit often disappears.

The same logic is behind the generation effect and the idea of learning from mistakes: producing an answer, even a wrong one, makes the correct answer more memorable once you see it.

Does pretesting work in real classrooms?

Sometimes. Classroom evidence is thinner and more mixed than lab evidence:

• It helped: in a 10-week psychology research-methods course, students took multiple-choice pretests at the start of several lectures. On the final exam they did 8–9% better on questions about pretested content, including related questions they hadn't seen before (Soderstrom & Bjork, 2023, as summarized by Pan and Carpenter).
• It helped briefly: in an introductory psychology course, students who answered a prequestion before a lesson remembered that content better at the end of class, but not the rest of the lesson. Everyone also took a short quiz at the end of class; a week later, both groups remembered the quizzed material better, and the prequestion group had no clear extra edge (Carpenter, Rahman & Perkins, 2018).
• It didn't help: in a chemical engineering course, prequestions did not improve learning or add to the benefit of end-of-class quizzes (Geller et al., 2017).
• It had a side effect: in an online course on climate change, students given pretests were more likely to drop out, though those who stayed did better on exams (Janelli & Lipnevich, 2021).

The pattern suggests pretesting works best when the questions target the key content, the answers are stated clearly in the lesson, and you find out the correct answers soon after.

Pretesting vs testing after you study

Which helps more: a quiz before you read, or after? Results are split. Pan and Sana's five experiments found pretests produced higher scores than tests taken after reading. Two other studies, one online and one with elementary school students, found tests after reading did better.

You don't need to pick one. They seem to do different jobs: a pretest prepares you to notice and encode new information, and recall afterward strengthens what you learned. Use both. For the research on testing after study, see retrieval practice and the testing effect.

When pretesting doesn't help

Pretesting has clear limits:

• No answers, no benefit. The effect depends on studying the correct answers afterward, through reading, a lecture, or feedback.
• Complex inference questions. Prequestions that require drawing conclusions not stated anywhere in the material tend not to help. Pretesting works best for facts and concepts stated clearly in one place.
• Unrelated guesses. If there is no meaningful link between the cue and the answer, as with random word pairs, guessing adds little.
• Material you can't understand yet. If the lesson itself is too hard to follow, you won't be able to connect the answers back to the questions.

Why pretesting feels like a waste of time

Most people underrate pretesting, even after it has worked for them. Asked how best to learn a list of word pairs, 78% of respondents in one survey rated plain reading as more effective than guessing first. In another study, participants ranked pretesting as their least effective technique even though it had produced their best memory.

Pan and Rivers found in 2023 that learners kept believing reading worked better even after repeatedly trying both methods. What changed their minds was seeing their own scores side by side. So try it on one chapter and compare: count how many pretested questions you get right a week later versus questions on material you only read. Our post on desirable difficulties explains why strategies that feel worse often work better.

A one-week test you can run on yourself

You need eight questions you haven't seen yet that are answered in one chapter: end-of-chapter review questions, a course question bank, or last year's quiz all work.

1. Before reading, pick four of the questions at random and write a guess for each. Don't read the other four.
2. Read the chapter as you normally would, then correct your four guesses.
3. A week later, answer all eight from memory without opening the book.
4. Compare your score on the four you pretested with your score on the four you didn't.

Every answer was in the same reading, so the only difference between the two sets is the pretest. Eight questions is a small sample, so run it on two or three chapters before you decide. If the pretested questions keep coming out ahead, you have your own evidence to keep the routine below, which is the kind of side-by-side result that changed minds in Pan and Rivers's experiments.

How to pretest yourself: a 10-minute routine

Use this before a textbook chapter or lecture. It adds about 10 minutes to your normal study time.

1. Write 5–8 questions (3 minutes). Pull them from the headings, learning objectives, end-of-chapter questions, last year's quiz, or the lecture slides if they're posted early. Aim for the key ideas, since the benefit stays with what you ask.
2. Guess (2 minutes). Give each question a written answer, even a wild one. Spend seconds, not minutes.
3. Study with the questions in view. Read or watch as usual. When you hit an answer, mark it.
4. Correct (2 minutes). Write the correct answer next to each guess in a different color.
5. Recall afterward (3 minutes). Close the book and answer the same questions from memory, plus one or two new ones.

Here's what it looks like for an intro psychology lecture on classical conditioning:

Getting all three wrong is normal and fine. You'll pay attention when the lecture reaches each one, and correcting your own wrong guess tends to make the right answer more memorable.

If your course has no question bank, an AI quiz generator can draft multiple-choice questions from a chapter PDF. Check them against the chapter, and keep the ones about the main ideas. For writing your own, see how to create effective quizzes. Then add the questions to your regular active recall reviews so the pretest becomes the first of several retrievals.