What the research says

222 Classroom Studies on Quizzing, in One Review

A 2021 review in Psychological Bulletin pooled 222 studies and 48,478 students to ask how much quizzing helps in real classrooms, not labs. The answer was a medium-sized boost.

By the Giga Quizzes team4 min read
A wall of hundreds of books stacked on their sides, in every colour
Photo: Robert Anasch, Unsplash, Unsplash License, cropped.
In this post
  1. 1What a meta-analysis can tell you
  2. 2The headline result
  3. 3What made the effect bigger or smaller
  4. 4Why quizzing works, according to the review
  5. 5The limits
  6. 6Putting the factors to work

Laboratory studies of the testing effect are so consistent that they can start to sound too good. Students learn word lists in a quiet room, take a practice test, and remember more a week later. But a classroom is noisy, lessons are long, and the material is a real course rather than a list. In 2021 a team led by Chunliang Yang and David Shanks gathered every classroom study they could find to ask whether the effect survives the move. They ended up with 222 independent studies and data from 48,478 students.

What a meta-analysis can tell you

A single classroom study can be swayed by one good teacher, one keen class or one easy topic. A meta-analysis pools many studies and asks what is left once those quirks average out. Yang and his colleagues included studies that compared quizzing with some other way of spending the time, such as restudying the material, doing nothing extra, or using another strategy like concept mapping, and that measured learning on course material.

Each study gives an effect size: the difference between the quizzed and the non-quizzed students, measured in standard deviations, so that results from different tests can be put on one scale.

The headline result

Across the 222 studies, quizzing raised achievement by g = 0.499, about half a standard deviation. The authors call this a medium effect.

How to read g = 0.5What it means in practice
A class at the middle of the rangeWith quizzing, a typical student moves to around the 69th percentile of the unquizzed group
A test with a spread of about 15 marksRoughly seven or eight marks higher on average
Compared with most classroom changesLarger than many, and cheap to put in place
A rough guide to a half-standard-deviation effect; the conversions are standard approximations, not figures from the review

Half a standard deviation is not magic, and it is an average over very different studies. Some found much more, some found little. That spread is the most useful part of the review, because the authors went on to ask what made the difference.

What made the effect bigger or smaller

The abstract lists the factors that changed the size of the benefit. In plain terms, they are:

  • What the comparison group did. Quizzing looks bigger against doing nothing than against an active strategy.
  • Whether the quiz matched the final test in format, for example multiple choice before a multiple-choice exam.
  • Whether the final test asked about the same material that had been quizzed, or about related material.
  • Whether students got corrective feedback after the quiz.
  • How many times the material was quizzed.
  • Where and when the quiz happened: in class or at home, before or after the lesson.
  • How long the programme ran, and how the study was designed.

A teacher-facing summary of the review adds the directions: feedback helped, more frequent quizzing helped, quizzes in class worked despite taking lesson time, and quizzes after a lesson did more than quizzes before it. The same summary notes that the benefit held across subjects and from primary school to university.

Why quizzing works, according to the review

The authors found support for three explanations, and they are not rivals: each probably plays a part.

  1. Additional exposure. A quiz is another encounter with the material, and the answer revealed afterwards is a second look at it.
  2. Transfer-appropriate processing. Practice that resembles the final test prepares you for the final test. Recalling an answer is closer to what an exam asks than reading one.
  3. Motivation. Knowing a quiz is coming makes students prepare, pay attention and keep up.

The third explanation is easy to forget in laboratory research, where students are paid to sit still. In a real course it may be the largest effect of all. A class that is quizzed every week is a class that reads before every week.

The limits

A meta-analysis is only as good as its studies. Classroom research is harder to control than laboratory work, and many studies compare whole classes rather than randomly assigned students. Studies that find nothing are also less likely to be published. The authors examined these problems, but no statistical method removes them entirely. What the review does show is that the testing effect is not a laboratory curiosity: across hundreds of real courses, it holds.

Putting the factors to work

Each moderator in the review is a decision a teacher makes anyway, so it is worth making them deliberately.

A quiz routine built on the review

  • Quiz after teaching, not only before it, and do it more than once on the same material.
  • Always show the right answer afterwards, and say why it is right.
  • Use the question formats your final assessment will use.
  • Keep quizzes in lesson time when you can; homework quizzes are easier to skip.
  • Run it all term, not for one week before the exam.

None of this needs new technology or a new course. The daily quiz study from Texas is one way to put the whole list into practice, and the post on practice tests versus re-reading explains the laboratory work this review set out to check.

Sources

  1. 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.
  2. Barton, C. (Tips for Teachers) (2024). Research Bite #4: Testing (quizzing) boosts classroom learning. Tips for Teachers (summary of Yang et al., 2021).

More from the blog

All posts →