Active recall, why quizzing yourself beats rereading

Active recall, also called the testing effect, means pulling information out of memory without your notes in front of you. You answer a question, complete a fill in the blank sentence, write down what you know about a chapter, and then check. The word test does not mean a graded exam here. A test is any situation where you produce the answer yourself.

In the review by Dunlosky and colleagues, practice testing is one of the two techniques with a high utility rating, along with spreading your study out over time (Dunlosky et al., 2013). For finals, AP exams or the SAT, this comes down to simple habits. You close your notebook and write the three causes of the Great Depression. You recite the product rule for derivatives. You redraw the diagram of mitosis without a model. The goal is not to get it right on the first try, it is to make the effort of recalling.

Experiment 2 by Roediger and Karpicke involved 120 students learning prose texts. One group read once and then quizzed themselves three times. The other reread four times. A week later, the first group remembered 61% of the ideas and the second 40% (Roediger and Karpicke, 2006).

Between the 5 minute test and the one a week later, the rereaders forgot 52% of what they knew, against 14% for those who quizzed themselves (Roediger and Karpicke, 2006). The table puts both results side by side. Read it with one important caveat. These figures come from short texts read in a lab, with a test on remembering the ideas. They do not say that a semester of law or chemistry is retained at 61%. They show a gap, measured in a specific setting, between two ways of reviewing the same material.

One detail of the same study explains why so many students reread. At 5 minutes, rereading gave a better score. And the students who reread were the most confident they would remember (Roediger and Karpicke, 2006). In the moment, rereading feels reassuring, and a week later the picture is the opposite.

This confusion is not specific to one text. With Swahili vocabulary, once a word was known, studying it again had no effect on delayed recall, while testing it again had a large positive effect. The students’ predictions of their own success were not related to how well they actually did (Karpicke and Roediger, 2008). Bjork and colleagues describe these illusions of mastery and suggest looking for desirable difficulties, situations a bit harder that force effort (Bjork et al., 2013). If a study session feels easy and smooth, it may be because it does not make you recall much.

A 2014 meta-analysis found that recall tests give a bigger benefit than recognition tests, such as multiple choice (Rowland, 2014). When you can, produce the answer instead of recognizing it. Write the definition, say the date, complete the fill in the blank sentence, instead of only picking among options.

Another result worth knowing is that active recall produced more learning than elaborate study with concept maps (Karpicke and Blunt, 2011). Making a beautiful concept map of your course pays off less than asking yourself questions on it. Keep a simple order. Understand the course first, then quiz yourself without looking, correct yourself, and repeat on different days. Spacing is covered in Spaced repetition, how long to wait between two reviews. The page How to study effectively, according to the research puts all this back into the context of the ten techniques.