Newsroom Search Newsroom home Newsroom Archive
Purdue News

* Memory and Cognition Lab

December 10, 2009

Student self-testing earns high marks as study tool

WEST LAFAYETTE, Ind. - College students who pore over their notes again and again as they prep for finals could use their studying time more wisely, according to new learning research from Purdue University.

"We know that self-testing, which happens when students practice retrieving knowledge, drives learning," says Jeffrey D. Karpicke, an assistant professor of psychological sciences. "Students can really benefit from testing themselves as they study by using something as simple as flashcards. However, the key is to not drop a flashcard once you feel you have mastered the material. Keep it as part of your rotation and keep practicing retrieval of that information."

Karpicke found in his study that college students are more likely to invest their time in repetitive note reading, and those who do practice retrieval spend too little time on it.

"My research found that this happens because there is an illusion about how much a person is actually learning while they are self-testing," said Karpicke, who is a cognitive psychologist and memory expert.

The illusion takes root when students feel answers come to them easily as they practice testing. For example, students using flashcards to study may eliminate certain cards when they believe they know that material well.

"This is called retrieval fluency," he said. "If you practiced recalling information even a few more times, it would produce big gains in learning and long-term retention. The reason people don't keep testing themselves is because they are tricked by retrieval fluency. The answer comes to mind so easily the first time that they think they know it and drop the card from further self-testing. But this is not a recipe for good long-term learning."

The research findings appeared in last month's Journal of Experimental Psychology: General. Karpicke conducted four experiments with 150 college students in various studying situations on Swahili-English vocabulary words. Students in each experiment learned vocabulary words from a computerized flashcard format, and then the conditions were varied based on studying techniques assigned by Karpicke or selected by the participant. The students returned a week later for final testing. 

No matter if the students selected their own studying strategy or it was assigned, they all learned better when self-testing all of the material from the electronic flashcard format. Students didn't do as well on the final test if they dropped material as they learned it during self-testing.

The students whose studying techniques were assigned received computer prompts on what to study and even how to study it at times. Some of the students in the experiment could select how they wanted to study, and they were likely to drop the vocabulary words they felt they knew well. As a result, many could not remember the words when they returned a week later for the final test.

"What is surprising is that we know practicing retrieval by self-testing is really powerful, and yet people don't use it, or don't use it well," he said. "These are college students who are generally successful academically, so this just shows how powerful the illusion can be."

Karpicke's work was funded by a dissertation research award from the American Psychological Association and a graduate research scholarship from the American Psychological Foundation/Council of Graduate Departments of Psychology. He will continue studying learning techniques, and his next project will focus on evaluating how course instruction relates to the studying strategies students select.

Writer: Amy Patterson Neubert, 765-494-9723,

Source: Jeffrey D. Karpicke, 765-494-0273,

Purdue News Service: (765) 494-2096;

Note to Journalists: Journalists interested in a copy of the journal article can contact Amy Patterson Neubert, Purdue News Service, at 765-494-9723,


Metacognitive Control and Strategy Selection:
Deciding to Practice Retrieval During Learning

Jeffrey D. Karpicke

Retrieval practice is a potent technique for enhancing learning, but how often do students practice retrieval when they regulate their own learning? In 4 experiments, the subjects learned foreign-language items across multiple study and test periods. When items were assigned to be repeatedly tested, repeatedly studied, or removed after they were recalled, repeated retrieval produced powerful effects on learning and retention. However, when subjects were given control over their own learning and could choose to test, study, or remove items, many subjects chose to remove items rather than practice retrieval, leading to poor retention. In addition, when tests were inserted in the learning phase, attempting retrieval improved learning by enhancing subsequent encoding during study. But when students were given control over their learning they did not attempt retrieval as early or as often as they should to promote the best learning. The experiments identify a compelling metacognitive illusion that occurs during self-regulated learning. Once students can recall an item they tend to believe they have "learned" it. This leads students to terminate practice rather than practice retrieval, a strategy choice that ultimately results in poor retention.


To the News Service home page

If you have trouble accessing this page because of a disability, please contact Purdue News Service at