Why does pulling information out of memory beat putting it back in?
Retrieval practice means actively trying to recall something—answering a question, writing down what you remember, or self-testing—rather than rereading or reviewing the material. Restudying is the passive alternative: going over the same content again. In a randomized trial with 119 participants, those who repeatedly retrieved a list of related words outperformed those who repeatedly restudied it on a free-recall test two days later, regardless of whether they received brain stimulation [1]. The retrieval group also showed better organization of the recalled items, which the authors suggest may be one route by which retrieval practice helps [1].
The benefit shows up outside the lab too. In a study with 177 Saudi undergraduate English-language learners, retrieval practice without feedback beat repeated studying on a test one month later, and adding feedback made the advantage even larger [4]. That one-month gap matters: it is long enough that the result reflects retention rather than the temporary boost that comes from having just seen the material.
Why does restudying feel more confidence-inspiring?
Restudying feels smoother because the material is right in front of you, so it is easy to mistake familiarity for mastery. Retrieval, by contrast, feels effortful—and that effort can be misread as a sign that learning is not going well. In one study, participants rated how much mental effort each item required and how well they expected to remember it, then chose whether to restudy or self-test [5]. Difficult items produced higher effort ratings, which lowered judgments of learning and made participants less likely to choose retrieval practice [5]. In other words, the very struggle that makes retrieval effective also makes it feel less appealing.
This is a case where the subjective signal points the wrong way. A study that tracked participants' predictions about whether they would remember facts found that restudying after a test improved the accuracy of those predictions [2]. That suggests restudying can make people's sense of their own learning better calibrated in some conditions—but feeling well calibrated is not the same as remembering more.
Is retrieval always better, or does it depend on how the material is studied?
The advantage is robust but not unconditional. When facts were encountered only briefly—three total exposures, akin to a single mention in a lecture—long-term memory was best when restudy was combined with self-testing rather than when restudy was replaced entirely by testing [2]. In that study, a study-study-test sequence beat both study-test-test and study-study-study after delays of two and seven days [2]. So the practical lesson is not "never restudy" but "test yourself and also revisit the material," especially when exposure is limited.
Timing of feedback also matters for retrieval practice. Immediate feedback after a retrieval attempt was more effective than delayed feedback for one-month retention in the EFL study [4]. And the tendency to avoid retrieval practice on difficult material can be shifted: when students received an instructional intervention explaining the strategy, they chose retrieval practice more often for both easy and difficult items, even though their initial preference had leaned toward restudying the hard ones [5].
What should a learner take away from the confidence gap?
Treat the feeling of fluency during restudying as a warning sign rather than a verdict. The studies that measured confidence or judgments of learning consistently find that subjective ease does not track long-term retention well [2][5]. If rereading feels effortless and self-testing feels clumsy, that is roughly what the evidence would predict—and the clumsy-feeling strategy is the one more likely to hold up after a delay [1][4].
A reasonable approach is to test yourself first, then restudy what you missed, rather than choosing one strategy exclusively. That combination performed best for briefly encountered facts [2], and retrieval practice on its own has repeatedly outperformed restudying at delays of days to a month [1][4]. The confidence boost from restudying is real, but it is a feeling about the moment, not a forecast of what you will remember later.
Sources used in this answer
Effects of combining retrieval practice and tDCS over long-term memory: A randomized controlled trial
In a randomized trial with 119 participants, retrieval practice outperformed restudying on a two-day delayed free-recall test regardless of tDCS condition, and recall organization was higher in the retrieval group.
A combination of restudy and retrieval practice maximizes retention of briefly encountered facts
For briefly encountered facts, a study-study-test sequence produced better long-term retention than study-test-test or study-study-study, and restudying after a test improved the accuracy of participants' predictions of remembering.
Test-Enhanced Learning: Using Retrieval Practice via Testing to Enhance Long-Term
The paper examines retrieval practice versus restudying for long-term retention and includes measures of confidence in answers, but the available abstract does not report a specific confidence comparison result.
Timing of feedback and retrieval practice: a laboratory study with EFL students
Among 177 EFL students, retrieval practice without feedback outperformed repeated studying on a one-month delayed test, and immediate feedback was more effective than delayed feedback.
Improving the Use of Retrieval Practice for Both Easy and Difficult Materials: The Effect of an Instructional Intervention
Students chose retrieval practice less often for difficult items because higher perceived effort lowered judgments of learning, but an instructional intervention increased their choice of retrieval practice for both easy and difficult materials.