How does immediate feedback or error checking after retrieval change the retention benefit of spaced practice?

Why a quick self-check can flip the usual retrieval advantage

Spaced practice works best when learners actively retrieve information rather than just reread it. But what happens right after a retrieval attempt matters. In three online experiments, people studied Swahili–English word pairs, then either restudied the pairs or tried to recall the English word from the Swahili cue across several spaced sessions, with a final cued-recall test two days after each session [1].

In the first two experiments, every trial required a memory rating: after retrieval, learners judged whether their answer was correct; after restudy, they judged whether they would have remembered the translation. Under those conditions, restudied items were remembered better than retrieved items, the opposite of the usual retrieval advantage [1]. In the third experiment, the memory rating was removed entirely, and retrieval practice then outperformed restudying at every amount of practice [1]. The authors discuss this reversal in terms of the reactivity of memory judgments, meaning that the act of judging your own memory may itself change how you process the material [1].

This is a striking result, but it comes from one set of word-pair experiments, so it should be read as a specific finding about that task and that kind of self-rating rather than a universal rule about all feedback.

What immediate feedback adds: extra practice and fewer repeated errors

A different line of work looks at feedback timing rather than self-ratings. Learning theory predicts that immediate feedback should improve retention, and one explanation is that immediate feedback provides additional distributed practice, which in turn helps learners keep producing correct answers instead of repeating mistakes [3]. In other words, feedback may work partly as another spaced learning event, not just as a correction signal.

This fits with the idea that the benefit of feedback depends on what it makes the learner do next. If feedback prompts another round of retrieval or restudy, it can add spacing. If it mainly prompts a quick self-judgment, as in the word-pair experiments, it may interfere with the retrieval advantage [1][3].

When feedback is built into real spaced training

In applied settings, immediate feedback is often bundled into spaced training programs. A randomized controlled trial with surgical residents used online spaced education in which learners received immediate feedback and educational material after each answer, and compared it with massed practice concentrated into a single day [2]. The spaced format improved long-term retention of diagnostic skills and produced transfer, meaning residents could apply what they learned beyond the exact training items [2].

That trial shows that spaced training with immediate feedback can work well in a real educational setting, but it did not directly compare spaced practice with and without feedback [2]. So it supports the value of the combined package rather than isolating feedback's separate contribution. Similarly, a microlearning approach in higher education combined retrieval practice, spaced repetition, and immediate feedback and found better retention than conventional instruction, again without a clean feedback-versus-no-feedback comparison [4]. An AI-powered spaced-repetition system likewise pairs retrieval practice with immediate feedback and reports retention benefits compared with unstructured review, but does not isolate the feedback effect [5].

How to think about feedback in your own spaced practice

The practical takeaway is that feedback after retrieval is not a single ingredient with a fixed effect. A quick correctness judgment inserted after every attempt may change how learners process the material and, in at least one word-pair paradigm, can erase or reverse the retrieval advantage [1]. Feedback that supplies the correct answer and prompts another spaced encounter with the material is more consistent with the idea that feedback helps by adding distributed practice and reducing repeated errors [3].

In real training programs, immediate feedback is often part of a package that also includes spacing and retrieval, and those packages have produced strong long-term retention and transfer in at least one randomized trial with surgical residents [2]. What remains less clear is how much of that benefit comes from the feedback itself versus the spacing and retrieval around it, because the applied studies did not run that direct comparison [2][4][5].

Sources used in this answer

Related research questions