Shows that retrieval through testing can improve later retention more than repeated study, especially over longer delays.
This paper is central to research on the testing effect: retrieving information can strengthen later retention rather than merely measure what is already known. That distinction changes the role of a quiz. A well-designed recall attempt can be part of instruction, not just a judgment delivered after instruction.
Students studied educationally relevant prose passages and then either restudied the material or completed immediate free-recall tests without feedback. The researchers varied the amount of study or testing and measured retention after five minutes, two days, or one week.
Repeated study helped on the very short-delay test, but prior testing produced substantially better retention on the delayed tests. Repeated study also increased learners' confidence in how well they would remember. That contrast matters: fluent rereading can create a strong feeling of learning while a difficult retrieval attempt produces less comfort but more durable access later.
After reading, close the material and reconstruct the main ideas before looking back. Use short-answer or free-recall prompts before recognition questions. Revisit the same idea after a delay, then compare the response with the source and correct omissions. Keep the stakes low enough that an unsuccessful attempt becomes diagnostic information rather than a reason to avoid practice.
In Nuxflo, this supports recall steps after lessons, delayed review queues, and feedback that distinguishes “not retrieved yet” from “not understood.” A useful sequence is attempt, inspect, correct, and retrieve again later.
The reported experiments used particular materials, schedules, and tests. Retrieval practice is not a substitute for initial explanation, and unsupported errors should not simply be rehearsed. Feedback, suitable difficulty, and alignment between what is practiced and what later needs to be used remain important design choices.
Primary source: Psychological Science DOI.
