Benefits of pretesting prior to retrieval practice are limited, unless used for prior knowledge-based personalization.

Abstract

An abundance of research has demonstrated that both posttesting (also referred to as retrieval practice) and pretesting (asking a learner for an answer to a cue before a study trial has been provided) can enhance the long-term retention of material. While the benefits of retrieval practice have been widely applied in various real-world applications, such as computerized tools that promote the memorization of factual materials, pretesting has seen limited real-world application. In this study, we examine whether and under which realistic digital learning conditions combining pretesting and posttesting can promote learning. In four experiments (total N = 210), we contrast learning conditions in which repeated retrieval practice is preceded by passive study to learning conditions in which retrieval practice is preceded by a test. In the first two experiments, we confirm and extend previous findings by demonstrating that pretesting boosts retrieval accuracy and reduces response times on subsequent retrieval repetitions, regardless of the accuracy of the pretest. We find these effects both when a fixed item repetition schedule is used and with performance-based, adaptive item scheduling that resembles popular digital learning tools. However, after three repetitions of an item, the initial advantage of pretesting disappears, calling into question its usefulness in applied settings that involve spaced repetition. In the final two experiments, we explore a more targeted use of pretesting, leveraging it to assess prior knowledge. Dropping items that were answered correctly during the pretest enhanced overall learning efficiency, especially for learners with moderate to high prior knowledge, without disadvantaging those with low prior knowledge. (PsycInfo Database Record (c) 2025 APA, all rights reserved).