Retrieval-based learning and prior knowledge.

Abstract

Retrieval practice typically benefits later memory more than restudy (i.e., the testing effect). The benefits of retrieval-based learning generalize across a range of materials and contexts, leading many cognitive scientists to advocate for broad educational implementation. However, educators and practitioners call for more research on factors critical to education (e.g., individual differences between learners). One key understudied factor is the amount of relevant prior knowledge held by the learner, which can moderate the effectiveness of other learning strategies (i.e., expertise-reversal effects). The current study investigated the potential moderating influence of prior knowledge on the learning benefits of retrieval. To assess the causal effects of prior knowledge on later learning strategy effectiveness, we randomly assigned online participants to be trained in multiple topics within one of two academic domains over 3 days. Participants then engaged in a more typical testing effect paradigm-they studied new scientific text passages related to their trained or untrained domain and completed two rounds of focused restudy or retrieval with elaborative feedback (repeated for the other domain). Two days later, participants took a final test on the previously learned information from both domains. Despite being rated as more effortful than restudy, retrieval practice led to greater overall test performance (with larger relative benefits on retention than transfer questions). Critically, despite a substantial benefit from training-induced knowledge, testing effects were nearly identical for high and low prior knowledge information. This suggests that prior knowledge is not a critical boundary condition of retrieval-based learning. Educational Impact and Implications StatementThis study compared the effectiveness of two learning strategies, restudy (i.e., focused restudying of examples or key ideas) versus retrieval practice (i.e., short answer practice questions with detailed feedback), for learners with low versus high prior knowledge of the to-be-learned scientific text passages. Retrieval practice led to greater overall performance than focused restudy (i.e., about þ10% or one letter grade higher) and these benefits were similar for learners with low and high prior knowledge. This indicates that retrieval practice is more effective than restudy when learning complex, educationally relevant information and that students with lower and higher levels of relevant prior knowledge will similarly benefit from retrieval-based learning.

The magnitude of the testing effect is independent of retrieval practice performance.

“…The null effects in Experiment 3 are especially striking given the magnitude of the effect sizes observed on retrieval practice performance ( d = 2.38 in Experiment 3a and d = 1.97 in Experiment 3b). Although this result might seem surprising, it dovetails with recent evidence that individuals’ susceptibility to the testing effect is independent of prior knowledge (Buchin & Mulligan, 2023). Specifically, Buchin and Mulligan’s high prior knowledge participants exhibited better retrieval practice performance than their low prior knowledge participants, the two groups of participants had comparable testing effects.…”

Retrieval‐based learning in special education

“…Yet, even with regard to such specific knowledge, findings are mixed. Some studies reported retrieval‐based learning to be independent from certain domain‐specific knowledge (Buchin & Mulligan, 2023; Xiaofeng et al, 2016). Moreover, measures of fluid intelligence have not been found to correlate with the strength of the testing effect or an even stronger testing effect was found for students with lower ability (Agarwal et al, 2017; Bertilsson et al, 2017; Brewer & Unsworth, 2012; Pastötter & Frings, 2019).…”

The Testing Effect in the Lecture Hall: Does it Depend on Learner Prerequisites?

“…However, it seems questionable whether a no-treatment control group is an adequate point of comparison for assessing the effectiveness of retrieval practice and the role of prior knowledge. Finally, the only study in which prior knowledge was manipulated experimentally found no moderating effect of prior knowledge on the testing effect (Buchin & Mulligan, in press; for null results using a pre-experimental measure of prior knowledge, see also Xiaofeng et al, 2016).…”