The reach of retrieval: Exploring Retrieval Practice in Authentic Primary School Classrooms

BACKGROUND
A key goal of primary education is to help students acquire knowledge, retain it over time, and use it in meaningful ways. However, recent large-scale assessments, such as TIMSS and PIRLS, have shown a decline in Dutch students’ performance in the domains of mathematics and language (Inspectie van het Onderwijs, 2023; Mullis et al., 2021; Mullis et al., 2023; OECD, 2023). These findings raise urgent questions about how schools can support students acquiring knowledge that not only endures but can also be transferred to new situations. From a cognitive-educational perspective, retrieval practice is considered a promising, evidence-based learning strategy (Adesope et al., 2017; Dunlosky et al., 2013; Rowland, 2014). By actively recalling previously learned information instead of restudying it, students strengthen memory traces and make knowledge more accessible for future use (Roediger & Karpicke, 2006). The effectiveness of retrieval practice has been convincingly demonstrated across hundreds of studies. Yet, research in authentic primary school classrooms across different domains remains scarce, particularly for more complex learning goals such as text comprehension and procedural rule application (Agarwal et al., 2021; Yang et al., 2021).

AIM
The aim of this dissertation was to investigate to what extent retrieval practice can be effectively implemented in authentic primary school classrooms to enhance primary school students’ mathematics and language skills. Four empirical studies were conducted, all within everyday classroom contexts and using existing teaching materials. The first three studies addressed this question from an instructional perspective, comparing retrieval practice, supported by spacing and feedback, with other learning strategies in the domains of mathematics, text comprehension, and verb spelling. The fourth study examined the question from a student perspective, exploring how students perceive the effectiveness of different learning strategies. Together, these studies provide insight into both the benefits and the boundaries of retrieval practice for improving students’ mathematics and language learning in authentic classroom settings.

STUDIES AND MAIN FINDINGS
The first study investigated whether retrieval practice improves fluent recall of multiplication facts compared to chanting the tables out loud. In this study 48 students practiced multiplication tables using both strategies in counter-balanced sessions. Retrieval practice consisted of spaced practice with flashcards, supported by feedback. Results showed that retrieval practice led to significantly higher accuracy and fluency in recalling multiplication facts than chanting the tables out loud, both after five minutes and after one week. These findings demonstrate that retrieval practice can effectively strengthen factual knowledge in real classroom contexts within the domain of mathematics. T he second study examined whether the benefits of retrieval practice extend beyond factual recall to deeper text comprehension. In a classroom-based withinsubjects experiment, 57 fifth- and sixth-grade students studied authentic school texts using either retrieval practice, summarization, or restudy (control condition). Two weeks later, both factual recall and text comprehension were tested. Retrieval practice led to better factual retention than summarizing or restudying, whereas summarization and restudy supported better comprehension of the text content. T he third study examined whether combining retrieval practice with worked examples could enhance both the retention and application of procedural knowledge. In this study, 105 sixth-grade students learned Dutch verb spelling rules using either worked examples alone or a combination of worked examples and retrieval practice. One week later, students were tested on both recall and application of the verb spelling rules. Results showed that adding retrieval practice to a worked-example approach substantially improved long-term retention of the verb spelling rules, but did not result in significantly better application. T he fourth study shifted the focus from instruction to students’ perceptions of learning strategies. In a large-scale survey among fifth- and sicth-grade students, perceptions of the effectiveness of several strategies, including retrieval practice, were examined. Students overestimated the effectiveness of more fluent but less powerful strategies such as rereading and highlighting, and underestimated more effortful but effective strategies such as retrieval practice. Nevertheless, some students showed emerging conditional metacognition, recognizing that certain strategies work better for specific school subjects. Moreover, several students indicated that combining strategies can be effective.

ANSWERING THE CENTRAL RESEARCH QUESTION
Across four empirical studies, three from an instructional perspective and one from a student perspective, this dissertation shows that retrieval practice, supported by spacing and feedback, can substantially enhance students’ mathematics and language skills in authentic primary school classrooms. Across all studies, retrieval practice consistently improved the retention of factual knowledge, confirming its robustness across domains and age groups. Moreover, findings from the student study suggest that primary school students are developing the metacognitive awareness needed to use learning strategies more deliberately and effectively, an important step toward self-regulated learning. However, the effectiveness of retrieval practice depended strongly on the type of learning goal. While it reliably supported the retention of factual knowledge, its benefits were less clear for complex learning goals that required comprehension or procedural application. From several theoretical perspectives, this pattern can be explained. According to the Transfer-Appropriate-Processing theory (Morris et al., 1977), learning is most effective when the cognitive processes engaged during practice resemble those required at test. Practice tasks that elicit only factual recall are therefore less likely to strengthen the inferential or procedural reasoning needed for deeper understanding or accurate application. Similarly, the Knowledge-Learning-Instruction (KLI) framework (Koedinger et al., 2012) emphasizes that distinct learning goals rely on distinct cognitive processes. Based on this framework and our findings, retrieval practice appears most effective for memory-based learning goals, while strategies such as worked examples may better support inductive learning. Finally, Cognitive Load Theory (Sweller et al., 1994, 2011) suggests that when retrieval and application are combined in more complex tasks, cognitive load may limit the integration and application of retrieved knowledge. Together, these insights indicate that retrieval practice is a powerful but limited strategy. Its success depends on how it is designed and how well it aligns with the underlying learning processes of the learning goal. T he success of retrieval practice in primary education depends not only on its evidence-informed implementation by teachers and textbook authors, but also on students’ motivation and metacognitive understanding of effective learning strategies. Many students still overestimate the value of fluent but less effective approaches, such as rereading or highlighting, and underestimate the power of effortful retrieval. Explicitly discussing how, when, and why retrieval practice works can help students recognize that the mental effort involved signals genuine learning rather than failure. Over time, this awareness can enhance both students’ engagement and their academic performance in mathematics and language learning. Future research should continue to explore how retrieval practice can best be adapted to different learning goals, by examining the cognitive mechanisms underlying retrieval effects and moderating factors such as task complexity and sequencing. Such research will help to clarify the conditions under which retrieval practice contributes not only to durable retention of factual knowledge, but also to comprehension and procedural application of knowledge. In addition, studies should investigate how students’ metacognitive knowledge and motivation can be strengthened, for example through explicit strategy instruction or awareness training.

EDUCATIONAL IMPLICATIONS
The findings of this dissertation provide clear guidance on how retrieval practice can be meaningfully implemented in primary education. For students, retrieval practice offers a way to strengthen long-term retention of knowledge across domains. Using flashcards, recalling facts from text material, or recalling procedural rules, helps students learn more effectively. Spacing retrieval sessions and combining them with feedback enhances results, while reflecting on the purpose and difficulty of retrieval promotes metacognitive awareness and selfregulation. For teachers, the key task is to create classroom conditions that make retrieval practice both effective and engaging. Short quizzes, flashcards, and open recall tasks can help consolidate factual knowledge across domains, especially when spaced and followed by feedback. By explaining why retrieval practice works, teachers can strengthen students’ motivation and understanding. Staying informed about the latest research enables teachers to implement retrieval practice more effectively. For textbook authors, the challenge lies in embedding retrieval opportunities systematically within instructional materials across domains. Short, spaced recall activities supported by feedback and reflection prompts help make the learning process visible. Keeping materials aligned with current research ensures that the most effective principles are consistently translated into classroom practice. Doing so will help bridge the gap between educational research and classroom practice and strengthen conditions under which retrieval practice can enhance mathematics and language skills of primary school students.

GENERAL CONCLUSION
This dissertation demonstrates that retrieval practice is a powerful yet limited learning strategy in primary education. It consistently strengthens the durable retention of factual knowledge in authentic classroom settings, but its benefits for comprehension and procedural application are limited and need further examination. To realize the full potential of this learning strategy, collaboration among researchers, textbook authors, teachers, and students is needed. Researchers should continue to clarify the mechanisms and boundaries of retrieval practice; textbook authors and teachers should translate these insights into instructional design and classroom practice; and students should learn to use retrieval practice strategically and reflectively. Together, these actors can create evidence-informed learning environments in which retrieval practice supports meaningful and lasting learning in primary education.