Does retrieval practise improve learning in real-life classrooms? A recent database will help us answer this question

Retrieval practise – the act of recalling prior knowledge and information to be learned – is a well-established strategy to improve long-term learning and memory. As previous reviews indicate (e.g., Nunes & Karpicke, 2015), the benefits of simply retrieving facts instead of rereading them have been demonstrated in the laboratory (e.g., the milestone study by Roediger and Karpicke, 2006) as well as in the classroom (e.g., McDaniel et al., 2011). However, do we know the full extent of retrieval practice's benefits in the classroom? Or in what circumstances might retrieval be helpful, or not? 

A systematic review of research on retrieval practise, conducted in classroom settings, was presented in October 2019 (Agarwal, Nunes, & Blunt, 2019). The study presents a free open-access database that combines the analyses of 50 experiments drawn from 37 peer-reviewed publications between 1999 and 2019 (click here to access the database and the project summary). Even though this study is still a work in progress, which has not yet been peer-reviewed, it provides instructors and researchers quick access to some facts about retrieval practice in classrooms.

For this review, the researchers searched the main scientific repositories and identified the experiments in which retrieval practise had been compared to «business-as-usual» practices in authentic classrooms. These experiments took place in classroom contexts, resorted to relevant learning materials, which were part of the assigned class topics, and took the same time to implement as the more commonly used "classic" strategies. The scientific articles in question were categorized according to particular conditions of interest, including:

• Educational levels (K-12 to college/university, and medical school);
• Content areas (including Biology, Chemistry, Physics, Psychology, History, Spelling and Vocabulary, and Statistics):
• Type of retrieval practice (e.g., lectures with quizzes, scenario-based quizzes, and quizzes with definition questions vs application questions);
• Retrieval and test formats (e.g., multiple-choice, short answer, free recall);
• Spacing (i.e. time lag) and frequency of the retrieval practise (e.g., every class, once a week, once a month);
• Comparison conditions (e.g., lessons without quizzes, infrequent high stakes exams, restudying);
• Provision of feedback (e.g., immediate, delayed, no feedback);
• Number of students tested;

The preliminary results of this review indicate that retrieval practise do increase learning in real-life classrooms. The 50 experiments reviewed yielded more than 60 comparisons of interest, of which 59 were translated into measures of the effect size of retrieval practice. Figure 1 summarises the effect size of retrieval versus common educational strategies, which rely on rereading or on letting students' study on their own. The effect sizes were measured by Cohen's d, which, in this case, represent the magnitude of the changes introduced by retrieval practice. Cohen’s d values between 0.2 and 0.5 are considered 'small' effect sizes; values between 0.5 and 0.8 are considered 'medium' effect sizes; and values equal or greater than 0.8 are considered 'large' effect sizes (Cohen, 1988). In applied research, even small effect sizes are important – imagine the importance of improving the learning of, at least, one student, or the overall learning of one class topic. The results show that retrieval practice appears to have hampered students' learning only in two cases. In five cases, the benefits of retrieval were very small; in 16 cases, the effects were small; in 15 cases, the effects were medium; and in 21 cases, the effects were large. In sum, in 88% of the examined comparisons, retrieval practise yielded significant and impactful improvements in students' learning. 

This review shows that, even though there are several variables that researchers cannot control easily in classroom contexts, such as absences, external motivators and commitments outside the classroom, these studies match the results obtained in highly controlled laboratory settings and support the advantages of using retrieval practice. So far, the results answer the question of whether retrieval practice improves learning in real-life classrooms – and the answer is "yes". However, future analyses of this database will provide information about the circumstances that might influence the benefits of retrieval practice, such as the grade level, the field of study, or the type of retrieval activity.