The role of variable retrieval in effective learning

Current research on memory indicates that learning is most effective when it involves spaced retrieval practice of study materials. Here, we investigated whether the benefits of both retrieval practice and spacing can be further boosted when variability is introduced across practice sessions. Across six experiments, participants learned translations of foreign vocabulary, with foreign words embedded in contextual sentences hinting at the meaning of these words. These sentences were then either kept constant or varied from one learning cycle to another. Whenever repeated practice took the form of retrieval from long-term memory with contextual sentences serving as cues (with or without feedback after retrieval attempts), variable sentences led to better learning of the meanings of the embedded foreign words than constant sentences. The benefits of variable retrieval were observed both immediately after study and after a 24-h delay, and they were larger when retrieval practice was spaced rather than massed. However, these benefits were not appreciated by the learners who judged learning to be more effective with constant rather than variable cues. This metacognitive illusion, misaligning the effectiveness of learning and its appraisal by learners, was confirmed in the seventh experiment which focused on learning lecture content. Thus, while spaced retrieval practice employing variable cues clearly produces robust benefits for memory performance, such benefits may be severely underappreciated by the learners.