Answers are grounded in this paper and cite their sources.
Extensive research has demonstrated that testing enhances long-term memory by strengthening retrieval processes for the tested memory traces. Such memory models explain why learners benefit from repeated testing, but they do not account for how learners construct understanding across multiple ideas. In contrast, theories of discourse comprehension emphasize the construction of inferences, whereby learners actively connect ideas into a coherent mental representation. These two perspectives make distinct predictions about what learners retain and understand. The present study tested these predictions by comparing a test-only procedure, in which learners encountered the material exclusively through questions and corrective feedback and never read the source text, against self-explanation and reading, measuring both fact memory and inference performance. In Experiment 1 (n = 254), despite superior retention of individual facts, test-only learning produced lower inference performance than self-explanation. Experiment 2 (n = 68) was a preregistered replication with a minimum 24-h delay between learning and test sessions, and which repeated the learning phase in two blocks, enabling participants to retrieve from memory the answers they had encoded. The introduction of retrieval practice improved recall of facts but did not facilitate inference performance. A mixed-methods analysis revealed that generating inferences during learning specifically predicted inference performance, whereas noting facts predicted fact recall, providing converging evidence that these strategies produce different mental representations. Furthermore, participants generated fewer inferences when self-explaining science texts compared with narratives, highlighting domain differences in knowledge integration. Together, these results demonstrate that while test-only learning strengthens memory for discrete information, it does not foster the integrative processing necessary for comprehension.
No takes yet. Share an insight, caveat, or question.
Motz et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: