Why Does Closing the Book and Testing Yourself Often Beat Rereading?

By the third reading of the same page, the words flow easily and can create a strong sense that the material has been learnt. Familiarity, however, is not the same as being able to retrieve something later. Retrieval practice means temporarily closing the book and reconstructing what you have just learnt from a question, heading or small cue.

It is not merely a way to measure performance. Each successful retrieval practises finding that knowledge from a cue and reveals which connections remain weak. After reading a paragraph about refraction, keeping the text visible preserves many visual prompts. Closing the book and answering “Why does a straw look bent?” requires you to rebuild the relationship among water, air, the path of light and apparent position. A later need for the knowledge is more like the second task.

Experiments can directly compare rereading with self-testing under controlled conditions, so they support a causal effect of retrieval practice on memory. Rowland's meta-analysis found an overall retention advantage for testing over restudy, with recall tests generally producing a larger benefit than recognition tests. The effect is not a fixed multiplier: material, spacing, success on the initial retrieval and the form of the final test all matter.

Self-testing is therefore not better simply because it is harder, and an error should not be left unchecked. If nothing can be retrieved, or a mistake is repeatedly recalled, practice may be ineffective or misleading; prompt checking and feedback are important. The central point is that rereading mainly increases present fluency, while retrieval practice trains the act of access that will be needed later. The two can work together, but “This looks familiar” should not be mistaken for “I can use this without the page”.

https://pubmed.ncbi.nlm.nih.gov/25150680/
https://pubmed.ncbi.nlm.nih.gov/20951630/