Retrieval practice may not benefit mathematical word-problem solving.

The retrieval practice effect refers to the fact that one or even multiple retrievals of memory content during the same period are more effective than repeated studying to promote future memory retention. It is effective for numerous declarative knowledge learning materials. However, studies have demonstrated that retrieval practice does not benefit problem-solving skill learning. This study used worked examples from math word problem tasks as learning materials, considering the retrieval difficulty as the main factor. Experiment 1 explored the effect of retrieval practice on acquiring problem-solving skills under different initial testing difficulties. Experiment 2 manipulated the difficulty of materials as a variable to ascertain the effect of retrieval practice on problem-solving skills under different material difficulty levels. Experiment 3 introduced feedback variables to facilitate the generation of the retrieval practice effect and examined the effects of various difficulty feedback levels on problem-solving skills learning. Results showed that, compared with restudying examples (SSSS), the example-problem pairs (STST) did not promote delayed test performance. As for the retrieval practice effect, as no differences or advantages were found in the repeated study group on the immediate test, the retrieval practice group generally outperformed the repeated study group on the delayed test. However, across the three experiments, we found no evidence of retrieval practice affecting results during an enhanced delayed test. Therefore, there may be no retrieval practice effect on acquiring problem-solving skills from worked examples.