ABSTRACT
The impact of retrieval practice on analogical-problem-solving performance was investigated using a complex, educationally relevant task. Participants studied a statistical hypothesis testing scenario and practiced recalling the material or repeatedly studied it. Participants then completed a final test either 5 minutes or 1 week later involving a novel hypothesis-testing scenario that shared an intermediate procedural strategy and superficial and structural similarity with the study scenario but that differed at a specific procedure level. When the final test was given after 5 minutes, no differences in performance were observed across conditions (d = 0.01). Crucially, on the delayed test, retrieval practice produced superior performance than did repeated studying (d = 0.81), whereby participants were better at applying learned knowledge to solve a novel problem.