Several schools introduced frequent low-stakes quizzes. Students improved substantially where each quiz was followed by feedback and required correction of errors, but schools that increased quiz frequency without changing feedback saw little improvement. Interviews indicated that students in the first group revised misconceptions between assessments rather than merely rehearsing answers. Researchers concluded that retrieval coupled with corrective feedback, not testing frequency by itself, drove the gains. They therefore advised schools to design quizzes as feedback loops rather than simply adding more tests.