Abstract
Instructional quizzes are frequently used in educational games. When they present correct answers after learners have responded, these quizzes can be used on their own for teaching new factual and conceptual knowledge (no additional learning materials are needed). In games, these quizzes are often unrelated to gameplay: gameplay can be viewed as a reward for answering quiz questions. This has been criticized in game-based learning literature as a "chocolate-covered-broccoli" approach. However, is it really a bad approach? Theories offer conflicting predictions concerning the instructional efficiency of in-game quizzes relative to bare quizzes (i.e., not embedded in games) and empirical literature is lacking. Here, we present a within-subject design study (N = 69), in which 10-12-year-olds learn from both an in-game quiz and a bare quiz and undergo immediate and 2-3 weeks delayed post-test on the quiz questions. A modest difference in learning outcomes favoring the bare quiz was found in the immediate post-tests (d = 0.46), but not in the 2-3 weeks delayed post-tests (d = 0.09). Children enjoyed the game more than the bare quiz (d (z) = 0.65) and 59 preferred the game in the free-choice period. The findings suggest that both a bare quiz and a quiz within a game have their place at the table for useful educational interventions: the bare quiz should be preferred in schooling contexts; whereas, the game in leisure time situations as a voluntary activity. In the latter case, it should be considered how the game and the quiz are integrated.