Many studies have shown benefits associated with engaging students in problem-solving activities prior to administering lessons. These problem-solving activities are assumed to activate relevant knowledge and allow students to develop some initial knowledge structures, which support understanding of the lesson. In this paper we report the results of two studies in which we investigated the underlying benefits of engaging in a preparatory activity--setting up experiments without running them or receiving feedback--prior to an interactive computerized lesson on experimental design compared to only engaging in the interactive lesson. We predicted that the seventh-grade participants who demonstrated some initial knowledge of the topic--experimental design--would benefit more from spending the whole time engaged in instructional activities. However, we expected students who did not demonstrate initial knowledge would benefit more from engaging in the preparatory activity, which would allow them to activate or develop initial knowledge that would aid their understanding of the subsequent instruction. The predicted condition by initial knowledge interaction was found in both studies. In Study 1, the benefit of only engaging in the instruction was found only for the lowest-knowledge of students who demonstrated initial knowledge. For students who did not demonstrate some initial knowledge, the benefit of completing the preparatory activity appeared to be due to the development of an understanding of the general goal of the activity rather than of specific knowledge of experimental design. Based on this finding, in Study 2, we investigated an initial "goal" by condition interaction. In fact, students who did not express an understanding of the task goal on the pretest benefited from engaging in the preparatory activity and students who did benefited more from the instruction. Again, this benefit appeared to be due to students' development of an appropriate understanding of the task goal during the preparatory activity.