This research investigates the role of virtual experiments and worked examples in the learning of the control of variable strategy (CVS). Sixty-nine seventh-grade students participated in this study over a span of 6 weeks and were engaged in worked example learning and/or virtual experimentation to study the knowledge and procedures associated with CVS. The results show that learning from only worked examples or virtual experiments yields minimal knowledge gain on CVS. While worked examples can be effective learning tools, additional instructional support and hands-on experiences are critical to elicit deeper understanding of CVS. Moreover, virtual experimentation alone is not sufficient given the cognitive demands it requires for students to discover CVS on their own. However, when integrated with worked example learning, conducting virtual experiments with computer visualizations can lead to positive learning outcomes. This work corresponds to the growing research on "productive failure," and opens up further discussions on optimal strategies for learning CVS.