While instruction on control of variables has been shown to be effective, especially when it encourages students to focus explicitly on rules or procedures, little evidence of application to novel problems has been obtained. We hypothesized that prompting students to understand their own learning processes while doing experiments involving control of variables would allow them to activate their repertoire of knowledge and strategies and learn in a way that would enhance transfer of learning. Students were assigned to one of four versions of a computer-based biology simulation learning environment, each employing a different type of prompt: reason justification, rule based, emotion focused, or none (control). Learning in this computer environment, college biology students designed and conducted experiments involving control of variables. Students' ability to solve both contextually similar (near transfer) and contextually dissimilar (far transfer) problems was assessed. The treatment groups performed equally well on contextually similar problems. However, on a contextually dissimilar problem, the reason justification group had significantly higher scores than the other groups. Qualitative data showed that the reason justification prompts directed students' attention to understanding when, why, and how to employ experiment design principles and strategies, and this in turn helped students to transfer their understanding to a novel problem.