A shortcoming of Massive Open Online Courses (MOOCs) is the lack of substantive integration of hands-on activities into online classrooms. We conducted a randomised control trial (RCT) in a MOOC where we sent the treatment group at-home lab kits with which they could perform experiments in their own homes. In our intervention research, we found that, at the end of the course, students in the treatment group had significantly higher exam scores and levels of self-efficacy to perform in the topic area than their control group counterparts. Students' growth in self-concept was not significantly different. Findings have implications for distance courses in engineering and other science, technology, engineering, and math areas that traditionally have laboratory components. Furthermore, our findings illustrate the potential amplifying effect of this learning environment configuration for engineering instructors. Finally, this study illustrates how the dispersed nature of MOOCs facilitates RCTs and a recursive process linking theory and practice.