Testing the effectiveness of two natural selection simulations in the context of a large-enrollment undergraduate laboratory class

Background: Simulations can be an active and engaging way for students to learn about natural selection, and many have been developed, including both physical and virtual simulations. In this study we assessed the student experience of, and learning from, two natural selection simulations, one physical and one virtual, in a large enrollment introductory biology lab course. We assigned students to treatments (the physical or virtual simulation activity) by section and assessed their understanding of natural selection using a multiple-choice pre-/post-test and short-answer responses on a post-lab assignment. We assessed student experience of the activities through structured observations and an affective survey. Results: Students in both treatments showed increased understanding of natural selection after completing the simulation activity, but there were no differences between treatments in learning gains on the pre-/post-test, or in the prevalence of concepts and misconceptions in written answers. On a survey of self-reported enjoyment they rated the physical activity significantly higher than the virtual activity. In classroom observations of student behavior, we found significant differences in the distribution of behaviors between treatments, including a higher frequency of off-task behavior during the physical activity. Conclusions: Our results suggest that both simulations are valuable active learning tools to aid students' understanding of natural selection, so decisions about which simulation to use in a given class, and how to best implement it, can be motivated by contextual factors. © 2017 The Author(s).