We assess the relative effectiveness of two different active learning complements to traditional lecture-based learning. Using a large introductory class at a large public university, we conducted an experiment designed to evaluate whether active learning approaches provide a significant improvement in a student's short-term retention of material over only attending a standard large lecture. In this Introduction to World Politics class, each teaching assistant taught one section using a common discussion lesson plan and one section using a brief role-play activity. Using multiple regression analysis, we find that the addition of an instructor-led discussion significantly improves student performance on the short answer portion of a brief post-activity assessment, but not on the multiple choice portion. The role-play sections perform significantly better on multiple choice portion than the lecture-only group. In comparing the two treatments to each other, we find no statistically significant difference in the performance of the two groups.