By Kit Clement
Information
Simulation-based inference has been advocated by educators and researchers for its power in helping students understand statistical inference at a deeper conceptual level than the consensus curriculum. This poster focuses on the Change Agents for the Teaching and Learning of STatistics (CATALST) curriculum, which is unique among simulation-based curriculum for its focus on probability modeling in TinkerPlots. I developed new activities that introduce concepts of regression to students and builds on their modeling knowledge for comparing two groups or populations with randomization tests. These activities were implemented in a small undergraduate introductory statistics class (20-30 students) at a large, public, urban university. Students’ learning through these activities was assessed using pre-post task-based surveys and post-interviews. Students who completed these new CATALST regression activities not only showed greater progress in their learning of regression concepts than those who took a more traditional course, but were more prepared to apply inferential concepts to a novel data scenario before even first formally learning this content. These results emphasize the importance of emphasizing generalization in hypothesis testing and distinguishing testing from other descriptive methods in linear regression like correlation.