Title: Evaluating change in learning from different forms of interactive visualizations with a large case study
Authors: Leanna House - Virginia Tech (United States) [presenting]
Lata Kodali - Virginia Tech (United States)
Abstract: Cutting edge software has been preciously developed that allows novice analysts to explore high-dimensional data interactively. The software, Andromeda, effectively responds to user inputs in the forms of interaction with data at the observation-level and/or parametric levels to create multiple Weighted Multidimensional Scaling (WMDS) projections. We evaluate the impact Andromeda has on student learning via a large-scale user study implemented in an introductory statistics course at Virginia Tech. This study includes approximately 150 students and was conducted in two different semesters. Using a Bayesian approach, we share our findings from this user study, including significant differences in mastery of WMDS, complexity of insights, and change in attitude toward engaging in data analyses.