Improving multivariable statistical thinking by teaching sources of variation

Abstract

In an increasingly data-centric world, where both consumers and creators of statistical information are required to understand the complex associations among the many variables that form the basis for good data-driven decisions, we pose the question: how can teachers of first and second courses of statistics introduce and build facility with multivariable thinking? The GAISE guidelines include multivariable thinking in the first course, and it’s natural to assume that multivariable thinking also must be a focus of the second course. However, there is still a gap between what GAISE promotes and what instructors teach when trying to reach students that need pedagogically-sound, conceptually-based methods and visualizations to develop multivariable thinking. In this workshop we focus on how teaching sources of variation can be a key strand in developing multivariable thinking that cuts across the entirety of both the first and second algebra-based courses in statistics.