Milo Schield
Session Ended
Information
The 2016 GAISE update recommended multivariate thinking and confounding. But none of their suggested methods allowed students to work problems with numeric answers. Students in Math1300 at the Univ. of New Mexico work problems (without computer software) that take into account the influence of a measured confounder. This is a game-changer for most of our students who deal primarily with observational data. Learn how it's done. See how controlling for a confounder can change statistical significance into insignificance -- and vice versa. See (visually) Simpson's paradox. Talk about including confounding in your introductory statistics class.