Broaden Your Statistical Horizons: GLMs and Multilevel Models for Undergraduates

• Activity Number: 575
• Type: Invited
• Date/Time: Wednesday, August 12, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Education
• Author(s): Paul Roback* and Julie Legler
• Companies: St. Olaf College and St. Olaf College
• Keywords: undergraduate curriculum ; generalized linear models ; correlated data methods ; multilevel models

Abstract

At St. Olaf College, students follow up a standard course in applied regression with a course covering generalized linear models and multilevel methods. Our philosophy with our Advanced Statistical Modeling course is to truly "broaden students' statistical horizons". Students explore modeling options when data contains non-normal responses, an inflated number of zeros, correlated observations, or predictors at several levels, while simultaneously solidifying core regression ideas. We also introduce students to more advanced topics such as likelihood functions, parametric bootstrapping, and error distributions. Given that many of our concentrators are not math majors, our goal is to build intuition and foster an appreciation for the wider world of statistical methods and ideas, in a way that's accessible even to students without a deep mathematical background. To accomplish this goal, we have developed our own materials, since most existing materials on these topics were written with graduate students in mind. In this talk, we will describe our successes and lessons learned over 10 years of teaching Advanced Statistical Modeling at St. Olaf.