Online Program

Friday, February 20
PS2 Poster Session 2 & Refreshments Fri, Feb 20, 5:15 PM - 6:30 PM
Napoleon AB

Viewing Problems Through a Multilevel Modeling Lens (303013)

Julie Legler, St. Olaf College 
*Paul Roback, St. Olaf College 

Keywords: Multilevel modeling, generalized linear models, longitudinal data, undergraduate education

At St. Olaf College, we teach generalized linear models and multilevel modeling as a core applied course to all our statistics concentrators. Even though this material often is left for the graduate level, we have found that undergraduates can understand it when presented in an intuitive and applied manner. Better yet, expanding their horizons beyond linear regression has allowed these students to apply multilevel models—including models with more than two levels, a longitudinal component, or a generalized response—to real-world data addressing a variety of research questions. In this way, the students’ modeling approach is more faithful to the true nature of the data they are examining, especially when responses are non-normal or observations are correlated. This poster will illustrate how the multilevel modeling lens is developed in undergraduates and how they have applied this lens to problems ranging from performance anxiety among musicians, to pitch count limits in baseball, to race and the New York City stop and frisk program, to comparisons of charter and non-charter public schools, to seed germination and prairie restoration.