An Active Approach to Statistical Inference Using Randomiztion Methods

• Activity Number: 196
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Education
• Author(s): Todd Swanson*+ and Nathan Tintle and Jill VanderStoep and Vicki-Lynn Holmes and Brooke Quisenberry
• Companies: Hope College and Hope College and Hope College and Hope College and Hope College
• Keywords: curriculum ; randomization methods ; permutation tests ; introductory statistics ; active learning

Abstract

We have significantly changed our introductory statistics course in both content and pedagogy. We take an intuitive approach to inferential statistics using permutation tests and other randomization methods. This means that in the first week of the course students understand the structure of a test, what a sampling distribution is, and what a p-value means. This allows us to cycle through the core logic of inference throughout the course. We examine confidence intervals, power, and relate randomization methods to traditional techniques. We have also moved from a mainly lecture based class to one that is driven by group work, self-discovery, active-learning and tactile demonstrations using case studies, projects, and research articles. Our poster will include an overview of our curriculum, assessment data, sample class activities, and examples of student work.