Nonlinear, Non-Normal, Non-Independent? A Course About Models for Situations When Classical Regression Assumptions Don't Apply

• Activity Number: 414
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Education
• Author(s): Alison Gibbs*+
• Companies: Univ of Toronto
• Keywords: statistics education ; curriculum ; statistical practice ; generalized linear models ; mixed models

Abstract

The University of Toronto has over 65,000 undergraduate students. Last year the Department of Statistical Sciences offered 18 one-semester courses at the 3rd and 4th year levels for students who have completed one of our sequences of introductory 2nd year courses. One of our newer offerings is a post-regression course in linear models that is designed to contribute to the development of expert data analysts with a collaborative bent. I'll talk about the hole we identified in our undergraduate program that led to the development of this course, and some successes and challenges we've experienced, including coping with the lack of an appropriate textbook. Despite our assortment of competing offerings, this relatively new course has taken a prominent place in our program as the "third" course in statistical practice.