Title: Teaching statistical thinking to statistics graduate students
Authors: Walter Stroup - University of Nebraska (United States) [presenting]
Abstract: Most statistics graduate programs have a core curriculum that includes, at a minimum, theory (mathematical statistics and probability) and statistical methods (linear models and - possibly - design of research studies). George Box once commented that at some point, the discipline of statistics had been hijacked by mathematics. Although our discipline has made efforts to address Professor Box's concern, it is still true that students earning Masters degrees in statistics tend to know far more about the mathematics and computing of statistics than they do about what I call ``the statistics of statistics''. Ideally, statistical methods courses should be viewed as science courses. Statistics is not mathematics, nor is it a ``branch of mathematics'', as many dictionaries misleadingly define it. Statistics is a bridge between mathematics and science (as broadly defined). In developing new courses for the University of Nebraska's revision of its Statistics MS curriculum, we have developed teaching tools and student activities designed to help students ``get inside the scientist's head''. Students learn the methods and their supporting theory, but more importantly, they learn how to use these methods to help scientists answer the questions that motivate them to plan studies and collect data.