Abstract:
|
Statistics educators---imbued with improved computing power---have advocated for a greater emphasis on randomization and simulation-based techniques for statistical inference in recent years. While these ideas are not new, the traditional treatment of inference in introductory statistics courses has focused on methods that approximate the sampling distribution of a statistic with a probability distribution. We describe an approach to teaching inference in an introductory statistics course---for students who primarily major in the sciences---that emphasizes randomization and simulation-based approaches, briefly discusses but largely glosses over mathematical approaches using probability theory, and treats normal-based approximations as an alternative technique. The overall conceptual goal is to understand the sampling distribution of the test statistic: the three approaches are just means to that end.
|