|
Abstract:
|
Despite being an important concept in statistical theory, sampling distributions are difficult to convey in introductory courses. In part these difficulties stem from preliminary topics (e.g., probability theory and statistical independence) that must first be mastered. However, even with simple discrete random variables, the rapid expansion of the calculations needed to visualize results as sample sizes increase can quickly introduce not only obstacles for student learning but also teaching challenges for instructors. Using readily available software to overcome computational issues, this presentation describes the construction of sampling distributions for two "competing" discrete processes in forms suitable for a single lecture to illustrate key concepts (e.g., unbiasedness). An example demonstrates an application involving small sample sizes (e.g., disease outbreak investigations).
|
