125 – SPEED: Modernization of What, How, and Where We Teach Statistics Part 1
Paradox Problems as a Tool for Understanding Statistical Reasoning
Bailey Farrar
Southern Illinois University Edwardsville
Andrew Neath
Southern Illinois University Edwardsville
Thought experiments in the form of paradox problems are useful for illustrating how our statistical reasoning may be correct, or how it may be flawed, and how our models for behavior in the sciences may be appropriate, or how such models may be incomplete. For example, the Monte Hall problem, the Exchange paradox, the Ellsberg paradox, and transposed conditional paradoxes are all interesting exercises in decision making, quantifying uncertainty, and statistical inference. In this presentation, we will take a careful look at what each of these paradox problems can teach us, and our students, about the best practices in statistics.