66-108 DC First-Year Seminar: Statistical Paradoxes-When You Can't Trust Your Own Eyes

9 units, Carnegie Mellon University

Humans are notoriously bad at probabilistic thinking. Were crushed if our hometown team loses when its the favorite, and ecstatic when we go on a lucky streak playing craps we constantly see patterns in randomness. Even Paul Erdos, one of the greatest mathematicians of the last few centuries, was famously wrong about a deceptively simple probability puzzle. When it comes to statistics, we often run into trouble when we rely on intuition we cant trust our own eyes. In this class we will tease apart some entertaining but important statistical paradoxes and biases. For example Why are better health outcomes reported when only medical screening improves How could David Justice have a better batting average than Derek Jeter in both 1995 and 1996 separately, but not in 19951996 combined Could most published research findings actually be false Why should attorneys be forced to learn Bayes Theorem Can we really know that smoking causes lung cancer This course will help you answer these questions, teach you to think critically about research and news reports, or at least give you some entertaining anecdotes to tell at parties. We will also use the statistical software R for handson exploration

Prerequisites: none listed

Courseweave needs JavaScript to show sections, workloads and planning tools.