Statistical Literacy
Taught by Matt Stevens
Matt is a political scientist and statistician with almost fifteen years of teaching experience at Columbia, the New School and NYU. He lives in Astoria, Queens with his wife and two cats, spending his spare time drawing, cooking, game designing and essay writing.
Statistical Literacy is a lecture course, with a few little games thrown in, but we use as little math as possible, and nothing more advanced than basic algebra, so beginners are welcome.
This course is devoted to the ideas behind statistics. These ideas are used in everything from sports to gambling, from physics to opinion polls.
We start with the question of causality: When correlation means causation, when it doesn't, and how experiments work into it. These ideas are key both to science and to everyday living. The kind of science you see in the newspaper will never look the same again.
Then we turn to summarizing variables. I'll show you some beautiful graphs, some horribly ugly ones, and some of the ways they can mislead you. We look at three meanings of "average," and how they can be used to tell different stories. We wrap it all up with "sigma" -- used in testing and engineering -- and the "standardizing" of test scores.
Next we look for order in the cloud. How to make sense of a scatterplot, what "correlation" means, and look at the all-important "regression effect," critical to understanding the "Sports Illustrated cover jinx." We'll touch on the Ecological Fallacy, and how it affects our view of Red States and Blue.
Finally, in the last section, we start by rolling dice and flipping coins to find that the "law of averages" isn't a law at all. That takes us to the Normal Curve, which helps us learn what pollsters mean by "margin of error" and what scientists mean by "statistical significance."
With these covered, you'll know just about all the statistics you need to understand the modern world.
*This class will mostly be lecture, with some class exercises and with some optional homework.
