Naked Statistics by Charles Wheelan | Book review
A new book strives to make statistics comprehensible.
Trying to convince the math-averse that the study of statistics is simpler than it seems is to wage a serious uphill battle. In his new book, Charles Wheelan—a senior lecturer at the University of Chicago’s Harris School of Public Policy—gives it a shot anyway, attempting to steer away from the dense arithmetic that snares most people during Statistics 101 and toward a reduced, common-sense understanding of how statistical models work and why they’re such enticing tools for anyone trying to prove a point.
Using everyday circumstances, Naked Statistics effectively illustrates the mathematical principles for data analysis. This reduces the amount of time we have to spend trudging through an algebraic thicket to comprehend them. But it still can’t entirely deliver on its promise to “strip away the dread.” Statistics relies on things like regression coefficients and the central limit theorem, which means the author can’t completely spare readers from complicated equations and formulas.
It’s hard to say whether this book will cause anyone uncomfortable with tricky math to do a 180. But Wheelan is trying to accomplish something besides just simplifying his subject matter. He seeks to show how truly fragile the process of building reliable statistical findings can be, and how they should always be viewed in proper context. Statistics deal in likelihoods and relationships, not absolutes; they are, therefore, no substitute for good judgment. This is an especially important point when considered alongside the vast numbers of political polls, test results and drug trials people are asked to evaluate every day.