Anyone who can make the central limit theorem and standard error relevant and interesting is doing the Lord’s work. Although seemingly impossible, Charles Wheelan has achieved the feat in Naked Statistics: Stripping the Dread from the Data (300 pages, 2013).
Wheelan’s strategy is to focus not on elegant formulas but on the underlying logic and relevance of statistics. He recounts how he found calculus and statistics dull in school, at least initially, because he could not understand the point of it all. His chief mission then is to inject provocative and insightful statistical tidbits—and boy does the strategy work.
Let’s be serious: how many people look forward to reading a book on statistics? Without a nudge from my MBA program, I would never have heard of or picked up this one. Yet I found myself enjoying the discussion and countless clever stories. Naked Statistics got me sharing insights with friends and applying correlation coefficients to an Astros baseball game in Houston.
The picking up was figurative, since I listened to the audio version. Its accessibility demonstrates Wheelan managed to transfer the most important elements of statistics into words and not numbers. Even with this constraint, his precision has garnered praise from practitioners. Tomi Mester, founder of Data36.com, says it is a must-read, “even if you know everything about statistics. It just makes things clear.”
The beauty of the digital era is that we can access the most brilliant and compelling thinkers with ease. Eager learners bypass the bored teachers we all knew growing up, who taught to the test or little at all. Meanwhile, Wheelan distills the best of statistics and offers it for mass consumption, application, and—gasp—enjoyment.
In some ways, Naked Statistics is a massive open online course (MOOC), a Statistics 101. It has all you need to know short of the tools for professional data analysis. Further, the nerds among us, who want to dive deeper, can turn immediately to online videos for deeper explanations (from the likes of Mester) of concepts mentioned but beyond the scope of the book.
Naked Statistics has two slight weaknesses. First, the content is heavily US-centric, so likely some of the stories will be less relatable to foreign readers. Second, although Wheelan carefully warns readers away from conclusions about causality, he slips up at times and makes inferences of his own.
In doing so, Wheelan exhibits the temptation media outlets face, since unexpected conclusions are more exciting than null hypotheses. His injection of a few personal opinions is restrained, though, and adds color to the mix.
His prose also makes for a rounded, personable book that invites readers to apply the lessons and think for themselves. These lessons are increasingly important in a world of rapidly expanding and often invasive data collection. This development raises moral challenges, whether we want to face them or not, and Wheelan does not shy away from bringing them to the reader’s attention.