Naked Statistics: A (Re)Introduction

"It is not the place to learn for the first time about medians and means, but definitely the place to remember what you were once supposed to know... [and] why you were supposed to know them." - New York Times

In Naked Statistics, Mr. Wheelan does an excellent job of introducing (or in my case, _re_introducing) key statistical ideas. If you have already gone through statistics in college, and are just looking for a refresher on how to think statistically but in a much more enjoyable and approachable way, then I think this book is exactly what you're looking for.

It is filled with excellent examples in the form of stories and anecdotes on topics ranging from:

• Should we keep stats on teachers, like we do with pro athletes?
• How does Netflix know what movies I like?
• How can we confidently know a particular drug is effective at treating things like autism?
• How many terrorist-related deaths have we had in the US after 9/11? (think: did the average number of automobile deaths increase due to fear of flying?)

The Appropriate Balance of Insight and Caution

Perhaps even more importantly, the author does an excellent job of constantly reminding the reader of the common pitfalls with data and data collection, as well as the common fallacies and misapplications. He makes it easy to get excited about the possibilities of analytical work in the age of big data, all while maintaining the appropriate level of respect and caution for such a powerful tool.

Anyone with a laptop has access to immensely powerful statistical software packages, but this should come with a warning: using these tools is the easy part. The hard part of statistics is in the analytical mindset and discipline that can only come from years of hard work (he uses the analogy of street hockey vs playing in the NHL to illustrate the difference between these two uses).

Too Elementary?

Some reviewers criticized the book for being too elementary, and perhaps they are right for those readers with regular exposure to statistical processes. But I think that misses the point, and the intended audience.

My only gripe was that the book did hit a slow point for me somewhere around chapter 2, after the attention-getting intro and before the probability coverage. However, I urge readers to push through whatever similar obstacles they hit along the way. The payoff is worth it.

Conclusion

An excellent book, by an excellent writer with a gift for making the banal and otherwise-dry topics interesting (I highly recommend his previous book, Naked Economics as well).

I came to this book as I entertain the idea of enrolling in graduate school for computer science. We are currently in the age of Big Data, and the world needs books like this to remind us of the immense raw power of statistics, both the positive and negative aspects. I also found it very fitting that in the authors own Conclusion, he decides to "finish the book with questions, rather than answers."

Many of these questions and topics sound similar to something you'd find in books like Freaknomics, or those by Malcolm Gladwell. Imagine content as entertaining as those books, but all while you're learning serious concepts such as Probability, Central Limit Theorem, Regression Analysis, etc.

And to drive his final point home, he reminds us that

"...formulas will not tell us which uses of the data are appropriate and which are not. Math cannot supplant judgement."

Well said.