Which is all fine for me, but what has it got to do with you? Well, maybe I’m not the only one who has felt daunted by the vast choice of books on statistics and made a few false starts at reading them. So I thought it was worth picking out the ones that have helped me the most so far and sharing them with you, in the hope that you’ll also find them worthwhile.

Disclaimer—I’m not trying to provide a list of statistics books that is either comprehensive or definitive. If you’re a statistics expert or have found a better way into the subject, great! Please add a comment, and I’ll be delighted to follow up on it.

## The Basics: Getting Familiar with a Mean

Most introduction-to-statistics books seem to start with *mean, median,* and *mode:* the three most frequently encountered *measures of central tendency* and the three simplest *descriptive statistics*. If you are comfortable with all of these terms, skip ahead to the next section.

If you have already started feeling a touch alienated and confused, start with Darrell Huff’s *How to Lie with Statistics*. (Published in 1954 and widely available, in print or second hand, from about $5.)

I chose this book as my Survey Book of the Month for March 2011 because I’ve loved it ever since I bought my first copy as a teenager. It’s a witty guide to the basics such as

- the difference between a mean and a mode
- the concept that correlation does not equal causation
- how to work out whether a graph is misleading

Huff’s examples are redolent of the 1950s. If you’d prefer a more modern text or would like to go into similar topics in more depth, *Misused Statistics* by Herbert F. Spirer, Louise Spirer, and A. J. Jaffe, from 1998, covers much of the same ground, but it’s illustrated with many real examples the authors gleaned from serious newspapers and official documents.

## The Immediately Useful: Basic UX Statistics

The next step up is to use some statistics, perhaps *any* statistics, in our user experience work.

Tom Tullis and Bill Albert’s *Measuring the User Experience: Collecting, Analyzing, and Presenting Usability Metrics* has 10 pages that take you through the basics you’ll need in user experience:

- basic descriptive statistics
- comparing means
- correlations
- chi-squared tests

They describe typical UX problems and explain how to use the appropriate Excel 2007 functions—having published their book in 2008. From there, it’s not too hard to work out what to do in newer versions of Excel.

Bonus—The rest of the book is well worth reading. It has lots of ideas for UX metrics that you can collect and a selection of case studies that show how the ideas work in practice.

## To Pass Your Statistics Exam: Many Options

If you need to take a statistics class in college, there’s a simple answer to the problem of which book to choose, isn’t there? Yes, it’s whatever your instructor recommends. Or even better, whatever your examiner recommends.

I’ve amassed quite a large pile of getting-started-in-statistics books that are aimed squarely at this market—for example, Dawn Griffiths’s *Head First Statistics*. Depending on your point of view, you might describe it as a lively, fun book that introduces statistics concepts using amusing examples *or* as relentlessly and unbearably winsome. Either way, the book’s emphasis is squarely on helping you become familiar with statistical formulas in order to pass an examination.

Other options in the same category:

*Statistics for People Who (Think They) Hate Statistics*, by Neil J. Salkind*Statistics for Dummies, Statistics for Dummies II,*and*Statistics Workbook for Dummies*, by Deborah Rumsey*The Cartoon Guide to Statistics*, by Larry Gonick and Woollcott Smith, originally published in 1993 and reprinted in 2000

From the point of view of a UX professional who just wants to understand enough about statistics to use the formulas and tests on practical problems, I found that they all missed the point for me. There was

- too much emphasis on the details of the formulas
- not enough emphasis on when to pick which test or formula to use
- little or no acknowledgement that I’m going to use a computer in most of them

Salkind is an exception to that last point. My edition of his book gives instructions for using SPSS throughout—which is fine if you’re a student and your college gives you SPSS access, but not much help for the lone UX professional without SPSS or the $5000 necessary to purchase it. While checking my references for this article, I found out that he’s also written a version of this book that refers to Excel. I wish I’d known that before purchasing his book.

## To Get Familiar with the Tests Without Formulas

Fortunately, I found one statistics book aimed at college students that firmly relegates all of the formulas to the back, where they’re out of sight: Frances Clegg’s *Simple Statistics: A Course Book for the Social Sciences*.

Clegg covers mean, median, and mode in the usual way, but before getting into normal distributions, she takes you into rank and sign tests. This appealed to me because these tests do not require you to make as many assumptions about the data, and in particular, do not require the data to come from a normal distribution.

The *normal distribution* is the bell-curve distribution, which has many convenient mathematical properties and is rather common in nature—for example, heights of students. Unfortunately, a lot of UX data doesn’t happen to follow a normal distribution—for example, think of the long neck, or Zipf, distribution that you get for search log data.

If I lost you at normal distribution, I apologize. Ignore what I’m saying, and think about reading Clegg instead. Like Huff, the book has a lightness of tone that appeals to me.

Because the formulas are at the back, you don’t encounter them until you’ve understood why you might need one and how to use it. And even when you do turn to them, you’ll find that Clegg describes every formula in terms of a series of simple steps before you get to the full-on mathematical symbols. If you don’t want to grapple with Excel’s statistical functions, you could possibly just use Excel as a sort of calculator to help you through the individual steps as she lays them out.