Usability (UX) researchers who want to apply classical statistics to their everyday decision making will benefit from reading this how-to guide. UX researchers learn how to make data-based decisions, such as for moderated usability studies, unmoderated usability studies, and surveys. The techniques, principles, and processes discussed in the book can be extended to other forms of UX research, including focus groups, live website analytics, and competitive research, which are not covered in the book directly.
The book focuses on enabling the reader to actually apply the statistical methods discussed by applying software to example case studies. Two software packages, Microsoft Excel and Statistical Package for the Social Sciences (SPSS), are used to illustrate the application of statistical techniques and show the resulting outputs applied to the data. Excel is easily available to readers. Although Excel is not a statistical package, it has a lot of statistical capabilities and can be supplemented with add-ons for advanced techniques that it may not have. Of course, SPSS is a familiar friend to many users of statistics. The book liberally uses screen shots of both Excel and SPSS to walk the reader through entering data, applying the appropriate statistical technique, and obtaining results for each case study. Although not mandatory, a reader may actually enter data (in those cases where the number of data points is relatively small) in one of the software applications and thus enrich the learning experience by actually doing while following along with the reading.
The use of clearly marked sidebars is frequent, judicious, and invaluable. Each conveys a significant point and introduces additional knowledge that can prove valuable to the reader. For example, “Sample Size: How Many Participants Do You Need for a Usability Test?” goes into background and logic that shows that even a small sample size of eight participants can be satisfactory.
The writing is clear and lucid. The authors make highly technical content, including the use of necessary equations, readable and understandable, and also enable the reader to apply the discussed techniques.
The book divides into 11 chapters and builds from fundamental concepts (which are always worth reviewing) to more advanced concepts. Chapter 1 is an introduction to a variety of statistical ideas and techniques, including the basics of the normal curve, confidence intervals, and hypothesis testing.
The next two chapters introduce the use of t-tests to determine if the means of two groups are significantly different from one another. Chapter 2 compares two designs (or anything else according to the authors) using independent sample t-tests. Starting with this chapter, a different case study serves as the basis for the discussion and illustration of how to apply Excel or SPSS to the case study data. Chapter 3 compares two designs using paired sample t-tests. This adds the wrinkle of “paired” versus “independent” samples of the previous chapter.
Chapters 4 and 5 essentially move on to hypothesis testing. Chapter 4 covers binomial-related hypothesis testing and confidence intervals using independent samples. Chapter 5 extends the approach of chapter 4 from independent to paired samples.
The next three chapters cover analysis of variance (ANOVA). Chapter 6 compares more than two means using one-factor ANOVA with independent samples and multiple comparison testing with the Newman-Keuls test. Chapter 7 goes on to more than two means using one-factor ANOVA with a within-subject design. Chapter 8 extends the discussion of comparing more than two means to two-factor ANOVA with independent samples.
The final three chapters discuss the use of regression analysis. Chapter 9 gets into correlation and the use of simple linear regression. Chapter 10 expands the regression analysis discussion to multiple linear regression and stepwise regression. Chapter 11, the concluding chapter, discusses logistic regression.
Beginning UX researchers should find the book absorbing and will be able to make practical their academic training in classical statistics. Experienced professionals should welcome a chance to review how the authors approach each topic, and can refresh and renew their understanding of what needs to be done for practical decision making. Even the most senior UX researchers should be able to glean valuable insights from the book. In other words, all UX researchers, no matter their experience level, should find this book valuable and so it is recommended.