This presentation will help UX practitioners use both descriptive and predictive statistical techniques to gain meaningful insight from data collected employing traditional UX research methods, including moderated usability studies, unmoderated usability studies, surveys and contextual inquiries.
Over the past five years, excellent books have been published regarding collecting, analyzing and presenting usability metrics. Arguably, the best ones in this category are the Morgan Kaufmann books, including Measuring the User Experience by Tom Tullis and Bill Albert, Beyond the Usability Lab by Bill Albert, Tom Tullis and Donna Tedesco, and Quantifying the User Experience by Jeff Sauro and James R. Lewis. These books do an outstanding job of instructing the UX professional how to choose the right metric, apply it and effectively use the information it reveals.
And yet, as we surveyed the UX research literature landscape, we saw there was very little discussion urging UX professionals to use predictive and other advanced statistical tools in their work. [The current books on usability metrics leave out the techniques often used for data analysis, such as multiple regression analysis.] But these statistical tools⎯which begin with basic correlation and regression analysis⎯are now fairly easy to access. In fact, if you have Excel, you probably have most of these tools already at your fingertips! In this presentation, we will attempt to:
- Show the real-world application of these techniques through some carefully chosen vignettes from the real world of UX design. By seeing parallels between the problems introduced and resolved in each chapter and your own work, the attendee should easily be able to ascertain the right statistical method to use in your particular UX research project.
- Provide clear insight into the statistical principles without getting bogged down in a theoretical treatise. But, we provide enough theory for you to understand why you’re applying a certain technique. After all, understanding why you’re doing something is just as important as knowing what you’re doing
- Minimize the amount of mathematical detail, while still doing full justice to the mathematical rigor of the presentation and the precision of our statements. In addition, many of our numerical examples use simple numbers.
- Focus on how to get the software to do the analysis. There are a few exceptions, in those cases where Excel does not provide a built-in solution, when we show you how to use other Excel commands to duplicate, as closely as possible, what Excel would do if the technique were a built-in available one. Also, we provide end-of-chapter exercises that encourage, demonstrate, and, indeed, require the use of the statistical software described. By the way, we do not apologize for writing our chapters in a way that does not insist that the reader understand what the software is doing under the hood!