Tutorial 28

Intuitive Statistics for HCI Professionals: Developing Understanding and Avoiding Errors

Jeff Johnson, Robin Jeffries, Sun Microsystems

Monday, May 8

Objective

After completing this tutorial, participants will (1) know why statistics are used, and when to use them, (2) have an intuitive grasp of statistical concepts such as probability and statistical power, (3) understand how to design simple but convincing experiments, (4) have an understanding of a few statistical tests that are useful for HCI practitioners, (5) be able to evaluate methods used by others, and (6) be a more critical reader of results reported in the popular press.

Content

This tutorial tries a new approach to teaching statistics: building intuition and common-sense understanding. The tutorial begins with simple probability using concrete examples, moves on to what it means to take measurements, and discusses what it takes to convince others about your results. The tutorial proceeds to statistical tests. The tutorial also describes common statistical errors, using examples from the popular press and the HCI literature. Tutorial participants will take and summarize measurements, design hypothetical experiments, plan the corresponding data analyses, and evaluate experimental analyses reported by others.

Audience

This introductory-level tutorial is for HCI professionals who conduct usability tests or other empirical studies. It is also for people who have never taken a statistics course or would like to add to their knowledge.

Presentation

Multimedia, exercises

Instructors

Jeff Johnson has worked in the field of human- computer interaction since 1978. He designed, co-implemented, and wrote the manual for an early statistical package for PCs. Robin Jeffries is User Interface Architect for the Developer Products Division of Sun Microsystems. She has published in a variety of areas including comparisons of user interface evaluation methods and cognitive studies of computer programmers. She has studied and taught statistics both as a statistician and a psychologist.

Keith Instone / instone@acm.org / 95-01-05