Measuring usability is an important topic that very much needs to be discussed. Being part of the conversation usually requires reading peer-reviewed journal articles in HCI. This, in turn, requires knowledge of the techniques and the jargon of statistics and experimental design (the foundation of numerical precision in empirical disciplines). To help move the discussions along, I've attempted to provide some clarification and explanations on statistical concepts you'll encounter: p-values, power, and confidence intervals. The examples come right from papers published in HCI proceedings. I've kept the discussion high-level and left out formulae, as it is more important to…
You must be a member of SIGCHI, a subscriber to ACM's Digital Library, or an interactions subscriber to read the full text of this article.
GET ACCESSJoin ACM SIGCHI
In addition to all of the professional benefits of being a SIGCHI member, members get full access to interactions online content and receive the print version of the magazine bimonthly.
Subscribe to the ACM Digital Library
Get access to all interactions content online and the entire archive of ACM publications dating back to 1954. (Please check with your institution to see if it already has a subscription.)
Subscribe to interactions
Get full access to interactions online content and receive the print version of the magazine bimonthly.