Presentation on theme: "Usability Testing: Easier is Better Cokie Anderson Associate Professor Oklahoma State University"— Presentation transcript:
Usability Testing: Easier is Better Cokie Anderson Associate Professor Oklahoma State University
“Traditional” Usability Testing Long, Complicated, Expensive Process Designing tests Recruiting “representative” participants Setting up facility for tests Convening focus groups Analyzing responses Amending website Writing reports Do it all over again
Drawbacks to this approach Difficult to design and implement Books on subject not user-friendly May require professional assistance Expensive Time consuming Requires months to design, administer, analyze results, revise website and repeat test Labor intensive Results may or may not be accurate
What if you don’t have time and $$$? Use the “common sense” approach: Don’t Make Me Think (2 nd ed.) by Steve Krug Berkeley: New Riders Press (2006) Author is respected usability consultant with over 20 years experience The book is very usable (a good sign the person knows what he is talking about) Short, easy to read, well designed Lays out the facts and myths of usability testing
The OSU Electronic Publishing Center’s experience The EPC is publishing the online version of the NEH-funded Encyclopedia of Oklahoma History and Culture. Usability testing one of the requirements of the grant. First attempt at testing, we tried the traditional approach. Hard to manage, time consuming, difficult Results not helpful; no redesign resulted
First usability testing attempt Recruited subjects from Library’s Bibliographic Instruction courses Gave participants tasks—entries to find via searches and browsing Designed feedback forms, where subjects were asked to rank how easy it was to find information on the site Timed subjects while they were completing tasks, interviewed them after test
First usability testing results Most feedback was “middle of the road” No one added comments on the form; in interviews they mostly commented on lack of images/ bells & whistles We spent three months on testing and obtained no helpful results Problems with the test: Our inexperience likely resulted in bad design Only limited entries completed, so only limited tasks could be assigned Used “convenience sample”
Usability Test Take 2: The Common Sense Approach Using the recommendations in Don’t Make Me Think, we selected a few people we knew well and could trust to tell us the truth, however brutal. We sat down with these people while they explored the website, encouraging them to think out loud as they went. The testers’ observations and complaints led to a total overhaul of the site design.
Usability testing on 10p a day (To paraphrase Steve Krug) Grab 3 or 4 “reasonably patient” people who use the Web. Offer them a “reasonable incentive” to participate, i.e. small stipend. Sit them down in front of a computer in any office or conference room. Show them the website and ask what they see. (Krug gives a sample script). Video record the session if possible for viewing by design team. Only one person other than the subject should be present during testing. Test early and often.
Encyclopedia site before 2nd usability test (As it was before AND after 1 st usability test)
Encyclopedia site after 2nd usability test
Encyclopedia browse page
Krug’s Usability Principles 1. Don’t make the user think! 2. Make the site as self-evident as possible, because: People don’t read; they scan. People don’t choose the best option; they take the first acceptable option (satisfice). People don’t (won’t) try to figure out how a site works; they’ll just muddle through. 3. Follow conventions. Make it as mindless as possible.
Krug’s “things that must die” Needless words Introductions, welcomes, project histories, and other “happy talk” Instructions No one is going to read them—they’re just going to muddle through. Instead, find a way to make it obvious what needs to be done Focus groups Not the actual people—the concept. Focus group ≠ Usability test
It’s not rocket surgery™ — Steve Krug’s corporate motto Usability testing does not have to be complex to be effective. It’s better to do frequent, “10p a day” usability testing rather than to avoid usability testing because you’re concerned about the time, expense, and difficulty. Easier can be better.
Bibliography Available from for £ for $21.58 German edition: €24,95
Questions? Cokie Anderson Associate Professor & Librarian Director, Electronic Publishing Center Oklahoma State University Stillwater, Oklahoma USA