Hail to the usability test

Things might have been different this weekend if ballots were treated like Web sites.

It sure was confusing getting to Inauguration Day. Ask Doris Brooks, a self-proclaimed "nifty-spiffy granny, somewhere north of 50 and south of 100," who lives in Palm Beach County, Fla.

She's as sharp as a tack, yet voting in the presidential election caused her pain. It was so confusing, she says, she almost voted for Pat Buchanan rather than Al Gore. After 45 minutes on hold, she complained loud and long to the Palm Beach County registrar of voters. She wasn't alone. The phones there rang off the hook all day.

You can blame the ensuing presidential mess on faulty hand recounts, postmark-free absentee ballots or hanging chads. But I believe Doris' -- and the nation's -- confusion could have been avoided with some simple "usability" testing.

Usability testing is the little-known science of testing products, software and Web sites to see if an average person can figure out how to use them.

Consider VCRs. Most are so perplexing we can't even set the time and date. But it is possible to design a VCR that's easy to program and configure. That's where usability testing comes in.

You've probably seen a focus group portrayed, or perhaps been in one yourself. A moderator leads a group of "real people" through a directed discussion, while marketing types observe behind a one-way mirror. Usability testing works much the same way. A single person is led through a series of tasks involving the device, Web site or software program under analysis as cameras record every move and product developers watch through the two-way mirror.

Usability testing started in the military during World War II with the goal of developing a simpler cockpit. The results helped newly minted pilots learn quickly to fly B-17s, P-51 Mustangs and other planes, helping us win the war. But usability remained a backwater until the late 1980s, when more people started using personal computers -- and found software more incomprehensible than a Sopwith Camel. That's when Microsoft, Apple and Lotus turned to usability testing to improve their products.

Sure, Microsoft Word and Excel are still complex, but they are much, much better than in 1987. Did you shop at Amazon.com this past holiday season? Thanks to usability testing, it's one of the easiest sites on the Web, incorporating features like "buy now with 1-Click" and a simple-to-navigate interface. The latest version of Netscape's browser, 6.0, benefits tremendously from usability testing. Its stripped-down interface is limited to simple navigation and search.

Jakob Nielsen, one of the world's top usability scientists, likens ballots to a Web site. And judging by what he saw on TV, Palm Beach County's ballots were awful: "Anyone trained in usability would know it's an obviously bad design. You wouldn't even have to test it."

It takes training to develop a useful ballot, just as it takes training to design a decent Web site. Nielsen recommends design principles, such as a clear link between the instruction and the action item (or between the candidate's name and the punch-hole area), removing clutter, and regularly testing designs to ensure that they work. A usability review board for ballots could help make sure errors don't thwart the will of the people. Sure, this involves research, "but we have four years to do it," Nielsen notes.

Not just computers and ballots are usability-challenged. Poor usability is everywhere. Look at the cars we drive: In some models of Saab -- generally an extremely easy car to use -- the makers have surprisingly put the ignition switch and window-control buttons between the front seats. Or take office equipment. I'm a certified geek, yet I've never been able to transfer a call from my phone at work. Copiers are still a nightmare: Even though sorting and stapling are easier than in the past, you still need a degree in copy-ology to get the paper stapled in the upper-right-hand corner. And how many times have you stood before the office coffee machine wondering how to turn the darn thing on?

Poor design makes us feel like idiots. We blame ourselves. But Nielsen gets us off the hook: "If many people make the same mistake, the conclusion is not that people are stupid but that the

design is wrong." And you can't change people. We're a constant, according to Nielsen. Designers typically try to fix people, with training or unfathomable manuals. But it never works.

So next time you're confused by a contraption, baffled by a ballot or irritated by the Internet, remember it's not your fault. Blame it on the designers and encourage them to seek help. If half the money spent on focus groups went instead to usability testing, the world would be a much less frustrating place.

Contributing Editor Jim Louderback is editorial director of techtv.

Illustration by PHILIP BROOKER for USA WEEKEND