Headline! Radio buttons originally controlled radios

If you’ve read About Face, you know about how to use radio buttons. But if you haven’t read Tog on Interface, you don’t know how they were invented, or why they’re called that. You really don’t need to, but it’s damn interesting, and while the book is outdated, Tog is funny as hell. Actually, it may be useful too; a friend just wrote to ask, ‘Is it radial or radio?‘ The reason it’s radio is because it mimics the behavior of those old analog car radios. They had a row of buttons to select a station, and pushing one in moved the analog needle to the preset frequency. Of course you can only select one station at a time, hence they’re a good metaphor for when you need a mutually-exclusive UI select widget. Also, you can plainly see which choice is selected and all the choices at a glance.

Hmmmm, I feel old.

Cost and Style

My post on eBay-as-Flea Market received a bit of attention, including — judging by the referers — some folks from eBay. Later discussions with Tanya and Owen refined these ideas a bit, namely:

  • I was a bit sloppy in my use of the word design. eBay’s design works, though the style of the site — by reflecting the home-spun vernacular of sellers — can be low style
  • Low style is not the same as clean style or bad style. Low is the vernacular, clean can be a default look, and bad actually works against the design. Someone somewhere probably has expressed this better with different terms, but lacking that knowledge I’m running with it.
  • Low style and low cost — giving the perception of a Flea Market — can be good. These are not value judgments, but judgments of value.
  • Organization and classification is often secondary to other factors that communicate value, such as style and cost.

  • And so on. The interplay of style and cost most interested me, so I created a little matrix with pretty pictures:

    It’s a bit rough, but could turn into a fun little tool to clarify product design and marketing (‘Which quadrant are you in? They’ll all good, but different.‘). Might be nice to create a book of them, including ye ol’ value-complexity matrix.

Litmus test for scent/meaning

I’ve plowed through the research on information scent, and while they seem to be learning something about how people think about links and navigation, it’s not clear if there’s anything actionable for designers to take away from it. So I continue to think about how to create scent, or really, how to make links meaningful to people. I think some of Spool’s research on link length is helpful.

Recently I’ve been playing with what I call the Litmus Test for Scent. It helps to quickly judge whether link text is effective in an appeal-to-common-sense way. Here’s the format:

OH! I really want to see what’s in [link name]!

So, for example, when I go to Amazon and see that little tab with my name on it, I say, “OH! I really want to see what’s on Victor’s page!

but not

OH! I really want to see what’s in Click here” or “OH! I really want to see what’s in [insider lingo]” etc.

Shifting information goals

Peterme nicely illustrates shifting goals in the information seeking process — complete with screen shots — revealing the complexity of navigation design: ‘My original “goal” was to learn about Ann Willoughby. On reading that page about Ann, my goal shifted…. Shifting and evolving goals are not only common — they are the norm.



Someone at work recently pointed out that the retail theory people, like the folks at Vanderbilt’s eLab, tried pinning navigation to the measurement of flow. It’s another example of a highly dynamic environment (psychology + HCI + lots of information) that makes simplistic approaches to navigation design look, well, simplistic.



I took the first step to addressing this with my navigation design research and method. It needs a lot of work, but I think it approaches something robust enough to acknowledge the complexity of navigation design yet accessible enough to use without interfering with the design work itself.

Theory: EBay as Flea Market

Let’s assume EBay looks the way it does (not great) because not a lot of attention was paid to the design. Now let’s say they had contracted the design to a professional services firm that practices user-centered design. What would the result look like? Most likely something pretty slick.

Conventional wisdom – at least with the folks I hang out with – says that auctions, plus EBay’s first-mover advantage – is such a compelling experience that people will tolerate the bad design. But what if EBay is succeeding because of its bad design? What if, like a flea market’s rough, seller-created environment, the amateur design communicates the idea of bargain?

A designer might have come to this conclusion – balancing some good global elements like navigation with lots of seller-created pages, letting the vernacular bubble up, however painful to look at – but maybe not. And even if the designer was to hit upon this idea, how hard it would be to sell, or even to think about selling, a poor looking design to the client.

The experiment to switch the quality of the design could certainly be run, and will be if EBay ever grabs the reigns in a big way and puts a pretty design in action. If popularity declined as a result, that would be quite a big insight into experience design.

Update: Sparked thoughts from Tanya, Gene, and Jason.

Compare Products

A quick comparison of ‘compare products’ functions, I’m particularly interested in the use of screen space and comparing several products. Yahoo shopping will let you compare a seemingly unlimited number of products in its columnar layout, so the visual scanning is done by horizontally scrolling. Same deal at NexTag and EspressoPeople. Dealtime also lets you select many items but paginates the results, five per page. Bizrate uses a pop-up alert when you try to select more than three items. MySimon compares prices, not features, and uses rows instead of columns. Canon UK offers three pull-down menus displayed horizontally and each column is populated dynamically (neat, if limited). Turbo Tourney just displays their three products, no interaction needed.

Holy crap, check out the Yahoo Shopping SmartSort. The interaction design – especially the most/least important radio buttons – is a little confusing, but it’s fast and useful. Neatest of all is the copy for each of the items, which changes dynamically, so it can say things a person would say, like

Toshiba Portege 2010 is a subcompact notebook. It is ranked first because it has the best Screen Size compared to the others in your top 10 results. This Notebook Computer is cheaper than Toshiba Portege R100 which is displayed next.

Algorithmically, that’s

Toshiba Portege 2010 is a [productType]. It is ranked [rank] because it has the best [strongestFeatureByUserWeight] compared to the others in your top [resultSetAmount] results. This Notebook Computer is cheaper than [nameOfNextItem] which is displayed next.

Click one of those radio buttons and the results resort and the copy changes accordingly. I’d love to see the look on the programmer’s face when the designer proposed this.

Published
Categorized as Unfiled