Latching radio buttons

by Joe Gillespie — Dec 1, 2001

This interface uses the 'latching radio button' concept but presents it in a more modern way where electronic switching supersedes mechanical springs and levers. Real electronic switches usually have a 'click' effect to provide 'confirmation of operation' even though it is not technically necessary. 'Touch' switches that provide no feedback can be very disconcerting.

home
products
downloads
support
contact

This navigational device uses the principle of 'latching radio buttons' but in a less literal way. This demonstrates that you don't have to compromise creativity or use tired old visual clichés provided that the underlying principle is sound. This doesn't look like a set of radio buttons but gives exactly the same functionality.

In a Web page interface, it is better to suggest an element's function before it has been used. Then, when it has been established exactly what it does, you can reinforce the communication by providing visual feedback. That's the real purpose of a rollover or hover item – confirmation. When you say "Hello" down a telephone, you expect to hear "Hello" back, it confirms you are communicating. The initial screeching noise you hear from a modem or fax machine before they actually start communicating is called "handshaking". This is feedback too and highlights the importance of the principle of audibly clicking switches.

In a GUI, it is easy to represent objects with pictures. Some other concepts like 'up', 'down', 'forward' and 'back' are simple to communicate visually too. Many non-visual concepts, however, are difficult or even impossible to represent with icons or small graphic symbols. They are inappropriate yet somebody always decides that they want to use icons regardless of the fact that they don't work. When they discover that they can't make them work, they inevitably add the familiar crutch called a 'tooltip'. If you have to resort to explaining graphic element with words, styling is getting in the way of logic and you have failed. There must be a better answer and you have to stand back from the problem to recognise and fix it. If you are unable to stand back from a problem and look at it objectively, get somebody else to do it for you. That's what 'idiot testing' is all about.

The only way to really test an interface design is to stick it in front of an unfamiliar and unbiased audience and observe, preferably without them knowing that you are watching. In market research, this is known as a 'goldfish bowl', usually consisting of a room with a mirrored window or hidden closed-circuit television camera. The process might seem underhand but it is effective. You probably don't have access to such a facility, but you should try to emulate the situation by making sure that you don't pollute or otherwise influence your results.

The word often left missing from 'Interface Design' is 'Human'. We are providing a conduit through which a person can interact with an electronic machine. The machine should do all the work in the process, not the human. Machines are there to make life easier for us, not to create more challenges. Interface design is a significant part of the Web designer's job and has little to do with hand coding HTML or fancy graphics skills - it's about plain, common sense.

Del.icio.us Digg Technorati Blinklist Furl reddit Design Float