Monday, May 4, 2009
Think before you test
I recently read “Usability Evaluation Considered Harmful (Some of the Time)”, a CHI ’08 paper by Saul Greenberg and Bill Buxton. The paper describes how, while it is not a written rule, it is an understood rule that CHI papers must have a user study to be considered. Because of this, standard usability studies are blindly applied such that the results are sometimes meaningless and other times harmful. For example, if the standard test results for a new interface show that the new interface is slower than old ones, then the new interface might be abandoned before it has time to mature. Tests can still be useful if they are well thought out and applied in appropriate situations. However, as things are now, too many ideas are being killed off too soon.
Subscribe to:
Post Comments (Atom)
I agree, the tests themselves arn't really at fault, but rather what the researchers do based on their results. The right time to perform those tests is also an important factor.
ReplyDeleteI agree that trial and error is not a valid method of design. If usability testing is given too much credit it will encourage this kind of reckless development.
ReplyDelete