Before I get started, let me say I’m not a designer offended by usability testing thinking it takes something away from my “art”. (I’ve done the art thing in a previous life) This post is about poor thinking taking away from what could be interesting usability findings.
Jacob Neilson it’s fair to say is Web 1.0, much like IE6 he made a significant contribution but has possibly not moved quite enough with the times. And much like everybody’s least favourite browser he is also still used be a lot of people. So I keep up to date on what he says in his Alert Box columns; sooner of later someone will quote them at me (and to be fair they often do have nuggets of decent information).
The previous Alert Box was quite interesting. It examined how people behave when scrolling and had some eye-tracking plots to illustrate it’s points. The takeaway message was also useful – you can encourage people to scroll by designing the page to look like scrolling will reveal something interesting to the user. Fine. That’s all good, if not revolutionary, stuff.
But the latest column really has really irritated me. The article referers to the same eye-tracking data used in the previous column but this time looking at how people look at the screen in a left/right axis rather than vertical scrolling behaviour. The jist of the article is something like this:
- We showed people some websites that use a conventional layout.
- People looked at these websites in the way that they look at websites that use a conventional layout.
- We’ve just shown that people look at (all) websites as in the way they look websites using a conventional layout.
- Conventional layouts are the best way to design websites because that’s how people look at them.
The thinking presented in it is simply a usability tautology: people looking at websites using conventional layouts look that them in the way they look at websites with conventional layouts. It’s a complete failure of thinking that then leads to the conclusion that “you deviate from conventional layout at your peril”. To say that you need to show us a non-conventional layout suffering because people are looking in the wrong places.
Bad usability thinking like this is harmful to websites because people (managers, clients, amateurs, etc.) who don’t question it will resist the ability of designers to create solutions that are specific to the problems they are being asked to solve. They become a sort of half-known dogma trotted out at design meetings—like the “three clicks” rule and the “everything above the fold” ideal, I can see “conventional layouts are always better” being the next myth designers have to wade though before they can be creative.
Conventions in web design are still young and they are there for a reason but if you are going to use them, use them because they are good for your design not because of the false logic of bad usability articles.
