How to stop your usability lab from feeling like one
Testing with real users is a fundamental part of a good design process. But most usability labs are clinical, artificial and cold. Here are three things you can do to make your lab more like the real world.
Make the environment feel like home
In their normal lives, people are much more likely to use the web sitting on a couch than a $1,500 Aeron chair. Your first goal should be to create an environment that feels more like someone’s living room.
Attending a user testing session can be intimidating, so you want people to feel as comfortable as possible during the sessions. But there’s also clear evidence that context has a big effect on memory and behavior. So an environment like this doesn’t just make people feel better, it leads to more accurate test results as well.
Test on hardware that reflects reality
Again, the closer you can get to the natural environment the better. The PC you use to test web based software should be as close as possible to the average. What exactly is the “average“ computer? You might be surprised.
The average computer (August 2014)
- Screen resolution: 1440 x 900 (source)
- Operating system: Windows 7 (source)
- Browser: Google Chrome (source)
- Form factor: Laptop (source)
These specs vary a bit depending on where you are (Navy is based in Australia so we have a slightly different setup).
Use the latest technology to observe, but keep it in the background
Just because your test PC is low tech doesn’t mean the rest of your setup needs to be. We use top of the line recording software with a remote observer in a separate room. We use web-cams with facial recognition to observe people’s reactions. We have a multi-directional microphone capturing audio.
But they key is to keep this technology in the background so that participants don’t get intimidated.
- The fastest computer you can find that meets the specs above.
- A broad range of common mobile devices (e.g. iPhone, iPad, Android phone, Android tablet etc.) — although beware Android fragmentation.
- Mr Tappy for attaching cameras to mobile devices.
- A webcam for recording the screen during mobile testing (Logitech’s C615works well and attaches easily to Mr Tappy).
- Another webcam for recording people’s faces during mobile testing (we use the Logitech QuickCam Orbit because it’s unobtrusive, can be “aimed” remotely, also includes facial recognition software).
- A microphone for recording audio. This is optional, and for mobile sessions we tend to use the microphone built in to the webcam.
- Morae for recording and analysing sessions (terrible interface, but otherwise a good product). It also lets you easily have an additional observer in another room, which can help minimise common biases like the observer-expectancy effect.
- Silverback is a cheaper alternative, but is limited to OSX only (problematic when testing desktop software or sites).
What about eye tracking?
This is a pretty complex issue that’s been covered well by other people, but on balance we’ve found the drawbacks of using this technology outweigh the benefits.
So you’ve got the lab, but what about the testing? We’ll be covering techniques for running usability sessions in upcoming articles.