If you don’t know where to start, start small. Sit with a friend on your computer. Explain that you want them to complete the task using software and watch what they do.
It helps to keep quiet while they really work. Write it all. "John spent 15 seconds looking at the screen before acting. He moused over the top nav to see if it contained popup menus. He first clicked "About Us" even though it wasn't central to his task." Etc.
Then use the acquired knowledge to help you develop more complex tests. Tests with different users from different fields of knowledge. More challenging tasks and more.
Fold them. A webcam mounted on a monitor is a good way to capture where their eyes move. A VCR running 45 degrees over your shoulder is a good way to capture your vision. Bonus points if you can synchronize them in time. Do not worry if you cannot do all this. Do what you can do.
Do not plan your test as if it is the last one you will ever need, and you want to achieve its perfection. No perfect. The only thing that comes to perfection is many iterations and many repetitions. You can only come close to 100% certainty, as the number of tests approaches the number of actual users of your software. Usually no one comes close to this number, but everyone should try.
And don't forget to retest people after you have included the improvement that you saw. The same people, different people, everything is in order.
Do what you can do. Do not cry what you cannot do. Just cry, so you can check, but not done.
pcorcoran
source share