Stop Guessing, Start Testing

Observing users tells you what is happening, but it never tells you why.

That is the massive difference between just watching people and running actual experimental research.

Experiments are the only tool we have to prove cause and effect. It works like a

criminal trial. You start with a Null Hypothesis, assuming your new design changes nothing.

Then you manipulate the Independent Variable, like a menu style, and measure the

Dependent Variable, like speed. But you have to be careful not to convict an innocent

design or let a bad one walk free. Stop guessing and start testing.