CONTROLLING variables:

making sure the only thing that changes in your test is what you plan on measuring.

MULTIPLE trials:

making sure that you don’t base your thinking on too few events.

OBJECTIVE measurement:

making sure that you measure using a system that others can use in the same way.