The only way to test a particular variable for sure is to leave all other variables constant, and if a change in ONE variable creates an overall change, then you know that that variable caused it. But if you change 2 variables, and something changes, you don't know which one changed it for sure.