Think of it this way: If you test only one variable then you know that the difference in the experimental and control setup is that one independent variable. If you test more than one you will not know which one made the difference. Hope this helps :)