Why is it advantageous for scientist to test only one variable at a time during an experiment?

If they test more than one variable, they wouldn't know if it was both variables or one variable that was responsible for any observed effect.