Sagot :
It depends on the case, but usually two variables should be in a controlled experiment.
you should do three variables so you can note the difference