What happens if more than one variable is changed in an experiment?

If a researcher fails to control the relevant variables in an experiment, you could end up changing a bunch of different variables in an experiment, and you wouldn't be able to tell which variable is responsible for the results of the experiment.