If we were to design an experiment to test the hypothesis that aspirin takes headaches away the independent variable would be blank and the independent variable would be