Suppose that the effect of consuming oatmeal on cholesterol level is thought to depend to some extent on two additional variables-gender and degree of fitness. For which of these two variables would it be easier to separate the data when exploring the relationship between oatmeal consumption and cholesterol level?