You want to know the effect of a new drug on an individual's blood pressure. You randomize people into two group: control group and treatment group. The treatment group is given the drug to take at home and the control group is not given any drug. When the people report back the next day, you measure their blood pressure and decide that the difference in average outcomes of treatment and control is the casual effect of the drug. Why is this experiment not ideal? What mistake is the researcher making?