The answer to this question is the random assignment to control experimeter BIAS. This is the method or the only way in order to control for experimeter BIAS as well as confounding variables such as the severity condition, the attitudes, the age and etc., The researchers or the person who conduct subjects to groups by random assignment.