Want to reduce the likelihood of any preexisting differences between the groups of participants assigned to different conditions in an experiment you should use