Dividing up subjects for analysis based on something other than the drug or treatment. This could be factors like age, gender, or other demographic information.