In: Statistics and Probability
2. Suppose that the researchers are interested in the effect of class size on only female student test scores. They consider using the following method to reconstruct their treatment and control groups: select all female students who have been assigned to small size classes as their new treatment group and select all female students who have been assigned to normal size classes as their new control group. Is this method problematic?
Treatment Group: All female students who have been assigned to small size classes
Control Group: All female students who have been assigned to normal size classes
This method is problematic because of the following reason:
Let
Number of students in small size classes = 15
Number of students in normal size classes = 30
Consider a school having 5 small size classes and 2 normal size classes.
As per the given experimental design:
Treatment Group: All female students who have been assigned to small size classes : Total students = 5 X 15 = 75
Control Group: All female students who have been assigned to normal size classes : Total students= 2 X 30 = 60
We note that total students in Treatment Group:= 75 is greater than total students in Control Group:= 60, which is against our definition of Treatment Group as small size and Control Group as normal size.
To aviod this problem, we should change experimental procedure as follows:
Treatment Group:will have all students of a single small size class.
Treatment Group:will have all students of a single normal size class.