I am using the State-Trait Anxiety Inventory (STAI) which gives values from 20-80. My experimental group had mean of 26.1, the control group had 29.6. (p=0.146). Would that be a reduction of 26.1 divided by 29.6 (88% of the value, a 12% reduction), or can we consider that since 20 is the floor, the experimental group scored 9.6 above the floor and the experimental group scored 6.1 above the floor, giving a 37% reduction?