Three programs are seeded with faults and then rigorously tested. The 
percentage of seeded faults detected by each test set and the percentage
decisions exercised by each test are recorded.

Each observation consists of the following variables: 
a) Fault Detection Effectiveness (FDE)
b) Decision Coverage (DC)
c) Program - CAL, COL, SORT

In this case consider FDE to be the dependent variable and DC the 
independent variable.