Some descriptions of FMEA suggest using ordinal scales from 1 to 10 for the severity, the occurrence, and the detection measures. Is it appropriate to multiply these numbers? Suggest a technique for comparing different risks that have been evaluated on these three measures.