REDUCTION OF MEASUREMENT SYSTEM VARIATION USING THE DISCRIMINATION RATIO

The standard deviation of Test-Retest error provides an absolute measu re of the error introduced by the measurement system. It does not, how ever, characterize the relative usefulness of the measurement system f or measuring a specific product. Such a measure is of particular inter est in a production environment. To determine the relative usefulness of a measurement system, one must have an estimate of product variatio n and a method for comparing Test-Retest error with product variation. In this paper we will discuss the use of the Discrimination Ratio to characterize and evaluate the relative usefulness of a given measureme nt system in a production environment. In particular, we will calculat e the Discrimination Ratio for a 4-point probe system and investigate the effect of probe head tip radii on the overall system variation. Re sults indicate that each implant recipe has an optimum four point prob e head as determined by the maximization of the Discrimination Ratio. Other four point probe heads could be used if larger system variations are tolerated. The Discrimination Ratio was found to be a critical pa rameter when interpreting results to minimize measurement system varia tions.