Abstract

The translation of sigma quality levels into number of defects depends on whether the process is centered. Motorola Corporation has popularized the linking of six-sigma quality with the rate of 3.4 defects per million parts. If a dimension of a product is normally distributed and if the target value for the dimension is identical to (or centered) at the mean of the values, then the sigma level is inversely proportional to the standard deviation from the mean. An example shows how six-sigma corresponds to 2 defects per billion parts. The discrepancy with the Motorola data is caused by Motorola placing the process mean 1.5 standard deviations off center. It may be unwise to plan for a process to drift so far from the target. Those who implement six-sigma initiatives ought to be aware of what this concept means.