Document Type



In evaluating modeling quality, risk level and general quality management "6 sigma rules" are popular among statisticians and quality engineers today. The "6 sigma rules" of Gaussian distribution are very simple and elemental. If switching the working environment from the probability measure based to the uncertain measure based, the simple "6 sigma rules" will be no longer simple. In this paper, we investigate the problems when facing Liu's uncertain normal distribution which can only facilitate interval variance or standard deviation. Consequently, the way to define "uncertain 6 sigma rules" are described and discussed and thus those to practical applications.