Pugh, G. Allen (1989, ASQC) Purdue University, Fort Wayne, IN
Two of the most important modelling tools in the professional life of a statistical process control practitioner are the Gaussian (normal) distribution and the computer. Unfortunately, the Gaussian distribution has no closed form solution for finding the enclosed area; and hence the probability. This circumstance has led to the ubiquitous normal table. Such tables are not useful with computer applications.
There are two frequent demands for the Gaussian when computing. These are (1) determining the area under the curve (probability) or its inverse and (2) generating normal random deviates for Monte Carlo simulation. Several methods for accomplishing these are compared.