Application of Statistical Techniques To Manufacturing Operations

A number of statistical techniques--among them analysis of variance, discriminant analysis, nonlinear regression, response-surface methodology, nonparametric statistics, bivariate normal analysis, and Grubb's method for estimating components of variability--can be used to advantage in the manufacturing environment. This paper describes the application of these techniques to selected plastics--molding and functional-testing operations in a large multiproduct organization. Specifically, it illustrates their value for identifying sources of variation, determining process and/or equipment capabilities, and optimizing production and inspection processes.