Transforming Non-Normal Data to Normality in Statistical Process Control

Youn-Min Chou, Alan M. Polansky, Robert Mason

Journal of Quality Technology · 1998 · 299 citations · 24 references

Concepts

TL;DR

Statistical process control assumes normally distributed quality characteristics, and non‑normal data can lead to serious errors, so transforming data to normality—often via the Johnson system—is a common remedy. The study aims to identify suitable transformations for non‑normal data by applying the sample quantile ratio together with the Shapiro‑Wilk test of normality. The authors use the sample quantile ratio and Shapiro‑Wilk test to determine the appropriate Johnson transformation for a given data set. Examples of fitting non‑normal SPC data are presented, and the effect of the Johnson transformation on an SPC procedure that estimates the population standard deviation is demonstrated.

Abstract

Quality characteristics analyzed in statistical process control (SPC) often are required to be normally distributed. This is true in many types of control charts and acceptance sampling plans, as well as in process capability studies. If a characteristic is not normally distributed, but normal-based techniques are used, serious errors can result. One approach to solving this problem is to transform the non-normal data to normality using the Johnson system of distributions. In this paper, we use the sample quantile ratio, in conjunction with the Shapiro-Wilk test of normality, to find a suitable transformation for non-normal data. Examples of fitting non-normal SPC data are presented and discussed. The effect of the Johnson transformation on an SPC procedure involving an estimator for the population standard deviation is studied using non-normal data and Johnson-transformed data.

References

24