Statistical Quality Control

Statistical Quality Control  
Name STAT 521
Affiliation Instructor Date
Statistical Quality Control  
Researchers are increasingly focusing on ways to implement nonparametric control charts to control and monitor various distribution parameters. Chakraborti & Graham (2019) focus on an extensive analysis of nonparametric (distribution-free) control charts and provided updated reviews of their previous articles on these types of control charts. The researchers analyzed the Shewhart, cumulative sum (CUSUM), and exponentially weighted moving average (EWMA) charts. However, in their discussions, the researchers primarily focus on Phase I and Phase II distribution-free control charts. They also pose suggestions for further research on nonparametric control charts. Reviews of research on nonparametric control charts help address the current state of studies on control charts and influencing further research and publications.
Nonparametric control charts are useful as process observations are not necessarily normal on parametric distribution. Chakraborti & Graham (2019) first reviewed the univariate nonparametric charts and multivariate charts. However, in the past papers, they focused only on the univariate nonparametric control chart and in the in-control run length distribution like other researchers. Chakraborti & Graham (2019) focused on the effectiveness of nonparametric control charts, both in the univariate and multivariate cases. Multivariate control charts are based on the concept of the data depth approach used to monitor how the study of the variables can be interpreted as control charts, just as the classic control charts. These types of charts and graphs detect changes throughout the process, just like in the case of univariate. The multivariate graphs that do not require normality in the data are advantageous than those requiring data normality.
The researchers highlighted that the Shewhart-type control charts are popular because of their ease of use and efficiency. The Shewhart charts are common in Phase I and Phase II analysis when monitoring the distribution of processes. While most nonparametric control charts are applied in Phase II analysis, Phase I analysis is important and common as the charts are simple and can detect large shifts (Chakraborti & Graham, 2019). At Phase I, the control charting problems are similar to the test of homogeneity (k-sample problem) (Chakraborti & Graham, 2019). Phase II charts of the Univariate nonparametric process monitoring are increasingly common. Adapting the two-sample nonparametric tests is useful when constructing the Phase II control charts.