Autocorrelation's Effect on Process Capability Analysis

Abstract

The deviation of process output is usually estimated through the range or deviation of the sub‐samples in the calculation of process capability indices (PCI) in quality management. There are autocorrelation between the adjacent data in many practical process outputs. But this influence on PCI is not considered in process capability analysis. The autocorrelation of output data can be represented with time‐series model. Based on the several stationary stochastic process models of the time‐series analysis, the impact of the sampling method, sub‐sample size and deviation estimating methods in standard statistical process control are analyzed via simulation for the processes with autocorrelation. The autocorrelation testing of process should be undertaken in these cases. The proper analysis methods are set up in the paper.

Citation

Yang, J. (2009), "Autocorrelation's Effect on Process Capability Analysis", Asian Journal on Quality, Vol. 10 No. 3, pp. 65-72. https://doi.org/10.1108/15982680911021197