EWMA-type scheme for monitoring location parameter using auxiliary information. (December 2017)

• Record Type: Journal Article
• Authors: Adegoke, Nurudeen A.; Riaz, Muhammad; Sanusi, Ridwan A.; Smith, Adam N.H.; Pawley, Matthew D.M.
• Is Part Of: Computers & industrial engineering. Volume 114(2017)
• Page Start: 114
• Page End: 129
• Publication Date: 2017-12
• Subjects: Auxiliary variable -- Location parameter -- Ranked set sampling and median ranked set sampling -- Simple random sampling; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2017.10.013
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: The proposed charts monitor shift in the process variable using auxiliary variable. Efficiency increased when there is large negative correlation between the variables. The proposed charts are very efficient in detecting small shifts. Abstract: Control charts are statistical methods used to detect shifts in the location parameter of a process that is monitored over time. Here, we propose an improvement to the performance of the classical exponentially weighted moving average (EWMA) control charts, by making use of auxiliary information that is correlated with the process variable. We present an EWMA P and its modifications: EWMA-type control charts based on a product estimator where the location parameter of the process is monitored using an auxiliary variable. The charts are developed using different sampling schemes: simple random sampling, ranked set sampling (RSS) and median ranked set sampling (MRSS), and we evaluate their performance using average run length, and other performance measures such as extra quadratic loss and relative average run length. It is observed that the proposed control charts are performing better than the classical EWMA control chart in monitoring shifts in the location parameter of a process, especially when a strong negative correlation exists between the process and the auxiliary variables. They are particularly efficient in detecting small to moderate shifts in the process.