Abstrait

DEMONSTRATING THE INVARIANCE PROPERTY OF HOTELLING T2 STATISTIC

Sani Salihu Abubakar, Abubakar Yahaya

Numerous control chart procedures were developed and their widespread applicability as vital tools in statistical process control cannot be overemphasized. Under normal circumstances, process monitoring and control should include several interrelated process variables. The Hotelling T2 statistic being one of the most popular control chart procedures can be utilized to detect whether the observed disturbance in a given process is as a result of some perturbations in the mean vector of the quality characteristics or in their variability. One of the most notable features of Hotelling T2 statistic is its invariance property. This article demonstrates this important property using a real dataset obtained from a well-known cement production company in Nigeria.