Abstrait

DETECTING ASSIGNABLE SIGNALS via DECOMPOSITION OF 2 T STATISTIC

Sani Salihu Abubakar, Osebekwin Ebenezer Asiribo, Abubakar Yahaya, Hussaini Garba Dikko

The proliferations of the various means through which data can be accessed technologically via computers have led to a huge enthusiasm in the simultaneous monitoring of several correlated quality characteristics. Hotelling’s 2 T is a famous statistic that is so significant in identifying a shift that is due to the mean vector of a multivariate normal distribution. When there is signals in a multivariate control chart, it is a clear indication that special cause variation exist in the process, as such effort must be made to identified which variable (s) is/are responsible for the out of control condition. In this study, we have demonstrated the use of mason, Young and Tracy (MYT) approach in identifying non-random quality characteristics.