Constructing Hotelling’s T² control limits for different α levels of significance

Abstract

When the quality of a product is controlled by combinations of two or more variables, a high level decision qui-ckly determines the out of control signals, while simultaneously, evaluating the variable(s) responsible for the false alarm rate (α) levels of significance. In this connection, bootstrap algorithm for constructing Hotelling’s T² control limits was developed and implemented using Visual Basic Code to resolve such problem. The results obtained using industrial production process data proved that the bootstrap control limit at different α levels of significance performed better when compared with other existing methods. Akaike Information Criterion was also applied to identify the variable(s) responsible for the out of control signals.