Influence of skewness and kurtosis on high sigma cause selecting control chart performance for bivariate cascade processes

Cause selecting control charts are used for monitoring two cascade production processes. They work on the principle of regression in which the regressor variate is adjusted for a response covariate. It is generally assumed that the regression models have an intercept term. In practice, the process data may belong to a regression model with or without an intercept term. When the two cascade processes are part of a high sigma environ, the geometric data requires a transformation to convert the process data into normality. In this work for a high sigma cascade process, the extent of normality achieved and measured in terms of skewness and kurtosis and its influence on lower control limit and Type II error respectively have been studied. The study is carried out within limits of four scenarios of the intercept models.