Summary

The presence of outliers and contaminations in the output of the process highly affects the performance of the design structures of commonly used control charts and hence makes them of less practical use. One of the solutions to deal with this problem is to use control charts which are robust against violations of the basic assumptions, like normality. The effect of using robust or other alternative estimators has not been investigated thoroughly in perspective of control charting literature. Some authors made their contributions in this foregoing concern. This thesis contributes to the same subject and develops different robust Shewhart-type and CUSUM-type control charts for process spread and process location.