Abstract
A combination of double sampling the chart and the variable sampling interval of the
chart (DSVSI
chart) increases the sensitivity in the small shift detection. A usual assumption of control charts is that the process observations are normally distributed. However, this assumption may not be true for some processes in the real industry. This paper presented the performance of DSVSI
chart under non-normality and compared it with the Shewhart
chart and the variable parameters
chart. The compared results show that the DSVSI
chart has the best performance in detecting small process mean shifts.