Abstract
Process capability indices are widely used in the industry as part of TQM practices. The reliability of indices like Cp and Cpk depends on the assumption of a normal distribution that underlies the production process. Departures from normality, especially skewness and kurtosis, can lead to situations where a process with a high Cp such as 1.5, can yield lower levels of quality than expected. In this paper, we study the effects of departure from normality (measured in terms of third and fourth central moments) on process capability indices and estimate the difference in quality levels to be expected on account of such departures from normality.