An Empirical study on application of Wilcoxon signed rank test in Statistical Process control when process target is fixed

Many statistical analyses directly or indirectly assume that the population from which the sample was drawn is normally distributed. However there are many cases in which the process distribution is skewed, and such that the normality assumption is not valid. If the process distribution is skewed, the false alarm rate grows larger as the skewness increases. In Statistical Process control also if data do not came from normally distributed population then the use of conventional Shewhart chart is inadequate. Whenever our process target is pre-fixed, for example in packaging industries, we can use Wilcoxon signed rank test statistic. Using this we can try to detect shift from hypothised (assumed) value. We used empirical data to check the successful applicability of Wilcoxon signed rank statistic.