What does normal or transformed-normal mean when referring to the distribution of the data?
The Shapiro Wilk normality test, for instance, tests the background values of a data set and statistically determines whether the distribution follows a normal (or bell-shaped curve) distribution. If the original data set fails the test, there are a series of transformations known as the Ladder of Powers that may be applied to determine whether any transformation helps fit the data to the bell-shaped curve. The Ladder of Powers was developed by Dennis Helsel and includes the following transformations: x, x1/2, x2, x1/3, x3, natural log, x4, x5, x6. If one of these transformations passes the normality test, all data are transformed prior to the construction of any limits. Unless plot transformed values is selected in the Configure Sanitas options window, Sanitas will back transform the final limit. In either case, it is always advisable to closely examine the statistical limit to confirm that it is appropriate based on the actual data. In certain cases, transforming data can create unus
Related Questions
- The distribution that shows up when I carry out a Capability Study is not Normal. Does this mean I can’t use conventional Control Charts and do I have to calculate Cp and Cpk in a different way?
- What is normal distribution in statistics and what are the three methods to test data sets for normality?
- What does normal or transformed-normal mean when referring to the distribution of the data?
