In this Weibull distribution example, a transformation of process data to achieve normality seems like magic, but it is not. See the attached white paper article link below to understand why this statement is made.
There are a lot of reasons that specific transformations make sense and should be used, such as a lognormal transformation of standard deviations and of time data. But, when it comes to the Weibull distribution, there is no logical or inferential information that provides any guidance on a transformation to normality, until now.
Weibull Distribution Example: A Transformation with Surprising Result
Rick Haynes has found a relationship between the Weibull distribution parameters and the optimal Box-Cox transformation lambda value. Through the use of the @Risk simulation program and Minitab, this article (link below) walks through the generation of an equation to predict a lambda value.
