Transform your nonnormal data with a Box-Cox transformation. Transformations are not usually necessary for control charts with subgroups. But if the data are very skewed, you might want to use a Box-Cox transformation.
If you specify historical parameters, you cannot use a Box-Cox transformation.
To use the Box-Cox transformation, select Use a Box-Cox transformation (W = Y^λ), then select the lambda (λ) value that Minitab uses to transform the data.
- λ = 0 (ln): Use the natural log of your data.
- λ = 0.5 (square root): Use λ = 0.5 (square root).
- Optimal λ: Use the optimal λ, which should produce the best fitting transformation. Minitab rounds the optimal λ to 0.5 or the nearest integer.
- Optimal λ for each stage (in a chart with stages): Use the optimal λ, which should produce the best fitting transformation for each stage in the chart. Minitab rounds the optimal λ to 0.5 or the nearest integer.
- Other (enter a value between -5 and 5): Enter a value for λ. Other common transformations are square (λ = 2), inverse square root (λ = −0.5), and inverse (λ = −1). If your control chart has stages, you can enter a separate λ for each stage.