Overview for Box-Cox Transformation

A Box-Cox transformation of your process data can help correct the following conditions:
  • The process data are not normally distributed, especially if the data are not collected in subgroups. For more information, go to Normality assumptions for control charts.
  • Subgroup variance is unstable because the variation in the data is proportional to the subgroup mean.

You can also perform a Box-Cox transformation when you create a control chart. For example, to transform the data when you create an Xbar-R chart, choose Stat > Control Charts > Variables Charts for Subgroups > Xbar-R > Xbar-R Options > Box-Cox.

Where to find this command

To perform a Box-Cox transformation, choose Stat > Control Charts > Box-Cox Transformation.

When to use an alternate analysis

If you have data values less than 0, use a Johnson Transformation.