Box-Cox transformations have always been a feature that has been tucked away under the red triangle options of Fit Model. In version 13 of JMP this functionality is brought to the foreground. It appears as default output when you choose the Effect Screening emphasis.
This is a strange change of behaviour that I’m not particularly happy about. However, I’m trying to focus on features new to version 13, so I will stick to the facts and hold back on my rant!
Given that JMP have chosen to stick this functionality in your face it’s a shame that the default graph leaves a lot to be desired:
So grab the resize handles and give it a sensible appearance:
Whilst the power transformation lambda is on a continuous scale I would rather think in terms of meaningful transformations. Instead of -0.318 I would be more interested in 0.0 i.e. a log transformation.
Conveniently JMP have extended the red triangle options to allow me to quickly investigate a model using this transformation: