Greetings,
I recently saw a presentation advocating the use of the Johnson Su
transformation (also known as the Inverse Hyperbolic Sine
transformation), as an alternative to the more familiar Box-Cox
transformation. I have not managed to come across much information on it
and I was hoping that someone on Allstat might point me to some relevant
references. I'm particularly interested in ways of estimating the
transformation parameters for specific data sets.
Thanks
-Mark
|