Hello
I'm in the process of fitting a multiple linear regression model
The Y variable is Gross pay per hour, which is very skewed.
I have tried several transformations (Log, Sqrt, 1/sqrt, 1/x) to try and get the distribution of variable to approximate normal, but I'm not having any luck - the distribution is still noticeably skewed - although it becomes negatively skewed when using the log transformation.
I'm not sure how to proceed, because the residuals of the model are consequently not normally distributed.
If anyone has any advice I would really appreciate it.
Thanks,
Nichola
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|