How to use ApproxStdErr from Nonliear model in a Tukey analysis?

Question

I am using a 4p nonlinear model to get the inflection points of curves defined by my data, theta4.  I would like to then run something like a Tukey analysis to determine if there are significant differences between these inflection points.  I have been looking for a way to pass the ApproxStdErr from the 4p model but have not found a solution.  It seems to me that this should be a fairly common situation and would appreciate any guidance in how it should be solved.

KarenC · Accepted Answer

Richard,

If you are fitting your models using the non-linear platform with a grouping variable for your different "groups" then there is an option called compare estimates that utilizes Analysis of Means to compare the parameter estimates across the groups.

Karen