Dear J,
This is a good question. No we don't plan to extend Stata's postestimation
predict command to work with
runmlwin. This would involve working out and implementing the prediction formula for a very wide range of multilevel models. The aim of
runmlwin is to make MLwiN functionality accessible to Stata users, not to directly implement MLwiN functionality in Stata. The latter would be much more ambitious!
However, the good news is that you can easily calculate a wide range of in-sample and out-of-sample predictions using
runmlwin. All you need to do is to pull back the residuals from MLwiN and then do a small amount of data manipulation.
So for example, if you ran the following two-level growth trajectories model for child weight:
use http://www.stata-press.com/data/r11/childweight, clear
. bysort id (age): gen occasion = _n
. generate cons = 1
. generate agesq = age^2
. runmlwin weight cons age agesq, ///
level2(id: cons age, residuals(u)) ///
level1(occ: cons) plugin
You would generate and graph the growth trajectory for the mean child as follows:
. generate xb = _b[cons] + _b[age]*age + _b[agesq]*agesq
. line xb age, sort
And you would generate and graph the child specific growth trajectories as follows:
. generate xbu = xb + u0 + u1*age
. line xbu age, connect(a)
For a wide range of examples, for both continuous and discrete response models, please see the
runmlwin do-files for the MLwiN User Manual.
http://www.bristol.ac.uk/cmm/software/r ... /examples/
I hope this helps
George