Re: st: Fw: QUESTION FOR LISTSERVER

[William Buchanan <william@williambuchanan.net>]

Hi Charles,

Users of the Statalist are asked to read the Statalist FAQ prior to posting queries.  Based on the information you've provided there is no way anyone can give you any reasonable response. When you estimate a model from the GUI you are provided with the syntax for the estimated model. You should include this syntax and the results in your query, and if possible use datasets that come with Stata so other users can attempt to replicate your results. 

Not everyone has Mplus, but those of us that do could probably give you a better answer if we see all of the syntax and output that you are attempting to compare.

HTH,
Billy

Sent from my iPhone

On Dec 13, 2012, at 15:46, Charles Van Liew <cavanliew@yahoo.com> wrote:

> Hello,
> 
> I am trying to figure out how to model a conditional latent growth curve analysis in Stata's GUI.  Basically, I am trying to mimic results that were obtained using MPlus.  I can model an unconditional latent growth curve analysis in the GUI by erecting a model with two latent variables (Intercept & Slope) that are indicated by the observed variables (e.g., 4 repeated measures of "x").  By constraining the constant values for the "x" observed variables to 0 (do not estimate), the x -> Intercept paths to 1, and x -> Slope paths to 0, 1, 2, and 3, respectively.  However, when I attempt free any of the paths (e.g., x3 -> Slope and x4 -> Slope), so that a conditional model can be estimated the results no longer remain consistent with MPlus.  Am I missing something?
> 
> Thanks!
> 
> Best,
> Charles
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/