Multilevel CFA with continous latent ... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Pedro Quinteiro posted on Thursday, December 06, 2012 - 11:03 am
Dear Professors,

I am trying to run a homologous multilevel CFA, meaning I am testing whether construct Y is equivalence at different levels of observation (individuals and teams).

Although the variables in the model are continuous, it happened that I had a better model fit when using ESTIMATION = WLSM (instead of the recomended MUML).

This is part of the sintax I am using:


ANALYSIS:
TYPE = TWOLEVEL;
ESTIMATOR = wlsm;
ITERATIONS = 100000;
CONVERGENCE = 0.00005;

MODEL:
%BETWEEN%
xbtik BY ik1-ik10;
xbttk BY tk1-tk10;
ik-ik10@0;
tk1-tk10@0;

%WITHIN%
xwtik BY ik1-iki10;
xwttk BY tk1-tk10;

OUTPUT: Stand Res;

Hope you can help me.

Best regards
Pedro
 Linda K. Muthen posted on Thursday, December 06, 2012 - 1:35 pm
I'm not sure why you think MUML is recommended. The default is MLR.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: