|
|
Factor scores as endogenous variables |
|
Message/Author |
|
|
I read posts RE: using factor scores as endogenous variables, along with the Plausible Values for Latent Variables Technical Report and the Multiple Imputation with MPLUS article. I would like to use a Bayes approach to test a model composed of those (Bayes-estimated) factor scores. Would you please point me to one or more examples of how to do this in MPLUS (I'm using Version 7.11)? Probably my being dense, but I could not find anything in my search of the web site nor the manual, nor could I figure out how (that is, syntax) to incorporate the information in the SAVEDATA: file for the final Bayes model testing... Thanks in advance... Michael |
|
|
See Example 11.7 for generating and saving the plausible values. See Example 13.13 for using them in an analysis. |
|
|
Hello, I plan to use the aforementioned technique for estimating factor scores; I plan to use the plausible values in a model with both latent and observed variables. I am choosing this method to reduce the computational burden - the model has 3 latent factors already, and I am looking to add 6 more (3 socio-emotional constructs at 2 time points). I have not seen a model that includes both factor scores and latent variables - is this reasonable? And, after estimation, can factor scores be modeled as dependent variables? |
|
|
Yes on both. |
|
Back to top |
|
|