Message/Author |
|
|
Is this command only available in Mplus Version 6? |
|
|
MODEL INDIRECT was introduced in Version 3 and has been available since then. |
|
|
I have attempted to use this command in my SEM. Versions 5 & 6 are not reading the command. For example, in the following syntax: DATA: FILE IS F:\DissertationData.dat; VARIABLE: NAMES ARE Y1-Y15; MODEL: F1 By Y1, Y3; (ETC) F1 ON F4 F3; (ETC) MODEL INDIRECT: F4 IND F3 F1; M+ is not reading the MODEL INDIRECT command. That is, all other commands are indicated in blue text but the MI command is not and output does not represent this command. Could you explain? |
|
|
Please send your output and license number to support@statodel.com. |
|
Xu, Man posted on Thursday, January 04, 2018 - 1:08 pm
|
|
|
Dr.s Muthen, I was wondering if the MODEL INDIRECT COMMAND works with ESEM anlysis? Thanks! Kate |
|
|
That should work, I think. |
|
Xu, Man posted on Wednesday, January 10, 2018 - 4:36 am
|
|
|
yes it works thanks! |
|
|
Hello I am using ESEM (it is fitting better than CFA-type approach, and in this case ESEM is more theoretically appropriate as well). My model also has indirect effects that I'd like to model. Usually I use bootstrap or bcbootstrap confidence intervals for indirect effects. But error message says that BOOTSTRAP is not allowed for analysis with EFA factors. I've previously been told to use the bootstrap CIs for indirect effects, but in this case it won't work. Any advice on this? |
|
|
Tricky. Bayes would have been a way to allow for the non-symmetric CIs that indirect effects sometimes require (hence bootstrap for ML) but ESEM is currently not available with Bayes. |
|
Back to top |