Constraining paths across levels PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Aislin Graham posted on Thursday, March 31, 2011 - 1:17 am
I have few questions about my analyses and the interpretation of my results for a two-level SEM:

1) Do I need to constrain path across the within and between models?

2) Should I be specifying residual error variances? If so, what will this tell me?

3) Am I specifying covariances correctly? What does this tell me?

CLUSTER = ID;
BETWEEN = SPP IPP ESPP;
analysis:
TYPE = TWOLEVEL RANDOM;

model:

%WITHIN%
!measurement model
DISCw BY MDI APS DEQ;
PSPSw BY PROM DISCLO DISPL;
DAw BY POMSD DACLG PANASX;
PM_BEw BY PM_BULIT PM_EDIB PM_EDDS;
N_DISCw BY N_MDI N_APS N_DEQ;

!variances and covariances
PSPSw DAw PM_BEw N_DISCw;

!path coefficients
PSPSw ON DISCw;
DAw ON PSPSw;
PM_BEw ON DAw;
N_DISCw ON PM_BEw;

%BETWEEN%
!measurement model
SPPb BY SPP IPP ESPP;
DISCb BY MDI APS DEQ;
PSPSb BY PROM DISCLO DISPL;
DAb BY POMSD DACLG PANASX;
PM_BEb BY PM_BULIT PM_EDIB PM_EDDS;
N_DISCb BY N_MDI N_APS N_DEQ;

!variances and covariances
DISCb PSPSb DAb PM_BEb N_DISCb;

!path coefficients
DISCb ON SPPb;
PSPSb ON SPPb;
DAb ON SPPb;
PM_BEb ON SPPb;
N_DISCb ON SPPb;
PSPSb ON DISCb;
DAb ON PSPSb;
PM_BEb ON DAb;
N_DISCb ON PM_BEb;
 Linda K. Muthen posted on Thursday, March 31, 2011 - 3:42 pm
These questions are too broad to be addressed on Mplus Discussion. I would suggest that you listen to the Topics 7 and 8 course videos on the website.
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: