|
|
Mediation Growth Curve Modeling |
|
Message/Author |
|
|
Hello Drs. Muthen & Muthen, I am trying to conduct a mediation analysis in Mplus using a longitudinal growth curve analysis. The DV (HARS) has 4 time points and both of my IVs (T1ASIm and T1DRS) are static. I wanted to have the mediator as a growth process but am unable to within warnings. I keep recieving "WARNING: THE LATENT VARIABLE COVARIANCE MATRIX (PSI) IS NOT POSITIVE DEFINITE."Also, under the Indirect Command Output, all it has are 0's and 999.000 under Est./S.E. Here is the syntax: ANALYSIS:ESTIMATOR = MLR;iteration=10000; MODEL:iHARS sHARS | T1HARS@-2 T2HARS@-1 T3HARS@0 T4HARS@1; T1HARS with T3HARS;T2HARS with T4HARS;iHARS with sHARS; iMFQ sMFQ | T1MFQfrq@0 T2MFQfrq@-1 T3MFQfrq@-2 T4MFQfrq@-3;T1MFQfrq with T3MFQfrq;T2MFqfrq with T4MFQfrq;iMFQ with sMFQ;T1HARS with T1MFQfrq;T2HARS with T2MFQfrq;T3HARS with 3MFQfrq;T4HARS with T4MFQfrq;T1HARS with T1ASIm;T1HARS with T1DRS;T1ASIm with T1DRS;T1ASIm with T1MFQfrq;T1DRS with T1MFQfrq; iHARS on T1ASIm T1DRS;sHARS on T1ASIm T1DRS; MODEL INDIRECT: iHARS ind T1ASIm;sHARS ind T1ASIm; OUTPUT: STANDARDIZED SAMPSTAT TECH1 TECH4 MODINDICES CINTERVAL Thank you! Chelsey |
|
|
Please send your output and license number to support@statmodel.com. |
|
Back to top |
|
|