|
|
 |
| Integration Algorithm and Outcome Mea... |
 
|
| Message/Author |
|
|
|
|
Hi Drs. Muthen,
I'm attempting to run a structural equation model that includes an interaction between latent variables. I'm aware that this requires me to use the Integration algorithm; however, something about that algorithm is preventing Mplus from displaying fit indices such as the RMSEA. If the interaction and integration algorithm are removed, the model runs as desired. My code is included below. Do you have any advice on how to fix this problem?
Variable:
Names are
id MEpos_d MEneg_d MI_d CEpos_d CEneg_d CI_d wkce
classes onet sex mat1 mat2 mat3 utl30 utl31
utl28r utl29r GPA;
USEVARIABLES ARE
MI_d CI_d wkce classes onet mat1 mat2
mat3 utl30 utl31 utl28r utl29r GPA;
MISSING = ALL (-9);
Analysis:
TYPE = RANDOM;
Estimator=MLR;
ALGORITHM=INTEGRATION;
Model:
self_c BY mat2 mat3 mat1;
dweck BY utl30 utl31 utl28r utl29r;
perf BY wkce GPA;
CI_d WITH MI_d;
onet WITH classes perf;
perf WITH classes;
dweck ON CI_d;
int | dweck XWITH self_c;
perf ON self_c dweck CI_d int;
classes ON self_c dweck CI_d int;
onet ON self_c dweck CI_d int; |
|
|
|
|
| These fit statistics are not available with TYPE=RANDOM. |
|
|
|
|
| Thank you for your quick response, Dr. Muthen! Is there a way to calculate an interaction between two latent variables without TYPE=RANDOM? |
|
|
|
|
| No. |
|
|
| Back to top |
|
|
