|
|
Multigroup SEM, complex data, and 3 c... |
|
Message/Author |
|
|
I am attempting to run a multigroup SEM on a complex data set that includes three categorical outcome variables. I keep getting error messages. I have included most of my syntax below. Any assistance would be appreciated. categorical = distress behavior suicide; missing = all(-999); weight = WEIGHT; strat = STR; cluster = SECU; grouping = ACE2_G7 (0=ACE0 1=ACE1 2=ACE2 3=ACE3); analysis: type = complex; estimator = MLR; model: Religion by...; Family by ...; Friend by ...; Teacher by ...; School by ...; Distress ON Age GENDER POVERTY...; Behavior ON Age GENDER POVERTY...; Suicide ON Age GENDER POVERTY...; Age GENDER POVERTY PAR_EDUC US_CIT O_DSM Black Hispanic Other Metro O_Urban WITH Religion Family Friend Teacher School; Distress WITH Behavior; Distress WITH Suicide; Behavior WITH Suicide; model ACE1; model ACE2; model ACE3; output: modindices standardized; *** ERROR in ANALYSIS command ALGORITHM=INTEGRATION is not available for multiple group analysis. Try using the KNOWNCLASS option for TYPE=MIXTURE. |
|
|
When you use maximum likelihood with categorical dependent variables, numerical integration is required. The GROUPING option is not available in this case. Multiple group analysis is done with TYPE=MIXTURE and the KNOWNCLASS and CLASSES options. When all classes are known, it is the same as multiple group analysis. |
|
Back to top |
|
|