Regression
Message/Author
 rick gibbon posted on Wednesday, November 03, 2010 - 4:58 pm
I am using mplus 3. In a regression analysis, my syntax is

PC4HEPROB ON PC1ses;
PC4HEPROB ON PC1age;
PC4HEPROB ON PC3NLE;
PC4HEPROB ON pc3finstr;
PC4HEPROB ON pc3nghdv;

where PC4HEPROB is the outcome and the other are five different covariates. These are 5 simple regressions without adjusting for any covariates in each regression. But the output gives me a multiple regression results
PC4HEPRO ON
PC1SES -0.059 0.035 -1.703 -0.090 -0.077
PC1AGE 0.015 0.004 3.973 0.022 0.172
PC3NLE 0.030 0.013 2.414 0.046 0.102
PC3FINSTR 0.128 0.038 3.361 0.196 0.154
PC3NGHDV 0.019 0.062 0.310 0.030 0.013

How can I get 5 simple regression results? Thanks a lot.
 Linda K. Muthen posted on Wednesday, November 03, 2010 - 5:03 pm
You would need to run 5 separate analyses each with one of the ON statements.
 rick gibbon posted on Thursday, November 04, 2010 - 6:02 am
What if I want to run several multiple regessions controlling for the same set of covariates? For example, suppose I am interested in the main effects of PC1ses, PC1age and PC3NLE on the outcome PC4HEPROB controlling for pc3finstr and pc3nghdv. So there are three multiple regressions
PC4HEPROB = PC1ses+pc3finstr+pc3nghdv;
PC4HEPROB = PC1age+pc3finstr+pc3nghdv;
PC4HEPROB = PC3NLE+pc3finstr+pc3nghdv;

How can I achieve this in mplus 3? Thanks.
 Linda K. Muthen posted on Thursday, November 04, 2010 - 6:06 am
You would need to run a separate analysis for each model. The MODEL command defines one model only.
 rick gibbon posted on Thursday, November 04, 2010 - 11:58 am
You meant the MODEL command defines one model only for one exogenous variable? Can I have several regressions with different exogenous variables under one Model command?
 Bengt O. Muthen posted on Thursday, November 04, 2010 - 12:32 pm
The MODEL command is not limited to one exogenous variable. Please see the examples chapters of the user's guide.