|
|
 |
| Using residuals in equations as predi... |
 
|
| Message/Author |
|
|
|
|
I would like to know how residual factors can be used as predictors of other factors in SEM. In my research I use EQs because this is very easily done. For instance, when I have three factors, F!, F2< and F3, I can regress F2 on F1 and F3 on F1 and the residual (or disturbance) of F2, that is D2. In the EQs language, this would be as follows:
F2 = *F1 + D2;
F3 = *F1 + *D2 + D3;
I have not been able to find how I can take the residuals of factors in Mplus and use them as predictors of other factors, as specified above.
I would appreciate it if you would help in this because I highly value the many functions of Mplus not included in EQS.
Best regards,
Andreas Demetriou |
|
|
|
|
This is described in a FAQ on our website (see top of left column) called
Regressing on a residual |
|
|
| Back to top |
|
|
