Path analysis and correlations PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Scott C. Roesch posted on Saturday, May 26, 2007 - 5:43 am
I have run a very basic path analytic model using the complex procedure using the primary code below. Using the code below, however, I do not receive the correlations among the exogenous variables (covariates) in the output file as traditionally would occur using the With statement in other MPlus procedures. Is this not a default option in the Complex procedure?

CLUSTER IS agencyid;
MISSING IS BLANK;
USEVARIABLES private agysup9 ebpasm
ed white q47 yrsagy male Q7;

ANALYSIS:
TYPE = COMPLEX MISSING H1;
ESTIMATOR = MLR;
ITERATIONS = 10000000;
H1ITERATION = 100000000;

MODEL:

agysup9 ON private;

ebpasm ON private agysup9
ed white q47 yrsagy male Q7;

OUTPUT: STANDARDIZED MODINDICES SAMPSTAT ;
 Linda K. Muthen posted on Saturday, May 26, 2007 - 6:38 am
In most parts of Mplus, the model is estimated conditioned on the covariates as in regular regression so the means, variances, and covariances of the covariates are not estimated as model parameters. If you want these values, you can obtain them using TYPE=BASIC;
 Scott C. Roesch posted on Sunday, May 27, 2007 - 8:05 am
Thanks Linda...let me pursue this a little further. When I constrain some of the correlations between covariates to 0 (e.g., agysup9 with private) not only do the degrees of freedom increase (as if MPlus is estimating this correlation), but it gives me the correlations among the other covariates in the output.
 Linda K. Muthen posted on Monday, June 04, 2007 - 7:33 am
If you mention the means, variances, or covariances of the covariates in the MODEL command, they are treated as dependent variables in the model.
 Scott R. Colwell posted on Wednesday, November 26, 2008 - 8:46 am
Do you have any examples of a simple path analysis / regression with LCA? Such as Y on X with 3 classes?
 Linda K. Muthen posted on Wednesday, November 26, 2008 - 8:55 am
Examples 7.1 and 7.2 are mixture regressions. You can expand the path analysis examples in Chapter 3 to mixtures.
 Suzanne Elgendy posted on Thursday, June 24, 2010 - 11:32 am
Dr. Muthen,

I was hoping you could help me understand the benefit of running a path model with observed variables set up as latent variables with one indicator versus simply running the model with the observed variable as a true observed variable. I was advised in the past to run a path analysis using latent variables with one indicator and am getting highly inadequate model fits. When I re-run these models using observed variables, the model fits are adequate. I am not quite sure what could be driving this difference.

Thank you,
Suzanne Elgendy
 Linda K. Muthen posted on Thursday, June 24, 2010 - 12:39 pm
There should be no difference between using observed variables versus putting a latent variable behind each observed variable if the residual variance of the observed variable is fixed to zero, for example,

f BY y;
y@0;

There must be some other difference.
 Mario Mueller posted on Monday, April 23, 2012 - 2:04 am
Hello,

I run a LPA with 2 classes as best solution and included the latent class variable as binary outcome in a regular path analysis.
Is there a way to link both analyses and do you have any examples for that? Or is this an appropriate procedure?

Thanks, Mario
 Linda K. Muthen posted on Monday, April 23, 2012 - 12:37 pm
In the LPA analysis you can regress the categorical latent variable on a covariate, for example,

c ON x;

where x can also be used as a dependent variable in the path model.
 Samuli Helle posted on Thursday, October 25, 2012 - 12:47 pm
Hi,

I'm conducting a multiple-group path analysis (a MANOVA-type of an analysis), but I'm a bit confused what would be the preferred way of modeling the IVs: should I include e.g. the covariances between the IVs into the model (which seems to treat the IVs as DVs?) or not (as in regular regression approach)? This troubles me, because including these covariances (even by setting them zero) seems to influence the sample size used in the analysis. I'm using MLR estimator.

Thank you very much in advance.

Samuli
 Linda K. Muthen posted on Thursday, October 25, 2012 - 2:08 pm
The default in Mplus is to use all available information and TYPE=MISSING. Missing data theory applies to only the dependent variables. If you include the covariates in the model instead of estimating the model conditional on the covariates, you make distributional assumptions about them. It is your choice whether it is worth doing this.
 Noud Frielink posted on Wednesday, February 03, 2016 - 7:43 am
Dear Dr. Muthen,

Using path analysis, we would like to test the following model:
- independent variable (or exogenous variable): HCCQ (or X)
- dependent variable: SWLS (or Y2)
- mediator variable: SRQ (or Y1)

The HCCQ exists of 15 items,representing together 1 factor
The SWLS exists of 5 items, representing together 1 factor
The SRQ exists of 8 items, representing 4 factors:
SRQext BY SRQcs2 SRQcs9;
SRQintj BY SRQcs1 SRQcs3;
SRQide BY SRQcs4 SRQcs11;
SRQinte BY SRQcs7 SRQcs8;

I'm wondering how I should include these four factors within the path analysis. Both the independent variable and the dependent variable has 1 factor, but the mediator variable has 4. Right now, we have the following syntax:

MODEL:
HCCQ BY HCCQ1-HCCQ15;
SRQext BY SRQcs2 SRQcs9;
SRQintj BY SRQcs1 SRQcs3;
SRQide BY SRQcs4 SRQcs11;
SRQinte BY SRQcs7 SRQcs8;
SWLS BY SWLS1 SWLS2 SWLS3 SWLS4 SWLS5;

SRQext on HCCQ;
SWLS on HCCQ SRQext;

MODEL INDIRECT:
SWLS ind HCCQ;

OUTPUT:
STDYX SAMPSTAT;


However, in this case we only take into account 1 of the 4 factors of the mediator variable (i.e., SRQext, and not SRQintj, SRQide, and SRQinte). Could you please advice us?

Thank you very much and kind regards,
Noud Frielink
 Bengt O. Muthen posted on Thursday, February 04, 2016 - 6:50 pm
Your ON statements should add all SRQ factors. Model Indirect will then give all the indirect effects.
 Noud  posted on Friday, February 05, 2016 - 2:52 am
Dear prof. Muthen,

Thank you very much for your response. I will add all SRQ factors to the ON statements.

I have one follow-up question, as there might be one additional problem. That is, in my previous post I somewhat simplified our model, as we have two mediator variables consisting of multiple factors (SRQ has 4, BPN has 6). Adding all SRQ and BPN factors to the ON statements results in the following syntax:

MODEL:
SRQext SRQintj SRQide SRQinte on BPNas BPNaf BPNrs BPNrf BPNcs BPNcf HCCQ;
SWLS on SRQext SRQintj SRQide SRQinte BPNas BPNaf BPNrs BPNrf BPNcs BPNcf;
BPNas BPNaf BPNrs BPNrf BPNcs BPNcf on HCCQ;

MODEL INDIRECT:
SWLS ind SRQext SRQintj SRQide SRQinte HCCQ;
SWLS via BPNas BPNaf BPNrs BPNrf BPNcs BPNcf HCCQ;

OUTPUT:
STDYX SAMPSTAT;

Is this correct, and can I interpret the outcome of this syntax in such a way that if model fit is adequate, that model as a whole can be accepted?

Thank you very much and kind regards,
Noud Frielink
 Bengt O. Muthen posted on Friday, February 05, 2016 - 5:35 pm
Give it a try and see if you get the effects you expect. If you get an error, send to Support along with your license number.

Even if the model fits well it can be misspecified.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: