No direct effects given in my output? PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Emily Rosenzweig posted on Monday, June 08, 2015 - 8:22 pm
Hi --

I am running an analysis where type=complex, and where I am also modeling a specific indirect pathway. However I would also like to see an estimate of (and significance testing for) the direct effect that remains between my DV and IV after the indirect effect is accounted for.

However even though the output header says "TOTAL, TOTAL INDIRECT, SPECIFIC INDIRECT, AND DIRECT EFFECTS" it is only giving me the sum of indirect effects and my specific indirect effect. What do I need to do to get the direct effect included in the output?

Many thanks!
 Linda K. Muthen posted on Tuesday, June 09, 2015 - 10:23 am
See page 691 of the user's guide. The last paragraph of MODEL INDIRECT shows the specification to use.
 Emily Rosenzweig posted on Tuesday, June 09, 2015 - 5:30 pm
Hi Dr. Muthen --

That last paragraph is exactly what I'm already using -- here is my current code:

Names are ID, FA, Fabinary, PK, subject, Hardeasy, Context;
Usevariables are ID, FA, subject, Hardeasy, PK;
subpopulation = (context EQ 1);
cluster is id;
type is complex;
PK on subject Hardeasy;
FA on subject PK Hardeasy;
model indirect:
FA via pk hardeasy;

But the output gives me no estimates of the direct effect. As I stated, it only gives me the sum of indirect effects and the one I specify.

Many thanks for your help!
 Bengt O. Muthen posted on Tuesday, June 09, 2015 - 6:08 pm
VIA is an indirect effect option. To get all effects, use

FA IND hardeasy;
 Emily Rosenzweig posted on Tuesday, June 09, 2015 - 6:22 pm
Hi Bengt --

Unfortunately that didn't seem to solve the problem :-(

Should I send you the dataset and input file to see if you replicate? I don't know if I'm missing something, but it seems like if the direct effect was there it would be clearly labeled, correct? Here is what my output looks like:


Effects from HARDEASY to FA

Sum of indirect

Specific indirect


After that it just repeats the same for the standardized variables...
 Linda K. Muthen posted on Wednesday, June 10, 2015 - 6:20 am
Please send the output and your license number to
 Timothy posted on Monday, July 04, 2016 - 5:09 am
Hi Linda,

I experienced the same problem (without the direct effect in the output file). I also used the "IND" command as you suggested and repeated the same problem as Emily did. Can you help me to get the direct effects estimates?

Thanks a lot.
 Linda K. Muthen posted on Monday, July 04, 2016 - 1:19 pm
Usually when you don't get the indirect effects it is because the components of the indirect effect are not present in your model. If this does not help, send your output and license number to
 fred posted on Saturday, November 04, 2017 - 11:58 am
Strangely enough I am not getting any direct effect estimates in my output of a fairly straight forward analysis. The model is:

model: Mediator by c1-c6;
Mediator on Pred cohab age;
Outcome on Pred Mediator age;
model indirect: Outcome ind Mediator

The output has only indirect effect listed:


Estimate S.E. Est./S.E. P-Value

Effects from PRED to OUTCOME

Indirect 0.035 0.012 3.054 0.002
 Bengt O. Muthen posted on Saturday, November 04, 2017 - 4:14 pm
Drop "Mediator" from your Model Indirect statement.
 fred posted on Saturday, November 04, 2017 - 4:47 pm
Thank you! Is this because there is only a single mediator?
 Bengt O. Muthen posted on Sunday, November 05, 2017 - 1:37 pm
No, it's just a convention. See the UG.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message