MGA - Different results - Why?

[Anna_Mueller]

I did a MGA with SmartPLS 3.0.

What is the difference between the value of path coefficient and total/indirect effect?

One path from my model is not significant regarding the value of Path Coeffincient, but it is significant regarding the value of total effect (which is by the way the same as indirect effect).

Why is the same path once significant and once not?

[jmbecker]

Model:
A --> B --> C and A-->C

Direct Effect: A-->C
Indirect Effect: A-->B * B-->C
Total Effect: A-->C + A-->B * B-->C

If your indirect is the same as the total effect, then you don't have a direct effect in your model.

[Anna_Mueller]

Thank you, maybe I wrote it a bit confusing:
I mean there are results for the topic "Path coefficient" and e.g. "Total Effect".
How is it possible that a Path A--> B is significant regarding the results of "Path Coeffiecient" but this path is not significant regarding "Total Effect".

[jmbecker]

Simply because they are not the same (see above).

[Anna_Mueller]

I am really sorry but I still did not get it :(.

Model:
A --> B --> C and A-->C

Direct Effect: A-->C
Indirect Effect: A-->B * B-->C
Total Effect: A-->C + A-->B * B-->C

I have a model like this:
A -> C -> D
B -> C -D
So there are two latent variables (A and B) that influence the latent variable C, and this one leads to latente variable D

There are no other connections/variables. This is the whole model with all interactions/connections/variables.

Regarding my Output of MGA (comparing the output of the three columns I get "Path Coefficient", "Total Effect" and "Indirect Effect"):

Path Coefficients:
Henseler's MGA:
A -> C is significant

Indirect Effect:
Henseler's MGA:
A -> C is not significant

Total Effect:
Henseler's MGA:
A -> C is not significant

It is a direct Path, but I really do not understand why THE SAME Path has different results regarding the significance.

Which one is relevant to look at?

[jmbecker]

Do you use the latest version 3.1.6? In an earlier version there was a bug regarding the indirect effect calculation.

If you have the latest version, then you somehow modeled an indirect relation between A and C (maybe through B?). Otherwise, there would be no result for the indirect effect of A-->C if the model includes only the direct effect.

[Anna_Mueller]

No, I have 3.1.5.
Can I download the new version without losing all my models/calculations?

Still the output of the "total effect" should be the same as the outpu if "path coefficient", but it is not

[jmbecker]

Can I download the new version without losing all my models/calculations?

Of course.