Incoherent Graph

[tem3]

Hello,

I am in the process of creating a very simple PLS model with 2 formative LVs (each with 4 observed variables) loading on 1 reflective LV (with 3 observed variables). I expected the LVs to turn from Red to Blue after connecting them, but they have not changed and I get an error message indicating that the "variable is part of an incoherent graph." I have made several models before and never had this problem. Any suggestions as to what I might be doing wrong?

Thanks very much,
Teresa McCarthy

[jjsailors]

Hello,

I am in the process of creating a very simple PLS model with 2 formative LVs (each with 4 observed variables) loading on 1 reflective LV (with 3 observed variables). I expected the LVs to turn from Red to Blue after connecting them, but they have not changed and I get an error message indicating that the "variable is part of an incoherent graph." I have made several models before and never had this problem. Any suggestions as to what I might be doing wrong?

Thanks very much,
Teresa McCarthy

Are the structural paths in red? The only thing I can think of is if you have this system set up:

A --> B
B --> C
C --> A

Which you probably don't, but it's all I can think of so I thought I'd put it out there!

[tem3]

Hi John,

Thanks for your reply, but unfortunately that's not the problem. My structural model is:

A ---> C
B ---> C

It's a very simple model, so I don't understand why I am encountering this problem.

Thanks,
Teresa

[amadeus]

I am also receiving this message.

Does anybody have a solution?

Michael Cuellar
Georgia State University

[tem3]

Hello Michael,

It has been a couple of months since I encountered this problem, so I don't recall exactly what created the problem. However, I know it must have stemmed from something I did incorrectly with my model specification. After I deleted and recreated the model, the problem was resolved. It must have been a problem with the way I created the first model, but it just wasn't obvious to me at the time. Good luck!

Teresa

[zanna.van.dun]

I also get this message while I have never encounted this problem before, not even in a rather complicated model. Does anybody know another answer that just redo the model...? I've tried several new variations and simple relations but every relation just stays red with the 'Incoheren graph' note.

[alan_scott]

I also get this message while I have never encounted this problem before, not even in a rather complicated model. Does anybody know another answer that just redo the model...? I've tried several new variations and simple relations but every relation just stays red with the 'Incoheren graph' note.

Same for me ... solved by re-drawing the model. But note that it is possible to accidentally enter a latent variable that is so small you can't see it on the screen - in this case the graph will be incoherent until you find this impossible-to-see-LV and delete it :-)

[Tommy_E]

I had the same problem, and it was an invisible variable, that was the problem.

I Found out that the best way to find that invisible variable was to use the "outline" box to find a variable that had a name that did not correspond to what was expected.

[RetoFelix]

The invisible variable was also my problem. The advice above saved my day; I checked out the Outline window on the left side and voila, there I had a LV (latent variable) I neved intended to use, named (in my case) "New_1". In case you can't see the Outline window, just go in the main menu to Window - Show View and click on the Outline option. Deleted this invisible LV and everything turned blue... :-)