hybrid multidimensional CFA model: a second order construct with first-order formative dimensions

[Jonas Van Hove]

Dear,

 

Please advise me on the following (find my conceptual model above). When I try to conduct CFA in order to specifically articulate that there are two distinct forms of passion - Pass.for work and Entr.pass., I receive the warning message;

In lav_model_vcov(lavmodel = lavmodel, lavsamplestats = lavsamplestats, :

lavaan warning: could not compute standard errors!

lavaan NOTE: this may be a symptom that the model is not identified

Syntax:

# model the relationships between each passion construct (pass.for work; Entr.pass) and its two dimensions of IPF... and ID... as formative!

# this is a second order construct with first-order formative dimensions which are themselves measured by several reflective manifest variables

#  ID is a single item and we look for interaction between IPF and ID.

Independentmodel <-'IPFpasswork =~ X1pw + X2pw + X3pw + X4pw

IDpasswork =~ IDwork

IPFentrpass =~ X1ep + X2ep + X3ep

IDentrpass =~ IDdev

Passforwork<~ IPFpasswork * IDpasswork

Entrpass<~ IPFentrpass * IDentrpass

'

fit<-cfa(Independentmodel, data = Passwork.and.entrpass)

Any help is much appreciated. I suspect that my model is not under-identified but please feel free to correct me if I am wrong. Or do I have a mistake in my syntax? Or is it because my 'formative model doesn't predict anything...'?

Thanks.

J

[Sunthud Pornprasertmanit]

You have factors that have only one indicator. I recommend you to avoid using latent variables for single indicators. For example,

Independentmodel <-'

IPFpasswork =~ X1pw + X2pw + X3pw + X4pw

IPFentrpass =~ X1ep + X2ep + X3ep

Passforwork<~ IPFpasswork + IDwork

Entrpass<~ IPFentrpass * IDdev

'

fit<-cfa(Independentmodel, data = Passwork.and.entrpass)

P.S. I am not good on the formative factors. I am not sure whether this script is still non-identified.

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at http://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/d/optout.

[Mikko Rönkkö]

Hi

The formative part is not identified. You always need reflective variables to identify the model, or alternatively you need to fix the formative paths to know values.

This should be identified:

IPFpasswork =~ X1pw + X2pw + X3pw + X4pw
IDpasswork =~ IDwork

IDwork ~~ 0* IDwork

IPFentrpass =~ X1ep + X2ep + X3ep
IDentrpass =~ IDdev

IDdev ~~ 0*IDdev

Passforwork<~ 1*IPFpasswork + 1*IDpasswork

Entrpass<~ 1*IPFentrpass + 1*IDentrpass

There was a similar question on the SEMNET just today or yesterday, and you might want to search the archives on that mailing list for the responses.

Mikko

--
Mikko Rönkkö
Postdoctoral Researcher
Institute of Strategy and Venturing, Aalto University
mikko....@aalto.fi
+358 (0)50 387 8155

http://tuta.aalto.fi/en/