Standard Error Printed as NA for Single Parameter

[Andrew Grotzinger]

Hello, 

I'm running a model that produces an NA for a single residual variance parameter and lavaan does not print any error messages when running the model. The point estimate for this specific is .0001, but it is freely estimated and not estimated at 0. I'm hoping someone could explain why lavaan prints an NA in this case, and also whether the delta and weight matrix are still appropriate to use for later calculations that we perform when a single parameter is printed to have NA for a SE. Code used to run the model and model output are attached. Thanks very much in advance for your help!

Best, 

  Andrew

[Mijke Rhemtulla]

It looks like this issue is a function of the sample size entered when you fit the model. If you do N = 4 instead of N = 2, the weight matrix is unchanged but that SE goes from NA to a (small) estimated value. 

[Andrew Grotzinger]

This is super informative, thank you! 

[Sassi Maria]

Thank you, so much, for the useful information.

Maria

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/4dd666f4-f1ec-49cd-9a8f-c84e44b85291%40googlegroups.com.

--

-- 
Prof. Maria Sassi
Department of Economics and Management
University of Pavia

President Genius Loci SrL - spin-off accredited by the University of Pavia

Local director Erasmus Mundus Joint Master Degree 
in International Development Studies

Orcid n.0000-0002-6114-6826

Via. S. Felice 5 - 27100 Pavia - Italy
t. +39-0382986465
fax. +39-0382986228
e-mail: maria...@unipv.it
skype: msassi_6