path analysis? is model fit required?

96 views
Skip to first unread message

Mugaahed Saleh

unread,
Jun 17, 2021, 10:59:29 PM6/17/21
to lavaan
Dear experts,
I want to conduct a path analysis model, (Diagram attached).
I am just confused how to move towards that model (path model without the single items of the questionnaire - aka structure model).
I mean, should I conduct CFA and get model fit first?
Should calculate model fit for the path model?

The questionnaire I used is standardized questionnaire, I just wanted to see the mediating roles.

Any help is appreciated,
kindest regards,
Muug
Path diagram.jpg

Pat Malone

unread,
Jun 18, 2021, 8:57:28 AM6/18/21
to lav...@googlegroups.com
Muug,

The difference between path analysis and SEM is that in SEM you estimate the measurement model (the CFA, more or less) simultaneously with the structural model (the path model, more or less).

What I think you're suggesting is running the CFA, creating observed scores based on it, then entering them into the path analysis. This has a few drawbacks, one of which is a well-fitting CFA is much less constrained than the model implied by taking observed scores, unelss you go out of your way to impose them. One implication is that the measurement model you're actually putting into use (sum scores, or whatever it is), is not actually the model tested by the CFA.

There are other important differences, but I think that's one of the big ones for your question: You won't actually be testing the measurement model you're using.

That said, if you are willing to accept previous measurement work as valid without examining it in your data, you can still go straight to path analysis. SEM has other advantages, including distilling the unreliability and minimizing bias caused by it, but skipping the measurement step is common in many fields.

Pat

--
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/23c9b094-be3c-4aee-96a5-e41c1e22cf62n%40googlegroups.com.


--
Patrick S. Malone, PhD
Sr Research Statistician, FAR HARBΦR
This message may contain confidential information; if you are not the intended recipient please notify the sender and delete the message.

Mugaahed Saleh

unread,
Jun 19, 2021, 12:28:16 PM6/19/21
to lavaan
Thank you so much for your reply. 
This clears my doubt very well.

Kindest regards.
Muug

Reply all
Reply to author
Forward
0 new messages