SV: SV: Question about structural equation modeling
Posted by
Raimundas Vaitkevicius on
Oct 11, 2013; 10:34am
URL: http://spssx-discussion.165.s1.nabble.com/Question-about-structural-equation-modeling-tp5722477p5722499.html
I guess you used CFA approach?
You developed a new model (factor structure) using your sample. How can you be sure that your model is correct in some broader sense, not only in your sample? Generally, validation on another sample makes sense, however, I think, it should not necessarily been done in the same study and described in the same article. I guess that you simply didn't have such a goal at the present - to validate your model in another sample. Maybe it would be
sufficient only to explain this to the reviewer properly, probably making corrections to the manuscript in order to clarify goals and methods of your study.
Från: stace swayne <[hidden email]>
Till: [hidden email]
Skickat: torsdag, 10 oktober 2013
17:23
Ämne: Re: SV: Question about structural equation modeling
Thank you for the feedback. I actually didn't develop the depression scale. The scale was used in a large study without prior psychometric testing. In my paper I establish the factor structure and then I did invariance testing across women and ethnic groups within the sample.
On Thursday, October 10, 2013 8:23 AM, Raimundas Vaitkevicius <[hidden email]> wrote:
If this is a new scale and/or model, it would be nice to validate it. You cannot use the same sample for developing the model and validating it.
Maybe you could say that your goal was to establish the model, you leave it's validating to the future research or other researchers.
It all depends on the context and pre-history of the scale/model.
Raimundas
Vaitkevičius
Vytautas Magnus University
[hidden email]
Från: stace swayne <[hidden email]>
Till: [hidden email]
Skickat: torsdag, 10 oktober 2013 3:20
Ämne: Question about structural equation modeling
I received a revise and resubmit on a manuscript focused on the establishment of a measurement model for a depression scale. One of the reviewers asked why I didn't use a validation sample. My question is, is it always necessary to have a validation sample. All opinions are welcomed.