IGSCA Features

[Claude Obadia]

Following the recent article in Managament International Review I have run a model with IGSCA (factors+components). Two problems: 1) I do not get the GFI fit index 2) Moderation analysis does not seem possible with a a mix of factors and components. Thanks in advance for your guidance.

[heungsun.hwang]

Thank you for your inquiries. 

The next version of GSCA Pro will report GFI in IGSCA. We are collecting a few more bugs to fix at the same time. A research publication is not always going in parallel to software implementation, although it would be nice if that is the case. Thank you for your patience. In the meatime, please use SRMR for model evaluation. 

You're right that IGSCA does not provide moderation analysis, which involves both factors and components. This is actually tricky technically. To my knowledge, there is no study about how to create the product term of a component and a factor. I am not yet sure if even the simple, product indicator approach would work in this case. It would be necessary to examine the performance of the product-indicator approach through extenstive simulation studies. Also, I'd prefer to develop a simulatenous approach for estimating the interaction effects and all other parameters at the same time by optimzing a single objective fuction. This would be quite changelling. In any case, I have been thinking of moderation analysis in IGSCA. Of course, it would be wonderful if you are intereted in pursuing this line of research.

Best,

Heungsun