Saturday, February 8, 2020
Full latent growth and its use in PLS-SEM: Testing moderating relationships
The article below explains how one can conduct a full latent growth analysis, in the context of structural equation modeling via partial least squares (PLS-SEM). This type of analysis can be viewed as a comprehensive analysis of moderating effects where the moderating variable is “latent”, not “disrupting” the model in any way.
Kock (2020). Full latent growth and its use in PLS-SEM: Testing moderating relationships. Data Analysis Perspectives Journal, 1(1), 1-5.
A link to a PDF file is available ().
Abstract:
A full latent growth analysis, in the context of structural equation modeling via partial least squares (PLS-SEM), can be viewed as a comprehensive analysis of moderating effects where the moderating variable is “latent”, not “disrupting” the model in any way. In this paper we illustrate such an analysis employing WarpPLS, a leading PLS-SEM software tool.
Subscribe to:
Post Comments (Atom)
5 comments:
Dear Prof. Ned Kock,
Thank you for sharing the article. May I request you to kindly advise on the following:
1. How to present latent growth results graphically to indicate moderation effect as done using other methods.
2. List of ABDC journals where moderation effect has been presented using latent growth method.
Thanks and regards,
Sanjeet
Hi Sanjeet. I would suggest something along the lines of the presentation in the article linked. It is a new method; you may try "latent growth" and "warppls" in Google Scholar.
Hello Ned,
I am using secondary data. I am testing the moderating influence of "quality certification" (dichotomous 1/0) on the relationship between Environmental Commitment (latent with 2 binary items) and process Innovation (binary). The moderation analysis shows a significant effect at 5% (beta = 0.08), however, when I try assessing this with a MGA approach, the significance of the difference is only at 10% (beta of full latent growth = 0.064). Which method shall I prioritise here please? Can I stick to the moderation analysis only? Note that NO collinearity issues exist with or without the inclusion of the moderation. Your help is much appreciated. Many thanks
I would use the method that yields the highest coeff., as long as there are no collinearity issues. Keep in mind, however, that no support for a hypothesis is a finding by itself.
Thank you so much Ned. That's very helpful!!
Post a Comment