How to compare models of dependent samples

Frequently asked questions about PLS path modeling.
Post Reply
Mari
PLS Junior User
Posts: 1
Joined: Mon Sep 26, 2016 4:04 pm
Real name and title: Mirja Hollmann

How to compare models of dependent samples

Post by Mari »

After googling with rather limited success I hope for a hint here:
I want to make an analysis that is analogous to a Multigroup Comparison. The difference is that I distinguish not between groups but want to compare models for 2 different dependent variables, where the data was collected on the same particiants thus making the samples dependent.
Concretely, I built two compositionally aquivalent models to predict 2 different types of behavior and want to test if the path coefficients differ. Given the data was collected on the same people the models are not independent of each other, thus I wonder how I would adequately test if differences between the path coefficients for model 1 and model 2 are significant?


Thanks a lot!
jmbecker
SmartPLS Developer
Posts: 1129
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: How to compare models of dependent samples

Post by jmbecker »

There is a nice article on how to conduct difference test between individual parameters in a model in PLS: Rodríguez-Entrena, M., Schuberth, F., & Gelhard, C. (2016). Assessing statistical differences between parameters estimates in Partial Least Squares path modeling. Quality & Quantity, 1-13.
Dr. Jan-Michael Becker, University of Cologne, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
olearus
PLS Junior User
Posts: 5
Joined: Sun Jul 10, 2016 10:47 pm
Real name and title: Macario Rodríguez Entrena, Phd

Re: How to compare models of dependent samples

Post by olearus »

Mari wrote:After googling with rather limited success I hope for a hint here:
I want to make an analysis that is analogous to a Multigroup Comparison. The difference is that I distinguish not between groups but want to compare models for 2 different dependent variables, where the data was collected on the same particiants thus making the samples dependent.
Concretely, I built two compositionally aquivalent models to predict 2 different types of behavior and want to test if the path coefficients differ. Given the data was collected on the same people the models are not independent of each other, thus I wonder how I would adequately test if differences between the path coefficients for model 1 and model 2 are significant?


Thanks a lot!
Dear Mari,
As the Dr. Becker pointed out, the following paper can help to solve your issue: Assessing statistical differences between parameters estimates in Partial Least Squares path modeling. The good news is that a couple of Excel sheets are available upon request to the authors.
Best regards
MAC
tasneem
PLS Junior User
Posts: 1
Joined: Sat Jul 25, 2020 2:13 pm
Real name and title: Tasneem

Re: How to compare models of dependent samples

Post by tasneem »

Even I have the same issue.
I have to compare two different models on smart pls.
Can I compare on the basis of tvalues?
Post Reply