Validate or Refute a hypothesis by comparing 4 models

Frequently asked questions about PLS path modeling.
Post Reply
etudiant
PLS Junior User
Posts: 8
Joined: Fri Mar 17, 2017 12:39 pm
Real name and title: M Abdo

Validate or Refute a hypothesis by comparing 4 models

Post by etudiant » Mon Jun 25, 2018 3:29 am

Hello,
I created four models (with same data set)

Code: Select all

 
    Gv1============>\
                     LA ========>PERF
    Gv2============>/
 
 PERF R Square=0,158
 LA==>PERF path coef=0,398	
 

Code: Select all

 
    Gv1============>\
                     LB ========>PERF
    Gv2============>/
 
 PERF R Square=0,281
 LB==>PERF path coef=0,530	
 

Code: Select all

 
    Gv1============>\
                     NLA ========>PERF
    Gv2============>/
 
 PERF R Square=0,295
 NLA==>PERF path coef=0,543
 

Code: Select all

 
    Gv1============>\
                    NLB ========>PERF
    Gv2============>/
 
 PERF R Square=0,328
 NLB ==>PERF path coef=0,573	
 


I want to validate or refute this hypothesis: "LA positively influences PERF more than the others (LB, NLA, NLB)"
How can I do that? with the comparison of R square of the 4 models ? with the comparison of Path coef? ...?

jmbecker
SmartPLS Developer
Posts: 821
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Validate or Refute a hypothesis by comparing 4 models

Post by jmbecker » Mon Jun 25, 2018 6:56 pm

I would model all constructs in one model. Then you can compare the path coefficients from each construct (LA, LB, NLA, NLB) to PERF.

You can then use the boostrapping approach by Rodríguez-Entrena, Schuberth, and Gelhard (2018) to compare the coefficients and assess if your hypothesis is supported.

Rodríguez-Entrena, M., Schuberth, F., & Gelhard, C. (2018). Assessing statistical differences between parameters estimates in Partial Least Squares path modeling. Quality & quantity, 52(1), 57-69.
Dr. Jan-Michael Becker, University of Cologne, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Ja ... v=hdr_xprf
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de

etudiant
PLS Junior User
Posts: 8
Joined: Fri Mar 17, 2017 12:39 pm
Real name and title: M Abdo

Re: Validate or Refute a hypothesis by comparing 4 models

Post by etudiant » Tue Jun 26, 2018 1:56 am

I would model all constructs in one model. Then you can compare the path coefficients from each construct (LA, LB, NLA, NLB) to PERF.
I can't model all constructs in one model : I have problems of validity ==> viewtopic.php?f=5&t=15982&p=26430#p26430

etudiant
PLS Junior User
Posts: 8
Joined: Fri Mar 17, 2017 12:39 pm
Real name and title: M Abdo

Re: Validate or Refute a hypothesis by comparing 4 models

Post by etudiant » Fri Jun 29, 2018 3:22 pm

Any help please !

Post Reply