r square =1

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
Carolin Löffler
PLS Junior User
Posts: 2
Joined: Wed Nov 03, 2010 3:58 pm
Real name and title:

r square =1

Post by Carolin Löffler »

I conducted a success factor study. So my dependent variable is sucess and the independet variables are e.g. the company support the project (scale 1-5), qualification of personnel (scale 1-5) and so on. Now - after runnung smart pls all values are sigificant and also logical to interprete. But my r square for the dependent variable success is 1.0.
Is this result reliable or did I do anything wrong?
christian.nitzl
PLS Expert User
Posts: 248
Joined: Sat Jul 25, 2009 1:34 pm
Real name and title:

Post by christian.nitzl »

Dear Carolin,

Your result of R2 = 1.00 is possible but not very realistic for social science. If your R2 = 1.00 is really true I would called it ‘over fitting’.

You should do some checks. How do you measure your depended variable? With the same indicators as the independed variables? If yes, then you get a R2 1.00. There could be also a failure in your data file. Furthermore R2 depends on proportion between the number of the cases and the number of the depended variable. Therefore you should also check the Adjusted R2. Some information how do you can calculate Adjusted R2, you can find here:

viewtopic.php?t=775&highlight=adjusted

I hope this helps!

Christian
Carolin Löffler
PLS Junior User
Posts: 2
Joined: Wed Nov 03, 2010 3:58 pm
Real name and title:

Post by Carolin Löffler »

Thanks! I will check it out!
wati
PLS Junior User
Posts: 2
Joined: Wed Jan 12, 2011 11:56 am
Real name and title:

R2+1

Post by wati »

My research also have R2 1. Thanks for helps..
Post Reply