Is there any way to justify my low r-squared? I have good loadings and my relationships are significant, but my r2 are just to low... 0.095, and 0.073.
If I have these low r2 does that mean my model is not good???
Thank you in advance for your help...
low r-squared
-
jesuscarmona
- PLS Junior User
- Posts: 1
- Joined: Thu Oct 08, 2009 11:58 pm
- Real name and title:
-
Diogenes
- PLS Super-Expert
- Posts: 899
- Joined: Sat Oct 15, 2005 5:13 pm
- Real name and title:
- Location: São Paulo - BRAZIL
- Contact:
Hi,
we should assess the measurement model: as the loadings are good, probably, we have good AVE and reliability.
In the seconde step, we should assess the structural model (low R2).
It is not a estimation "problem", it is a research result. The LV were measured properly, but they have low relation on each other.
It is recommended to explain these results from the context (empirical issues).
Best regards,
Bido
we should assess the measurement model: as the loadings are good, probably, we have good AVE and reliability.
In the seconde step, we should assess the structural model (low R2).
It is not a estimation "problem", it is a research result. The LV were measured properly, but they have low relation on each other.
It is recommended to explain these results from the context (empirical issues).
Best regards,
Bido
-
iris_afandiphd
- PLS Senior User
- Posts: 28
- Joined: Thu Jun 24, 2010 5:38 am
- Real name and title:
-
christian.nitzl
- PLS Expert User
- Posts: 248
- Joined: Sat Jul 25, 2009 1:34 pm
- Real name and title:
Hey,
In following article on page 504 you can find an example for an explanation of a relative low R2 (<0,30) in sense of the empirical context as prof. diogenes recommended:
Eberl, Markus (2010): An Application of PLS in Multi-Group Analysis, in: Esposito Vinzi, V./Chin, W./Henseler J./Wang, H. (Hrsg.): Handbook of Partial Least Squares: Concepts, Methods and Applications, Berlin, S. 487-514.
In my opinion low R2 should be seen very critical in PLS because one strength of PLS over LISREL is his power of explanation. Furthermore small R2 could be occur by chance. In an questionnaire there is also a problem with common method variance. There could be a small R2 because people answer likert scales in an special manner.
Best regards,
Christian
In following article on page 504 you can find an example for an explanation of a relative low R2 (<0,30) in sense of the empirical context as prof. diogenes recommended:
Eberl, Markus (2010): An Application of PLS in Multi-Group Analysis, in: Esposito Vinzi, V./Chin, W./Henseler J./Wang, H. (Hrsg.): Handbook of Partial Least Squares: Concepts, Methods and Applications, Berlin, S. 487-514.
In my opinion low R2 should be seen very critical in PLS because one strength of PLS over LISREL is his power of explanation. Furthermore small R2 could be occur by chance. In an questionnaire there is also a problem with common method variance. There could be a small R2 because people answer likert scales in an special manner.
Best regards,
Christian
low r square
Dear Prof. Christian and Prof. Bido
Thank you for solving my all the queries. with your guidance i was able to do PLS modelling
I have Two models with sample 1 and sample 2 done in two regions with similar characteristics. When i did PLS modeling separately for these two samples R square is .35 for sample 1( sample size 148, total indicators=40, total constructs =11) and for sample 2 R square came as 31.( sample size 430, total indicators=40, total constructs/factors =12) .
Now when i combined the data of these two samples and did the combined analysis, the significance of my factors and their sign are almost similar to the findings in sample 1 and sample 2, BUt in this case my R square is 0.27. I am not able to get my I am getting this Rsquare low.
What can i conclude from this.
Thanks
Ruchi
Thank you for solving my all the queries. with your guidance i was able to do PLS modelling
I have Two models with sample 1 and sample 2 done in two regions with similar characteristics. When i did PLS modeling separately for these two samples R square is .35 for sample 1( sample size 148, total indicators=40, total constructs =11) and for sample 2 R square came as 31.( sample size 430, total indicators=40, total constructs/factors =12) .
Now when i combined the data of these two samples and did the combined analysis, the significance of my factors and their sign are almost similar to the findings in sample 1 and sample 2, BUt in this case my R square is 0.27. I am not able to get my I am getting this Rsquare low.
What can i conclude from this.
Thanks
Ruchi
-
christian.nitzl
- PLS Expert User
- Posts: 248
- Joined: Sat Jul 25, 2009 1:34 pm
- Real name and title:
Dear Sir
Thanks a lot for solving the query
But reporting this R square of 0.27 for the combined analysis and giving the explanation of that there may be group differences will be justifiable ?
Please let me know as I am writing my final report and will be presenting my analysis.
Thanks
Regards
Ruchi
Thanks a lot for solving the query
But reporting this R square of 0.27 for the combined analysis and giving the explanation of that there may be group differences will be justifiable ?
Please let me know as I am writing my final report and will be presenting my analysis.
Thanks
Regards
Ruchi