SmartPLS 2 & 3 giving diff. R2, PathCo for the same project

Report and discuss bugs of SmartPLS 2. To report bugs of the new SmartPLS 3 software, create a ticket in our helpdesk.
Post Reply
MMVS
PLS Junior User
Posts: 6
Joined: Tue May 12, 2015 6:17 am
Real name and title: Mark

SmartPLS 2 & 3 giving diff. R2, PathCo for the same project

Post by MMVS » Tue May 12, 2015 11:34 am

I ran the Simple Model.splsm that came with Corporate Reputation Moderation project file with Full_Data_contract.csv [125 records] on SmartPLS 2 and I got similar results from the book. I then ran the same file from the project on SmartPLS 3 and achieved different results.. R2 different.. Path Coefficients different.. Just wondering how or why the report output is not the same. Bug?

jmbecker
SmartPLS Developer
Posts: 879
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by jmbecker » Wed May 13, 2015 8:15 am

Usually it a problem in the data settings.

e.g.,
viewtopic.php?f=3&t=3461
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

User avatar
Hengkov
PLS Super-Expert
Posts: 1619
Joined: Sun Apr 24, 2011 10:13 am
Real name and title: Hengky Latan
Location: AMQ, Indonesia
Contact:

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by Hengkov » Thu May 14, 2015 4:50 pm

Hi,

The errors generally also occurs in charging bootstrapping on SmartPLS 2 (number of samples and number of resamples).

Regards,

MMVS
PLS Junior User
Posts: 6
Joined: Tue May 12, 2015 6:17 am
Real name and title: Mark

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by MMVS » Fri May 15, 2015 6:22 am

Hengkov wrote:Hi,

The errors generally also occurs in charging bootstrapping on SmartPLS 2 (number of samples and number of resamples).

Regards,
For this particular exercise.. I just imported the project files into v2 and v3, selected the Full_Data_contract.csv case file and then ran a simple PLS Algorithm procedure.
The SmartPLS 2 results seems to be consistent with the one in the book while version 3 is giving me another.
I didn't really change any settings for the PLS Algorithm procedure on v2 and v3.

jmbecker
SmartPLS Developer
Posts: 879
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by jmbecker » Fri May 15, 2015 11:58 am

Please try to change the number format (i.e., decimal sign; dot or comma) in PLS3 data file and see if result change.
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

User avatar
Hengkov
PLS Super-Expert
Posts: 1619
Joined: Sun Apr 24, 2011 10:13 am
Real name and title: Hengky Latan
Location: AMQ, Indonesia
Contact:

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by Hengkov » Fri May 15, 2015 1:45 pm

Hi,

It may be true that you are not inconsistent data formats when used in SmartPLS 2 and 3. To make all this clear, I provide the data and the results of SmartPLS 2 and 3 at the links below, you can try it. The results are 100% the same and hopefully there are no more questions like this.

https://www.dropbox.com/s/i1vt7vajdwiid ... .docx?dl=0

https://www.dropbox.com/s/m8zxop37a1tqs ... e.csv?dl=0

Greetings,

sbax123
PLS User
Posts: 11
Joined: Wed Sep 16, 2015 4:49 am
Real name and title: Sam B.- PhD Candidate

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by sbax123 » Wed Sep 16, 2015 5:15 am

Hi MMVS:
I am a PLS newbie as well. I am just doing the same Case Study from the 'A Primer on Partial Least Squares Structural Equation Modelling (PLS-SEM) book'. I have just completed the activity (first PLS Algorithm run for the Corporate Reputation project (344 rows). I have no SmartPLS Version 2 to compare with (I am currently using the months free trial of Version 3), but my results are definitely different than the ones in the book.

-R2 for Customer Loyalty .563 (V3) instead of .562 (book figure) could just be a rounding error. But the difference for Customer Satisfaction is .314 (V3) instead of .295 (book figure)
-All other figures are off e.g. .498 (V3) instead of .504 (book figure) between Customer Satisfaction and Customer Loyalty

I have re-run the algorithm using the two different number formats, checked and re-checked the datafile I downloaded for the activity (from the book's website). But cannot figure out what is going on! :(

Could someone please help! I know this is a post from May, but I still haven't been able to find the answer to this. I know that there is no reason why the results should be different, but I am at my wits end with this :( - I am sorry Hengkov, but I want to know why my results are different to the one in the book :(

Any ideas would be cool!! :)

sbax123
PLS User
Posts: 11
Joined: Wed Sep 16, 2015 4:49 am
Real name and title: Sam B.- PhD Candidate

Re: SmartPLS 2 & 3 giving diff. R2, PathCo for the same proj

Post by sbax123 » Thu Sep 17, 2015 3:19 am

I have it sorted. All constructs are reflective in this model, there are no formative constructs (i.e. the arrows were going in the wrong direction) - rookie mistake! :(

But I think this happened when I clicked on one connector and chose to 'change mode of measurement model', and assumed that this referred to only the one connector. In reality it actually changes the whole lot of connectors to the indicators.

But at lease we have an answer to this question now. Check which way your arrows are going! - do they match the ones in the book?!!

Thanks to everyone who read my crazy ramblings :(. But I'm hoping the next person doesn't spend as long as I did on this.

Cheers,
Sam

Post Reply