PLS PREDICT RESULT

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
norhafiz83
PLS Junior User
Posts: 3
Joined: Sun Nov 25, 2018 9:59 am
Real name and title: Nursaadatun Nisak binti Ahmad

PLS PREDICT RESULT

Post by norhafiz83 »

Hi PLS experts,

I have conducted a PLS Predict analysis test on the final data of my study (there are two dependent variables in my study).
The results of the analysis showed that one of dependent variables is negative in value. Thus, here I would like to ask three questions
with regard to the result:

1. Why do the negative result happen?
2. How do we solve it?
3. What is the best way for me to report it in my thesis?

Thank you in advance for your attention and for your sharing of information.

NISA
User avatar
cringle
SmartPLS Developer
Posts: 818
Joined: Tue Sep 20, 2005 9:13 am
Real name and title: Prof. Dr. Christian M. Ringle
Location: Hamburg (Germany)
Contact:

Re: PLS PREDICT RESULT

Post by cringle »

Please upload a screenshot of your model (with results) and the PLSpredict results table.
norhafiz83
PLS Junior User
Posts: 3
Joined: Sun Nov 25, 2018 9:59 am
Real name and title: Nursaadatun Nisak binti Ahmad

Re: PLS PREDICT RESULT

Post by norhafiz83 »

Dear Prof,

I have attached herewith screenshot of my model (with results) and the PLS Predict results table (as below).

LV Prediction Summary

RMSE MAE Q2
CEKP 0.507 0.365 0.008
DPL 0.513 0.372 0.040
EMMT 0.486 0.363 -0.072


Thank you in advance for your attention and for your sharing of information.
Attachments
FRAMEWORK- MODEL.png
FRAMEWORK- MODEL.png (51.48 KiB) Viewed 4809 times
jmbecker
SmartPLS Developer
Posts: 1281
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: PLS PREDICT RESULT

Post by jmbecker »

A negative value for Q² means that you have less predictive power than simply using the average value of all observations.

That may have several reasons. Your measurement model might not be good enough or your structural model is weak. Usually, the indicator level predictive results also guide you in a direction.
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
norhafiz83
PLS Junior User
Posts: 3
Joined: Sun Nov 25, 2018 9:59 am
Real name and title: Nursaadatun Nisak binti Ahmad

Re: PLS PREDICT RESULT

Post by norhafiz83 »

Dear Dr,

Thank you for the feedback. Anyway how should i interprete it in my thesis.

Thank you in advance for your attention and sharing of information.
jmbecker
SmartPLS Developer
Posts: 1281
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: PLS PREDICT RESULT

Post by jmbecker »

As I said before: Your model does not have sufficient predictive power. Using a simple mean of the latent variable is more predictive than your model. This shows that you model is not good in predicting endogenous indicators.
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
Post Reply