#SmartPLS4

The session focuses on how to assess the predictive power of the model using PLSPredict in SmartPLS4. The sessions starts with the concept of Predictive power, the theory behind it, the process of predictive power assessment, and finally practical illustration of how to assess predictive power using PLS Predict in SmartPLS4.
#smartpls #smartpls4 #predictivepower #predictivevalidity #plssem #sem #TrainingSmaple #HoldoutSample #OutofSamplePredictivePower

[@asmansur22]

Highly contributive. jazakAllahu kairan.

[@researchwithfawad]

Thanks. Ameen. I am glad you liked it.

[@seyedjafarmousavi6973]

Thank you for another great tutorial.

[@researchwithfawad]

Thanks. I am glad you liked it.

[@faizakhalidkhan7942]

Assalam o akailum.. sir you are doing great job. These videos are really helpful me and other researchers.

[@researchwithfawad]

Glad to hear that

[@2011lobito]

Felicitaciones muy clara la explicación

[@researchwithfawad]

Thanks. I am glad you liked it.

[@Alhamzah_F_Abbas]

Thank you

[@researchwithfawad]

You're welcome

[@yarnjamin]

Thank you very much, Prof for the tutorial. Your videos have supported my research tremendously. I have a question here, I did the PLSpredict calculation but only the dependent result was shown in the report. I am using a higher-order model with a reflective-reflective construct, a reflective-formative construct and a moderation variable. My concern is only dependent items were shown in the report. Can you please help me to rectify the issue? Thank you in advance Prof!

[@researchwithfawad]

I am glad you liked it. Can you post the question with screenshot on SmartPLS forum. This will help in better understanding of the issue.

[@zeinabazein3752]

can you tell the order of the result tables in the structural model of two stage disjoint method , ? is it 1. path analysis, 2. direct& total effect and indrct 3. specific indirect indirect , 4. R^2 f^2 and Q^2 5 RMSE

[@chefberrypassionateresearcher]

Sir, shall i do this once for complete data set or for all the data groups. Later i will do MGA

[@researchwithfawad]

If your focus is on each group, do for each group

[@rashialagh283]

Sir pls predict should be reported of latent constructs valued model or original modl

[@researchwithfawad]

Please watch the video again for details.

[@z61237]

Hello sir
I have a total of 23 endogenous indicators, out of which 12 indicators have high LM. According to the rule, the model has low predictive power. Can we mentioned it in the paper

[@researchwithfawad]

I rather not report it at first.

[@farahfarhana4014]

Salam Dr. if the number of indicators having lower PLS-SEM-RMSE are lesser by 2, can we delete 2 indicators with high prediction errors to make it equal so that predictive power move from low to medium?

[@researchwithfawad]

WAS. You may if those two items do not affect how you ha e conceptulized the construct.

[@farahfarhana4014]

can you explain more? example : I have 6 items under the construct. So I remove 2 but I would like to know the basis of choosing to remove that items, is it due to higher prediction errors or any other way to judge?@@researchwithfawad

[@researchwithfawad]

Does deletion of two items invalidate how you have defined the concept

[@user-bp5mg7bx1m]

As my all predictive powers are red except one...how to solve this issue
Plz help

[@researchwithfawad]

Overall, you can say it is acceptable.

[@rashialagh283]

Sir do we need to report both Qsquare and rmse for predictive power

[@rashialagh283]

And do we need to report rmse of each item

[@researchwithfawad]

Report Q Square, you do not need to report fits when using SmartPLS

[@rashialagh283]

Ok thnku sir

[@rashialagh283]

Sir I want to ask is Qsquare and pls predict different