#SmartPLS4

[@dorothyj1893]

Thank you so much sir for sharing the knowledge for free, ppl like you are very less.

[@researchwithfawad]

Thanks and welcome

[@mercychemutai3987]

Thank you so much sir! Your videos are truly invaluable. I appreciate that you have painstakingly covered every possibility in data analysis. For me it was the point of what to do in absence of a GV, then if one of the LOCs is insignificant. The guidance on reporting is always my favorite part. Blessings!

[@researchwithfawad]

You are most welcome

[@hamahrapeaahsemawi7202]

Thank you very much for your help Sir. your vids help a lot. May Allah always bless you and your family. appreciate much.

[@researchwithfawad]

Alhamdulillah I am glad you liked it.

[@lina.r11]

Hello thank you for the video. May I know what happened if we have one construct with two indicators, with one factor loading is above 1 and the other is below 0.4. VIF is less than 5. Should I delete the items? Also if the model has second and third orders (9 first order, 3 second order, and 1 third order) . will the approach be the same? thank you

[@researchwithfawad]

Please check the VIF values for the indicator. If there is third order, Yes, use the same approach.

[@PatriciaOmordia]

Thank you sir for the knowledge shared. If a construct with three indicators has a Cronbach Alpha of 0.6 (red), composite reliability of 0.8, and AVE > 0.5. What should be done if the cronbach alpha is not improving after deleting an indicator.

[@researchwithfawad]

Thanks for watching. No need to delete any indicators of CR is over 0.70.

[@B.Rights]

I don't think I would have been able to get my analysis for my dissertation done without your videos. Thank you so much. I did not include GV questions in my survey and have seen you and Dr Ringle comment that the Reliability test is not necessary for second order formative model. I really need a reference to help me here, I am very worried about justifying my model without this step. Could you please offer any advice? Thankyou!

[@researchwithfawad]

Thank you for watching. I am glad that the videos helped. Here is the reference Sarstedt, M., Hair Jr, J. F., Cheah, J. H., Becker, J. M., & Ringle, C. M. (2019). How to specify, estimate, and validate higher-order constructs in PLS-SEM. Australasian marketing journal, 27(3), 197-211.

[@B.Rights]

@Research With Fawad Thank you so much for your reply! However, I have already read this article and it notes that the convergent validity (reliability) test is the first step in assessing a reflective-formative construct. I cannot find any published literature that justifies skipping this step.

[@B.Rights]

My apologies, I meant to say 'redundancy' not reliability. As the first step in evaluating the formative HOC.

[@aarti3180]

Hello @@B.Rights How did you manage this, i am facing the same problem. please help

[@B.Rights]

@@aarti3180 I did not include this step. My supervisors were not worried.

[@LijuanHu-s8m]

Dear Prof., in this disjoint two-stage approch of the video, can I know why you put your moderators direct effect to your DV before you run the measurement model? Why is it not the moderating effect of the culture to DV in this model?

[@researchwithfawad]

Since the objective of MM is to assess the constructs, not the effects.

[@LijuanHu-s8m]

Can I understand when we do MM, usually use the direct effect of moderators in our model but not the moderating effect?@@researchwithfawad

[@researchwithfawad]

Yes, dont add moderating effect.

[@souljachiraq4559]

Prof, kindly design a reasonable model and share a data set that are suitable for students to practice all the concept explained in the video series. It will increase the usability of the video series for beginners.

[@researchwithfawad]

Thanks for the suggestions. Hopefully soon.

[@Alhamzah_F_Abbas]

Much appreciated prof.

[@researchwithfawad]

Thanks I am glad you liked it

[@anitalow3628]

dear Sir, for disjoint approach as demonstrated in this video, my further questions are (1) in stage 1, should we report measurement model assessments for all constructs (including the reflectively-measured LOCs as well as other all non-hierarchical constructs (either reflective or formative))? (2) if question 1 is YES, should we report again the measurement model assessment for all non-hierarchical constucts in stage 2?

[@researchwithfawad]

Yes, reliability and validity is assessed at lower level and if HO is formative, assess and report its validity.

[@meinedoraa]

dear sir, if my path coefficient of hoc formative is below 0.7, but the vif is below 5 is it okay to proceed?

[@researchwithfawad]

If you are validating the HOC, see the VIF, Weights, and Loadings.

[@stouts0899]

Big Thanks Dr, very helpful..... Plz Dr could u share with us the template and PowerPoint

[@researchwithfawad]

Pleasure. I am glad you liked it. You can email me on [e-mail] and i will share some documents that may help.

[@anitalow3628]

sir, is this example (internal marketing) published in journal already?

[@researchwithfawad]

The paper is in review.

[@anitalow3628]

@@researchwithfawad noted. thanks.

[@fariya949]

Sir if we don't have global variable and we have already collected the data then what should we do to apply higher order construct

[@researchwithfawad]

Report results without testing for global variable.

[@mansour6629]

Question please: How to proceed with analysis if you dont have a single item for the convergent validity?

[@researchwithfawad]

You can assess VIF, outer weights, and loadings

[@mansour6629]

@@researchwithfawad Thanks! Planning for publication. Would not having "convergent validity" for the formative construct affect publishing?

[@researchwithfawad]

No it wont

[@mansour6629]

@ResearchWithFawad Thanks for your time and effort! Appreciated!