Request for Guidance on robustness checks using PLS SEM

[Tushar saha]

Dear Neeraj Sir, Professors and Scholars,

The query seeks clarification on the procedure for determining the number of path combinations.

Considering the model, twenty-two bootstrapped total paths are presented. However, the current research tests only sixteen hypotheses. The hypotheses consist of ten direct paths and six mediating paths. The analyses are conducted using R Studio.

Inquires

Part 1: Checking the Nonlinear Effects using Ramsey’s RESET

For this purpose, path combinations are framed according to the number of incoming arrows that point toward DV. The combinations are as follows:

> resettest(EEE ~ GGG + HHH + FFF, power = 2:3, type= "regressor", compositscore_004)

> resettest(DDD ~ EEE + FFF, power = 2: 3, type= "regressor", compositscore_004)

> resettest(CCC ~ DDD + EEE, power = 2:3, type= "regressor", compositscore_004)

> resettest(AAA ~ DDD + EEE, power = 2:3, type= "regressor", compositscore_004)

> resettest(BBB ~ DDD + EEE, power = 2:3, type= "regressor", compositscore_004)

Subsequently, each combination is checked between DV and IDV levels, such as

HHH

HHH2= HHH^2

HHH3= HHH^3

Model_HHH= summary(lm(EEE ~ HHH+ HHH^2+ HHH^3))

Model_HHH1= lm(EEE ~ HHH+ HHH^2+ HHH^3)

effectsize::cohens_f(Model_HHH1)

.

.

And so on.

Inquiry 1: Is the selection of five distinct paths for Ramsey’s RESET testing appropriate based on the model?

Part 2: Endogeneity test using the Gaussian copula approach

In this regard, please suggest the initial steps to map copula terms within our model, particularly given the interrelationship positioning among variables.

Inquiry 2: How do we begin fitting the copula terms?

Inquiry 2a: Could we consider final DVs, specifically AAA, BBB, and CCC, along with their corresponding IDVs?

Inquiry 2b: Otherwise, proceed with the submodel that represents only our hypotheses, particularly the mediation level?

 Inquiry 2c: Or, start with how many incoming arrows connect toward the DV? For example:

Part 3: To address CMV, the marker variable was incorporated into our questionnaire instrument. After deducting the effect of the marker, the resulting composite latent values were used for both analyses above.

Inquiry 3: Is it correct to proceed with both analyses using the composite latent values?

 

Thank you for being so supportive.

Regards,

Tushar

[Neeraj Kaushik]

Dear Tushar

You've analyzed correctly

1. Non-linearity is checked only for the DVs

You've checked all paths that point towards the DVs, which is appropriate for Ramsey’s RESET testing.

However do not check the square or cubic effect individually as its an endless process.

Decide based on Ramsey's test.

2. You have written the equations correctly for the Endogeneity test using the Gaussian copula approach.

Now run each one by one.

3. Ideally, the marker variable should not be correlated with any of the DV or IDV.

If so, no further analysis is needed.

If there is any correlation, then you shd use marker variable, like a control avr and nullify its effect.

Best wishes  

--
The members of this group are expected to follow the following Protocols:
1. Please search previous posts in the group before posting the question.
2. Don't write the query in someone's post. Always use the option of New topic for the new question. You can do this by writing to dataanaly...@googlegroups.com
3. It’s better to give a proper subject to your post/query. It'll help others while searching.
4. Never write Open-ended queries. This group intends to help research scholars, NOT TO WORK FOR THEM.
5. Never write words like URGENT in your posts. People will help when they are free.
6. Never upload any information about National Seminars/Conferences. Send such information
in personal emails and feel free to share any RESEARCH-related information.
7. No Happy New Year, Happy Diwali, Happy Holi, Happy Birthday, Happy Anniversary, etc. allowed in this group.
8. Asking or sharing Research Papers is NOT ALLOWED.
9. You can share your questionnaire only once.
---
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAMn1K5Ar%3DQ4jH%3Dp12Qvx%2BqCCXZOF0rXLjW6P_dRMtGed50aBuw%40mail.gmail.com.

[Tushar saha]

I sincerely appreciate, Sir, your review and confirmation of the inquiries.

Thank you, Sir, and the DataAnalysis group once again.

Tushar

To view this discussion visit https://groups.google.com/d/msgid/dataanalysistraining/CAAd%3Dc8PfXKE8wgq-uA0poLVmYWX8YEMxGa0YuVbzCrz35Bxubg%40mail.gmail.com.