Mediation in PLS SEM

[Pooja26.s]

Dear sir,

This is in reference to the PLS SEM 3 playlist on your Youtube channel which I have been following for development and validation of my model.

Sir, I have three constructs, for which I have applied mediation effect. 

I have obtained results where total effect and indirect effects are significant but direct effect is not significant.

As per your video, I am able to interpret that if direct effect is not there, we could not have mediation, however, in my model Indirect Effect is highly significant.

Does it make my model complete worthless? Or it qualifies to be a case of Full mediation. I am confused....

Can you please guide how should I write the interpretation of the same.

[Neeraj Kaushik]

Dear Pooja

In AMOS as well as SmartPLS, you've to run the model 2 times:

Step-1 Check the effect of IDV on DV in the absence of Mediator

a) If no effect here, no mediation is possible

If sig effect of IDV on DV, then Mediation is possible

Step-2  Check the effect of IDV on DV in the presence of Mediator

b) If sig Indirect effect (i.e. effect of IDV on DV through Mediator), then Mediation is present

c) If sig Direct effect (i.e. effect of IDV on DV), then Partial Mediation present else Ful Mediation present

Best wishes

--
Protocols of this Group:
 
1. Plz search previous post in group before posing the question.
2. Don't write query in someone's post. Always use the option of New topic for the new question. You can do this by writing to dataanalys...@googlegroups.com
3. Its better to give a proper subject to your post/query. It'll help others while searching.
4. Never write Open ended queries. This group intend to help research scholars NOT FOR WORK THEM.
5. Never write words like URGENT in ur posts. People will help them when they are free.
6. Never upload any info about National Seminars/Conferences. Send such info on personal emails. And feel free to share any RESEARCH related info.
7. No Happy New Year, Happy Diwali, Happy Holi, Happy B'day, Happy Anniversary etc. allowed on this group.
8. Few months back there was a facility for asking & sharing the Research Papers. Now there is no provision of asking for the research paper here.
 
Let’s make a better research environment.
---
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 on the web visit https://groups.google.com/d/msgid/dataanalysistraining/12880b58-6d7c-4b61-b0f9-9bf8d8c88312n%40googlegroups.com.

[Fatih ÇELİK]

Dear Pooja,

The Baron & Kenny (1986) approach is a very strict approach. Contemporary mediation analysis is now used more often. It is argued that there may be an mediation effect in the direct effect not being meaningful. Please check the resources below.

Sources:

1. Zhao, X., Lynch, J. G., & Chen, Q. (2010). Reconsidering Baron and Kenny: Myths and Truths About Mediation Analysis. Journal of Consumer Research, 37(2), 197-206. https://doi.org/10.1086/651257

2. Rucker, D. D., Preacher, K. J., Tormala, Z. L., & Petty, R. E. (2011). Mediation Analysis in Social Psychology: Current Practices and New Recommendations. Social and Personality Psychology Compass, 5(6), 359-371.

3. Preacher, K. J., & Hayes, A. F. (2008). Contemporary Approaches to Assessing Mediation in Communication Research. In A. F. Hayes, M. D. Slater, & L. B. Snyder (Ed.), The Sage Sourcebook of Advanced Data Analysis Methods for Communication Research (ss. 13-54). Sage Publications.

4. MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7(1), 83-104. https://doi.org/10.1037/1082-989X.7.1.83

5. Hayes, A. F., & Rockwood, N. J. (2017). Regression-Based Statistical Mediation and Moderation Analysis in Clinical Research: Observations, Recommendations, and Implementation. Behaviour Research and Therapy, 98, 39-57. https://doi.org/10.1016/j.brat.2016.11.001

6. Hayes, A. F. (2018). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach (Second Edition). Guilford Publications.

7. Hayes, A. F., & Preacher, K. J. (2013). Conditional process modeling: Using structural equation modeling to examine contingent causal processes. In R. G. Hancock & O. R. Mueller (Ed.), Structural equation modeling: A second course (2nd ed., ss. 219-266). IAP Information Age Publishing.

[Pooja Sharma]

Dear Neeraj sir,

Sir I am confused after the post as provided  https://www.smartpls.com/documentation/algorithms-and-techniques/mediation

Here, as per the diagram, p1 and p2 are the deciding factor for determining whether mediation analysis is possible or not.

Presence of Direct effect has not been mentioned as a criteria.

Can you please give some insights on the above mentioned post?

Shall be grateful to you

To view this discussion on the web visit https://groups.google.com/d/msgid/dataanalysistraining/CAAd%3Dc8ME5KBFLLJmuhE5n7MLS0a5_VhEwyxzvSSr0046Bk92zg%40mail.gmail.com.

--

Regards,

Pooja Sharma, M.COM , Ph.D.

[pushpendu chand]

Hi Pooja,

Without being specific to your question, generally, the following three conditions should be present for the mediation effect:

Mediation effect analysis is performed subjected to the fulfillment of following preconditions, (1) the direct relationship without mediation effect should be statistically significant; (2) the indirect effect post mediation should also be statistically significant; and (3) If the variance accounted for (VAF) is greater than 80%, it represents full mediation, 20% ≤ VAF ≤ 80%  means partial mediation and VAF<20% indicates there is no mediation effect. 

Preacher KJ, Hayes AF (2004) SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav. Res. Methods,
Instruments, Comput. 36(4):717–731.

Ignore if it doesn't meet your purpose.

regards

Pushpendu 

To view this discussion on the web visit https://groups.google.com/d/msgid/dataanalysistraining/CA%2B%3DB4ynK4re5y4nME2ZSq8NGgktGezwXCrUVa%3DCZEHEYMh37%2Bw%40mail.gmail.com.

[Atul Shiva]

VAF can be applied only in case there do exist any direct effect between IDV and DV

Secondly, Apply Step 1 as guided by Dr Neeraj Sir. Record the Values

Then apply step 2 and check whether the Direct Effect increase or decrease in the presence of MEdiator to report COmplementary or Competitive MEdiation as per Zhao et al, 2010.

In PLS SEM, you must have observed that in case there is no direct relationship between IDV and DV, still the results are provided. Why so? As the process of bootstrapping is applied, one can investigate the Indirect Effects for Mediator with the help of p-values. You can repot whether mediation is significant or not.

In my opinion, you must read paper of Mediation by Prof T Ramayah from Sarawak Society, Malaysia

To view this discussion on the web visit https://groups.google.com/d/msgid/dataanalysistraining/CAOWONsnN7ocYu%2B75tj88JObKaa3LswFJ1Ny%2BkcPcpxS%2B%2B6gTXg%40mail.gmail.com.

--

Atul Shiva
Assistant Professor in Commerce
Chandigarh University
atul...@gmail.com
8427001687

[Drmeenakshi Gandhi]

Dear Sir

Attached is the paper

Please guide if this is the one yu are referring to 

Wishes and regards

Dr. Meenakshi Gandhi

Professor - Entrepreneurship & Marketing, 

Website: www.meenakshigandhi.com

Linkedin profile: https://www.linkedin.com/in/drmeenakshigandhi/

Contact No. +91-9811507303

School of Business Studies,Vikekananda Institute of Professional Studies 

Affiliated to Guru Gobind Singh Indraprastha University, Delhi

 

Research Interests - Women Entrepreneurship, Marketing, Women Empowerment & Leadership

"I in every individual must stand for innovation, not for inhibition or imitation. It is better to fail in originality than to succeed in imitation." 

- Raghunath Mashelkar (Chairperson - National Innovation Foundation)

 

To view this discussion on the web visit https://groups.google.com/d/msgid/dataanalysistraining/CAHmXBTJkvuVf%2B1efvq%3D31fx760JhVo35v7pT-GhHOfr9MPzgdg%40mail.gmail.com.