mediation effect

This forum is closed, and read-only.
Locked
lama
PLS Junior User
Posts: 2
Joined: Fri Oct 22, 2010 2:45 pm
Real name and title:

mediation effect

Post by lama »

Hi everyone


I am trying to test the mediation effect of some veriables in the model which am trying to test. and I realized that there are some mediation effect there. I know I can test the simple mediation using the bootstrapping outpot, but when I read in the net some authors do not recommand using simple mediation with SEM and they recommend using multiple mediation methods.

there are Preacher, K. J., & Hayes, A. F. macros which help in calculating multimediation but this macro is design to be used with SPSS and SAS.

I have read apaper in which the auther mentioned that he used the outpot of latent variable score from SmartPLS as an input to use in calculating multiple mediation by indirect macro in SPSS.

my question is...

if there any other way to test the multimediation effect using SmaartPLS
would you help me in it


regards
vonbergh
PLS Expert User
Posts: 60
Joined: Tue Jan 26, 2010 10:06 pm
Real name and title:

Post by vonbergh »

Hi Lama,

Checking for mediation in SmartPLS is more straightforward than in SPSS and SAS. How to do it:
(1) Create all relationships between independent variables (IV) and dependent variables (DV) according to your model
(2) Create the mediating variables (MV)
(3) Create the mediating relationships between the IV and MV and between MV and DV. Now you should have "triangles" between IV, MV and DV.
(4) Estimate the path coefficients
(5) Perform the bootstrap.
(6) Assess which relationships are significant. If all relationships between an IV, a MV and a DV are significant, then the relationships IV -> DV is mediated by MV. The path coefficients imply to what degree.

Good luck
Dennis
saloualabib
PLS Junior User
Posts: 2
Joined: Sat Sep 01, 2012 7:52 pm
Real name and title:

Re: mediation effect

Post by saloualabib »

lama wrote:Hi everyone


I am trying to test the mediation effect of some veriables in the model which am trying to test. and I realized that there are some mediation effect there. I know I can test the simple mediation using the bootstrapping outpot, but when I read in the net some authors do not recommand using simple mediation with SEM and they recommend using multiple mediation methods.

Hi,
Please , I need some articles about mediator variable in pls

Thanks



there are Preacher, K. J., & Hayes, A. F. macros which help in calculating multimediation but this macro is design to be used with SPSS and SAS.

I have read apaper in which the auther mentioned that he used the outpot of latent variable score from SmartPLS as an input to use in calculating multiple mediation by indirect macro in SPSS.

my question is...

if there any other way to test the multimediation effect using SmaartPLS
would you help me in it


regards
vonbergh
PLS Expert User
Posts: 60
Joined: Tue Jan 26, 2010 10:06 pm
Real name and title:

Post by vonbergh »

Hi Zaloua, you may consider reading Zhao, Lynch and Chen (2010) Reconsidering Baron and Kenny: Myths and truths about mediation analysis, Journal of Consumer Research, 37(2), 197-206, available at http://www.jstor.org/stable/10.1086/651257 greetings, Dennis
Juan Pablo ROMAN-CALDERON
PLS Junior User
Posts: 1
Joined: Fri Dec 16, 2011 9:12 am
Real name and title:

Bootstrap confindence intervals according to Chin (2010)

Post by Juan Pablo ROMAN-CALDERON »

Hi,

I'm trying to calculate the significance of two indirect effects within my PLS structural model. I followed the two-step approach suggested by Chin (2010):

-"Perform N bootstrap resampling (e.g., 1,000 resamples) and explicitly calcu- late the product of direct paths that form the indirect path being assessed".
-"Estimate the significance using either percentile bootstrap or bias corrected corrected bootstrap..."

A reviewer told me that my results are illogical (.12-.13) (-.21 - -.20). He said that this is practically impossible as that means that the standard errors are essentially zero (,002).

Could anyone explain me what does all that mean? I'm not very familiar with statistics.

Any suggestions on how to respond to this comment and/ or another approach to calculate the bootstrapped CIs?

Thanks a lot,

JP

Chin, W. W. 2010. How to Write Up and report PLS Analyses. In V. E. Vinzi, W. W. Chin, J. Henseler & H. Wang (Eds.), Handbook of Partial Least Squares (pp. 655-690). Berlin: Springer-Verlag
Locked