[Mediation-information] Reg a doubt in mediation

M Subbiah sisufive5 at yahoo.com
Thu Mar 8 17:38:53 CET 2012


I would like to convey that, we dont have a formal
training in mediation-moderation analysis, and through the reference cited
above and other such materials we are able understand and implement the methods
(in R specifically)

We have attempted the procedure elaborated in (2) for a data
set (pertain to a study on extent of search by students) using 3 independent
variables (IV) and one mediating (M) variable. All IV, Mediating and dependent
variables (DV) are metric. 
 
 Please find the attachment for the diagram depicts the relation  


We have attempted the procedure (in R) as 
b <- lm(MOT ~ TP + PC + PK,
data=MOD3)
c <- lm(EXS ~ MOT+TP + PC + PK,
data=MOD3)
OP <- mediate(b, c, sims=10,boot =
TRUE, treat="PK", mediator="MOT")
 
Your
valuable clarification needed:
        I.            We run three times by replacing the IV in “treat=
“. Is this a right approach to study the mediating effect?
      II.            Please explain how to interpret the outputs
(though we understand the point and interval estimates) ACME, Mediation Effect,
Direct Effect, Total Effect, Proportion of Total Effect via Mediation.

It will be of more useful if you
could help us to clarify and use the methods more reasonably well
 We thank in advance for your advice and
suggestion
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/mediation-information/attachments/20120308/35cec083/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Diag.doc
Type: application/msword
Size: 24576 bytes
Desc: not available
URL: <http://lists.r-forge.r-project.org/pipermail/mediation-information/attachments/20120308/35cec083/attachment-0001.doc>


More information about the Mediation-information mailing list