Hi all, this is on mediation.
I have 3 IVs (Let's say it is Commitment, Leadership, Resources) 1 MV (Training) and 3 DVs (Organization citizenship, trust and sharing) which in turn will affect firm's financial performance.
As mentioned by Baron and Kenny, to establish whether a variable can be consider as a mediator, 3 conditions must be met:
a) Significant relationship between IV and MV
b) Significant relationship between MV and DV
c) Significant relationship between IV and DV with the absence of MV
My questions:
a) In modeling the relationship in PLS, do I model it individually meaning:
Commitment (DV) to Organization citizenship to Financial Performance
and find out the t-stats or I put all the variables in one model without
the MV to find out whether the condition (c) above had been fulfilled?
b) The same goes with finding out for criteria (a) and (b), model all the
variables together or model them separately?
c) In PLS, how do we determine whether the mediation is full or partial?
Thanks.