Dear friends,
i've a little problem with a study about mediation. I will call my dependent
variable "Y", the independent "X" and "Z" will be
my mediator variable...
Using a linear regression, i can say that there isn't a direct effect of X
on Y because the coefficient of X isn't significative, but i've tried to
use anyway the Sobel's test to find a mediation relation.
The result was positive, and with a p-value of 0.03 i can say that Z have a
mediator effect on the relation between X and Y.
Now my question: is right think about a mediator effect of Z on the relation
between X and Y, also if there isn't a direct effect????
Waiting for your opinion, thank you so much.
Costantino Milanese
[[alternative HTML version deleted]]
