Hi,
I wonder if it's appropriate to ask this question here but don't know where to ask this question.
I want to know if the relationship between age (X) and a behaviour score (Y) is mediated by BOLD signals in a brain region (M) and thus did a mediation analysis (using TD Wager's tool). Results showed that path a (X to M), path b (M to Y), and the mediation effect (ab) were all significant. Original path c (total effect of X on Y) was significant but, after controlling for the mediation effect, the unmediated direct path c' became insignificant.
I think this means that the relationship between X and Y was COMPLETELY mediated by the mediator M. However, studies have proved that other factors, such as a physiological parameter of the spinal cord (M'), is correlated with X and Y (and maybe also correlated with M). However, we did not collect data of M'.
My concern is:
If the mediation effect is partial, that is, c' was significant, people can argue the possibility that M' might have a role in X-Y relationship. Since the effect here was COMPLETE, can I say that M is the most important factor to mediate X-Y relationship, and argue that it is not necessary to have M' data?
Thanks for any suggestions.
Mark
|