fully mediation or not?

Question

I am using Process Marco Mediation Model

The results are as follows:

X-->Y total effect is not significant (c path= -.06, P=.109)

X--> M significant (a path= -.34, p=000)

M--> Y significant (b path= .13, p=.21)

X--> Y not significant (c' path= -.02, p=737)

Indirect effect is significant (=-.05, CI[-.088, -.006])

Is this still fully mediation? How do I interpret the result? Thanks

Answer 1

The author of the Process Macro, Hayes, believes that "full" and "partial" mediation are outdated terms, hence the Process macro offers no guidance on making this distinction.

In the Hayes Process framework, you can conclude that M mediates the effect of X on Y i the indirect effect of of X and M on Y is statistically significant.

However, If you were asked to go back a run a "Baron and Kenny" style analysis of mediation, it would fail at the first step, a bivariate regression predicting x from y, which would be not significant.

It appears to me that both practices are currently in use.

Good luck! I'm going to take a Tylenol and suggest you do the same.