4.1 What about Barron and Kenny?

On page 121 of the text, Hayes opined:

Complete and partial mediation are concepts that are deeply ingrained in the thinking of social and behavioral scientists. But I just don’t see what they offer our understanding of a phenomenon. They are too sample-size-dependent and the distinction between them has no substantive or theoretical meaning or value of any consequence. I recommend avoiding expressing hypotheses about mediation or results of a mediation analysis using these terms.