MEDIATED ASSOCIATIONS: PARADIGMS AND SITUATIONS
James J. Jenkins UNIVERSITY OF MINNESOTA
The concern with mediate associations and mediation processes generally is ancient (as these things go in psychology), dating back at least to the British associationist philosophers. Research in the area has lagged far behind, of course, and is largely concentrated in the last twenty-five years. Even within this period it is easy to plot a rapidly accelerating curve of research endeavors and theoretical papers devoted to mediation, illustrating a growing concern with mediation phenomena and the adaptation of experimental tools, some new and some old, to the task of understanding the processes. This literature has been reviewed elsewhere (e.g., Jenkins, 1959; Kjeldergaard and Horton, 1960; Goss, 1961a, 1961b), and it is not my purpose to recapitulate it here.
The endeavor of this paper is to review some experimentation with which I have been closely involved, to illustrate ways in which the emphasis of our attack on mediation problems has shifted, and to propose a view of mediation which, I believe, suggests important directions for future research.
In our concern with research in mediation, I believe all of us have been attempting to demonstrate two major propositions: First (aimed at our doubting colleagues), mediation effects really can be found, and second (aimed at each other), mediation is "properly" inferred by X____________________