An Alternative Framework for Defining Mediation

Abstract

The present article provides an alternative framework for evaluating mediated relationships. From this perspective. a mediated process is a chain reaction, beginning with an independent variable that affects a mediator that in turn affects an outcome. The definition of mediation offered here, presented for stage sequences, states three conditions for establishing mediation: (a) the independent variable affects the probability of the sequence no mediator to mediator to outcome; (b) the independent variable affects the probability of a transition into the mediator stage; (c) the mediator affects the probability of a transition into the outcome stage at every level of the independent variable. This definition of mediation is compared and contrasted with the well-known definition of mediation for continuous variables discussed in Baron and Kenny (1986), Judd and Kenny (1981), and Kenny, Kashy, and Bolger (1997). The definition presented in this article emphasizes the intraindividual, time-ordered nature of mediation.

Mediation in experimental and nonexperimental studies: New procedures and recommendations.

“…The total effect of X on Y is the sum of the indirect effect a × b and the direct effect cЈ. In 14 Collins et al (1998) and MacKinnon and colleagues (MacKinnon, 2000;MacKinnon et al, 2000) recommended dropping the first step of Baron and Kenny (1986) for a different reason than we do. They considered inconsistent mediating variables that may have effects that go in opposite directions, so the total effect may seem to disappear.…”

“…In structural equation models, for example, the standard errors of parameter estimates are typically smaller when paths that are known to have zero weight are eliminated from the model. the first step of Baron and Kenny's (1986) classic approach (e.g., Collins et al, 1998;MacKinnon, 2000;MacKinnon et al, 2000). Especially for distal processes, for which the usual bivariate tests of association have limited power, we recommend that the mediation analysis proceed on the basis of the strength of the theoretical arguments rather than on the basis of the statistical test of X on Y.…”

“…We have assumed simple linear regression models in our discussion, but more complicated nonlinear alternatives should be considered for binary variables (Collins et al, 1998;Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001). Bootstrap methods can be used with nonlinear models as well as the linear models we have considered in our illustrations.…”

“…Statistical approaches to the analysis of mediation have been discussed extensively in the psychological literature (Baron & Kenny, 1986;Collins, Graham, & Flaherty, 1998;James & Brett, 1984;Judd & Kenny, 1981;Kenny, Kashy, & Bolger, 1998;Maccorquodale & Meehl, 1948;MacKinnon & Dwyer, 1993;MacKinnon, Krull, & Lockwood, 2000;MacKinnon, Warsi, & Dwyer, 1995;Rozeboom, 1956;Sobel, 1982). In this article, we focus on the approach to mediation analysis that was articulated by Kenny and his colleagues (Baron & Kenny, 1986;Judd & Kenny, 1981;Kenny et al, 1998).…”

Provoked Vestibulodynia: Mediators of the Associations Between Partner Responses, Pain, and Sexual Satisfaction

“…The overall relationship between the independent variable and the dependent variable does not have to be significant in order to test for mediation (Collins, Graham, & Flaherty, 1998;Kenny, Kashy, & Bolger, 1998;MacKinnon, Krull, & Lockwood, 2000;MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002;Shrout & Bolger, 2002). In these cases, mediation may be pursued because of prior theoretically or empirically-driven hypotheses which suggest Mediation in Couples with PVD 15 that the relationship between the independent and dependent variable only exists through the mediator (Collins et al, 1998). It is also possible that the direct and mediated effects of an independent variable on a dependent variable have opposite signs, in which case their effects may cancel each other out, resulting in a non-significant overall relationship (MacKinnon et al, 2000).…”