Moderation in the actor–partner interdependence model

Abstract

Potential moderators of effects in the actor–partner interdependence model (APIM) include variables that vary within dyads, between dyads, or both between and within dyads (i.e., mixed moderators). Another factor in the moderation of the APIM is whether dyads are indistinguishable (e.g., same‐sex friendship pairs) or distinguishable (e.g., heterosexual couples). For each possibility, what are the potential moderator effects (up to 8), how they might be estimated and tested, and how they can be interpreted are discussed. Submodels are also presented, based on patterns of moderation of the actor and partner effects, which are statistically simpler, more conceptually meaningful, and more powerful in testing moderator effects. Example analyses illustrate the recommended steps involved in an APIM moderation analysis.

Concurrent and longitudinal dyadic polynomial regression analyses of Big Five traits and relationship satisfaction: Does similarity matter?

“…On the grounds that no interaction effect or higher order coefficient term reached significance at the .01 significance level, we followed Garcia, Kenny, and Ledermann (2015) and tested whether simple APIM -by constraining all higher terms to zero -was not significantly worse than the polynomial regression models. The test of small difference in fit between the models did not yield any significant results suggesting that the simpler APIM displays the data equally well compared to the dyadic polynomial regressions (see supplemental online material for model fits of these and subsequent models).…”

Life Problems and Perceptions of Giving Support: Implications for Aging Mothers and Middle-Aged Children

“…To consider whether the frequency of one’s own support given in the mother-child relationship moderates the dyadic associations between life problems and support perceptions, two interaction terms were added in the second step of the model (i.e., Own problems X Own support, Partner problems X Own support; Garcia, Kenny, & Ledermann, 2015). In the third step, we added two three-way interaction terms to determine whether these moderation effects differed on the basis of child gender (i.e., Own problems X Own support X Child gender, Partner problems X Own support X Child gender).…”

Baby with the Bathwater? Examining the Relationship between Video Game Use and Relationship Outcomes and the Moderating Effects of Attachment Behaviors among Married Casual Gamers

“…Finally, the moderating effects of own and partner's attachment behaviors was tested. As described in Garcia, Kenny, and Ledermann (), attachment behaviors is a mixed‐dyad variable. Thus, there were four potential moderating effects per partner (a: own BARE x own time → own outcomes; b: own BARE x own time → partner outcomes; c: own BARE x partner time → own outcomes; d: own BARE x partner time → partner outcomes).…”