MEDYAD is a regression-based computational macro for SPSS, SAS, and R that conducts mediation analysis with distinguishable dyadic data. It is designed to handle models such as the Actor-Partner Interdependence Model extended to Mediation (APIMeM) and special cases and extensions of this model. MEDYAD estimates the direct and indirect effects of each actor’s (X) on their own and their partner’s outcome (Y ) through their own and their partner’s mediator(s) (M).
If you wish to use MEDYAD in your research, please use the following citation: Coutts, J.J., Hayes, A.F., & Jiang, T. (2019). Easy statistical mediation analysis with distinguishable dyadic data. Journal of Communication, 69(6), 612-649.
An advantage of MEDYAD is its flexibility with model estimation. It allows up to 12 mediators (6 if the mediator is measured on both members of the dyad or 12 if the measure is a feature of the dyad or only measured on one of its members). MEDYAD also allows for specification of 1 or 2 Xs and 1 or 2 Ys (depending on whether the variable is between-dyad or mixed). Bootstrapping methods are available for inference about specific indirect effects and contrasts between them.
In addition to the syntax version, a dialog box exists based on SPSS’ graphical user interface (GUI). Not all options are available in the dialog box, but it can be permanently installed and used for simple model estimation.
There are numerous other features available to researchers including (but certainly not limited to) options for controlling for covariates, printing descriptive statistics, and obtaining pairwise contrasts between specific indirect effects. MEDYAD, and other macros produced by the MAC lab, can also be found at afhayes.com.