Dyadic Data Analysis Service (APIM)

Dyadic data analysis with APIM — Dyadic Data Analysis (APIM)
Actor and partner effects in paired data

For couple, parent–child and other paired data we fit the Actor–Partner Interdependence Model, which is the correct analytical design for relational research.

Paired data are not independent observations, and treating them as such inflates significance. The APIM handles that interdependence explicitly and, more usefully, separates the effect of a person’s own characteristic on their own outcome (the actor effect) from its effect on their partner’s outcome (the partner effect).

Before estimation we establish whether the members of the dyad are distinguishable (as in heterosexual couples or parent and child) or indistinguishable (as in same-sex friends), since this determines how the model is specified and how the parameters are constrained.

Mediation and moderation extensions of APIM are available where the theory calls for them.

What We Establish

Distinguishability

Whether dyad members can be told apart on a meaningful variable, tested rather than assumed.

Actor and partner effects

Both estimated and reported separately, with the interpretation stated plainly.

Extensions

Mediation, moderation and over-time APIM where the design supports them.

How the Process Works

Initial assessmentYou share your research question, your dataset and your target journal. Together we decide which analysis is appropriate.
Proposal and planScope, timeline and fee are agreed, and what you will receive is set out in writing.
AnalysisAssumption checks, model specification and estimation, with interim findings shared with you.
Delivery and handoverAPA-format tables, a written interpretation and a handover conversation. We do not request authorship for standard analysis work.