Actor-Partner Interdependence Model
Also known as: APIM, Dyadic Actor-Partner Model, Interdependence Regression Model
The Actor-Partner Interdependence Model (APIM), formalized by Kenny, Kashy, and Cook, is the standard framework for analyzing dyadic data in which two people's outcomes depend on both their own and their partner's characteristics. For each member of a dyad, the model estimates an actor effect -- the influence of a person's own predictor on their own outcome -- and a partner effect -- the influence of the partner's predictor on the person's outcome -- while explicitly modeling the statistical non-independence of the two members' scores. For example, a person's relationship satisfaction may depend on their own attachment anxiety (actor effect) and on their partner's attachment anxiety (partner effect). By simultaneously estimating these effects and accounting for the correlation between partners, the APIM avoids the bias of treating dyad members as independent and reveals how individuals in relationships shape each other, making it indispensable for research on couples, families, and other interacting pairs.
Key highlights
- Estimates both self (actor) and partner influences simultaneously.
- Properly models the non-independence inherent in dyadic data.
- Distinguishes interpretable interdependence patterns.
- Flexible via multilevel or SEM estimation for distinguishable and indistinguishable dyads.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use the APIM whenever you have dyadic data -- couples, parent-child pairs, friends, coworkers, patient-caregiver dyads -- and want to estimate how each person's outcome depends on their own and their partner's characteristics while properly handling non-independence. It is the default model for couples and family research. It is less appropriate when members rate many others (where the Social Relations Model fits), when the dyad members are indistinguishable in ways that complicate role assignment without proper handling, or when only one member is measured. Researchers must decide whether dyads are distinguishable (for example, by role or gender) or indistinguishable, since this affects model specification and estimation.
Strengths & limitations
- Estimates both self (actor) and partner influences simultaneously.
- Properly models the non-independence inherent in dyadic data.
- Distinguishes interpretable interdependence patterns.
- Flexible via multilevel or SEM estimation for distinguishable and indistinguishable dyads.
- Requires data from both dyad members on predictors and outcomes.
- Cross-sectional APIMs do not establish causal direction of partner effects.
- Indistinguishable-dyad estimation imposes constraints that must be handled correctly.
- Mediation and longitudinal extensions add substantial complexity.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
What are actor and partner effects?
The actor effect is the influence of a person's own predictor on their own outcome, such as how one's own attachment anxiety affects one's own satisfaction. The partner effect is the influence of the partner's predictor on the person's outcome, such as how a partner's anxiety affects one's own satisfaction. The APIM estimates both simultaneously while controlling for each other.
Why must non-independence be modeled?
The two members of a dyad share a relationship, so their scores are correlated; treating them as independent observations biases standard errors and can lead to wrong conclusions. The APIM explicitly models the correlation between partners' residuals, which both corrects the statistics and captures shared influences beyond the measured predictors.
What is the distinction between distinguishable and indistinguishable dyads?
Distinguishable dyads have members who differ on a meaningful variable (such as role or gender) that can order them, allowing separate effects per role. Indistinguishable dyads (such as same-sex friends) lack such an ordering, so the model must constrain effects to be equal across members and use specialized estimation. Choosing the right specification is essential for valid APIM results.
Sources
- 1.Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic Data Analysis. Guilford Press.ISBN 9781572309869
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Actor-Partner Interdependence Model. ScholarGate. https://scholargate.app/social-psychology/actor-partner-interdependence-model