How Couples’ Social Networks Shape Relationship Satisfaction and Communication

Every romantic couple lives inside a web of other relationships. Friends, family members, and acquaintances watch, advise, interfere with, and celebrate the partnerships around them, and psychologists have long suspected that this surrounding network matters for how couples feel about each other. A new study published in Trends in Psychology by James Stein and Lucy Rios of Utah Tech University provides some of the clearest evidence yet that what people believe about their social networks is woven directly into the quality of their romantic relationships, shaping not only their own satisfaction and intimacy but also their partner’s.

The research team recruited 80 couples, yielding 160 individual participants, and collected dyadic data between April 2022 and March 2024. Dyadic data of this kind is prized in relationship science because it allows researchers to model two people whose outcomes are statistically intertwined. Rather than treating each partner as an independent observation, the study used the actor-partner interdependence model, a statistical framework developed by Cook and Kenny that separates two distinct pathways of influence. Actor effects capture how a person’s own perceptions predict their own relational outcomes, while partner effects capture how one person’s perceptions spill over to predict their partner’s outcomes. This distinction matters because relationships are genuinely interdependent systems, and ignoring that interdependence can distort or obscure the very effects researchers hope to detect.

To preserve the statistical assumption of independence, couples completed all substantive measures in separate rooms, and each individual provided informed consent twice, first jointly as a couple and again individually before answering questions alone. The project was approved twice by Utah Tech University’s institutional review board, first in January 2022 and again in April 2023, and was conducted in accordance with the Declaration of Helsinki and the guidelines of the Belmont Report. All survey items used seven-point Likert scales ranging from strong disagreement to strong agreement, a format that allowed participants to express graded levels of certainty and endorsement rather than simple binary judgments.

The study focused on two families of network-based perceptions. The first is network uncertainty, the sense of ambiguity people experience about how their social network members view, influence, or evaluate their romantic relationship. Uncertainty has been a central construct in relational turbulence theory and in the theory of motivated information management, both of which argue that ambiguity about a relationship, whether it originates inside or outside the dyad, fuels emotional and communicative upheaval. Stein’s earlier work had developed and validated a dedicated network uncertainty measure, and this study extended that line of inquiry by embedding the construct within a full dyadic analysis of couples.

The second family of perceptions concerns network interdependence and facilitation, the degree to which couples perceive their social networks as intertwined with, and supportive of, their partnership. This builds on a long tradition in social psychology. Classic work by Milardo, Johnson, and Huston in the 1980s documented how patterns of interaction with network members shift as close relationships develop, while Parks, Stan, and Eggert showed that romantic involvement and social network involvement are deeply entangled. Felmlee’s influential social network perspective on dyadic stability captured the idea memorably: no couple is an island. More recently, Sprecher and Felmlee have examined how network pressure shapes decisions to enter relationships, and Stein’s own prior studies established measurement models for network and partner interdependence, network interference, and network facilitation.

A methodological strength of the new study lies in how it measured the couples’ shared social world. Rather than asking each partner separately about their own friends and family, the researchers had couples assess network overlap together using a duocentric model of assessment, following the primer on duocentric social network analysis published by Kennedy and colleagues in the Journal of Marriage and Family. A duocentric approach places the couple at the center of a single combined network, capturing the overlapping constellation of people connected to both partners. This is a meaningful departure from ego-centric designs that treat each partner’s network in isolation, and it aligns with the reality that long-term couples typically share friends, in-laws, and communal obligations.

The results revealed a dense pattern of significant actor and partner effects linking network-based perceptions to relational outcomes. Measures of network uncertainty and network interdependence and facilitation shared a variety of significant associations with measures of relationship satisfaction, intimacy, relational talk, and network-based relational talk. In practical terms, when one partner perceived greater uncertainty about the network, that perception was associated not only with that person’s own relational experiences but, in several cases, with the partner’s as well. The same held true for perceptions of network support and interdependence: how one member of a couple viewed the couple’s embeddedness in their social world carried consequences for both members of the dyad. These partner effects are the study’s most striking contribution, because they demonstrate that network perceptions are not merely private psychological states but shared relational resources, or liabilities, that cross the boundary between partners.

The findings also connect to a growing literature on how couples communicate with and about their networks. Prior research has explored relational talk, the conversations partners have about the state and future of their relationship, and how such talk relates to uncertainty within courtship. Knobloch and Theiss, for example, used a longitudinal actor-partner interdependence model to show that relational uncertainty shapes relationship talk during courtship, and Knobloch and Knobloch-Fedders linked relational uncertainty to depressive symptoms and relationship quality. The new study extends this framework outward, showing that uncertainty and talk about the network itself, what the authors call network-based relational communication, form a distinct layer of couple communication worthy of measurement in its own right.

The authors frame their contributions as theoretical, methodological, and applied. Theoretically, the study bolsters the case that relationship science should treat the social network as an active ingredient in romantic outcomes rather than background context, consistent with relational turbulence theory’s emphasis on how external involvement generates uncertainty and reactivity. Methodologically, the combination of duocentric network assessment with dyadic modeling offers a template for future studies that want to capture both the structure of couples’ shared networks and the interdependent perceptions partners hold about them. In applied terms, the findings suggest that practitioners working with couples may benefit from attending to how each partner perceives the couple’s network, since uncertainty or perceived interference emanating from friends and family may be felt by both partners, even when only one person voices it.

The study is not without boundaries. The data were collected from couples volunteering for a university-based project, with participant rewards paid out of pocket and no external funding, and the data are not publicly available, though they can be shared with reviewers or editors after anonymization. As with any correlational design, the significant associations between network perceptions and relational outcomes do not by themselves establish causal direction; it remains possible that dissatisfied partners perceive their networks more negatively, rather than the reverse. Still, by demonstrating that network uncertainty and network facilitation travel together with satisfaction, intimacy, and relational talk across both members of a couple, the study makes a compelling case that the dyadic bubble is porous. The people surrounding a couple, and the beliefs partners hold about those people, appear to be part of the relationship itself, measurable, consequential, and shared.

Subject of Research: Dyadic study of how romantic partners' perceptions of their social networks relate to relationship satisfaction, intimacy, and communication using actor-partner interdependence modeling.

Article Title: Network-Based Perceptions and Relational Outcomes: An Actor-partner Interdependence Modeled Approach

Article References: Stein, J., & Rios, L. (2026). Network-Based Perceptions and Relational Outcomes: An Actor-partner Interdependence Modeled Approach. Trends in Psychology. https://doi.org/10.1007/s43076-026-00530-x

Image Credits: AI Generated

DOI: 10.1007/s43076-026-00530-x

Keywords: social networks, romantic relationships, actor-partner interdependence model, relationship satisfaction, network uncertainty, intimacy, relational communication, dyadic data, duocentric networks, relational turbulence theory, network facilitation, couples research