Social Relations Model

David Kenny's Social Relations Model (SRM) allows researchers to study patterns in the thoughts, feelings, and behavior of individuals across relationships, as well as to examine the uniqueness of each relationship. Additionally, the model addresses the most important statistical challenge for relationships research that involves dyadic data—data collected from more than one person in a relationship. Specifically, the model accounts for nonindependence, the situation in which responses from persons in a relationship are related to each other. For example, a husband's marital satisfaction is likely related to or influenced by his wife's satisfaction. Nonindependence is more than a statistical nuisance because relationship research is often explicitly concerned with the degree of nonindependence between individuals and the processes that explain noninde-pendence, how one person's thoughts, feelings, and behavior influence another person. Designed explicitly for the analysis of dyadic data, the SRM offers researchers an important tool to understand relationships. This entry reviews the SRM and illustrates its application to different relationships, including friendships and families.

Although the SRM applies to dyadic data, the analysis requires information from groups of at least four individuals. Thus, the model is most easily applied to the study of work groups, friendship groups, and families. The most commonly used group design is a round robin. Consider a study of self-disclosure among a group of mutual friends. In a round-robin design, each person would rate how [Page 1558]much he or she discloses to each other friend. Similarly, if actual self-disclosure is measured, each person would interact one on one with each friend, and the disclosures of each person in every interaction would be recorded.

The SRM proposes that any dyadic measure (one person's report about or behavior toward another person) may reflect three important effects: actor, partner, and relationship. To illustrate, imagine that Mike reports trusting his friend Susan. Mike is the actor and Susan is the partner. The actor effect measures the degree to which Mike generally trusts or distrusts his friends. The partner effect assesses whether all of Susan's friends tend to trust or distrust her. Finally, the relationship effect indicates whether Mike especially trusts Susan beyond his general tendency to trust and the tendency for others to trust Susan. The presence of actor effects indicates individual differences in the tendency to be trusting, whereas partner effects point to something about the target person that elicits trust from others. The relationship effect reflects either the unique combination of two persons or a relationship property that emerges over time. These effects are interpreted in a similar manner for actual behavior. For example, consider the tendency of individuals to compliment friends while doing a cooperative decision-making task. An actor effect indicates that some individuals tend to compliment everyone, whereas others are less complimentary. A partner effect would demonstrate that some individuals are complimented more than others. The presence of a relationship effect would indicate that compliments are particularly prevalent in some friendships and less prevalent in others.

Recently, the SRM has been applied to families. In a family analysis, each family role (e.g., mother) has an actor and a partner effect. Relationship effects refer to the particular role combinations. For example, consider a study that examines criticism in families consisting of a mother, a father, and a daughter. The actor effect for mothers would indicate whether mothers tend to criticize all family members to a similar degree. The father partner effect would reveal whether fathers tend to be criticized to a similar degree by both mothers and daughters. The mother-father relationship effect would show whether criticism is especially pronounced in that particular relationship.

...