In our previous lesson, we established a rigorous process for formulating a precise causal question, identifying the intervention, outcome, and target population. This discipline is the essential first step. Now, we shift our focus from a single causal arrow to the intricate system in which it operates. Your research on the husband-wife-bull triad is not about isolated individuals, but about a connected, interdependent emotional system.
Today's lesson addresses the core complexity you identified in your research: the triadic data structure. Our objective is to formally define the six relational perspectives—three individual and three dyadic—that comprise this system. By treating each of these as a distinct unit of analysis, we can begin to untangle the web of influences on emotions like jealousy and guilt. This framework is the foundation for all the advanced modeling we will explore later.
From Individuals to Interdependent Systems
In any multi-person dynamic, especially one as intimate and complex as a triad, analyzing individuals in isolation is insufficient. The participants' emotions and behaviors are not independent. The husband’s emotional state influences the wife, which in turn can affect her interactions with the bull, and so on. Family systems theory provides a powerful conceptual lens for this. It posits that a group is more than the sum of its parts, characterized by interdependence and complex, layered relationships.
A paper by Schrodt, Soliz, and Braithwaite on stepfamily dynamics offers a clear introduction to this perspective. While the context is different, the underlying structure of a three-person system with distinct roles is directly analogous.
A Social Relations Model of Everyday Talk and ...
This paper introduces key principles from family systems theory that are highly relevant to your work. It frames the "triangle" as the fundamental unit of a relational system.
Please read the first part of the "Theoretical Perspective" section. Focus on the three key principles mentioned: interdependence and wholeness, and the idea of complex relationships and subsystems. Pay particular attention to the mention of Bowen's theory, which identifies the triangle as the fundamental building block of any system.
This systems perspective requires a statistical approach that can honor this complexity rather than assuming it away. The Social Relations Model (SRM) is purpose-built for this task. It allows us to decompose the variance in an outcome (like an emotional rating) into components attributable to the different perspectives within the system.
Defining Perspectives with the Social Relations Model (SRM)
The SRM dissects interpersonal data to answer questions like: "Is the husband's jealousy a product of his personality, something about his wife's behavior, or the unique chemistry between them?" It achieves this by partitioning the variance of a measured outcome into three primary components:
- Actor Effect: The consistent tendency of an individual to feel or behave in a certain way across all their relationships. This is their personal baseline or disposition.
- Partner Effect: The consistent tendency of an individual to elicit a certain feeling or behavior from others across all their relationships. This is the common effect they have on their partners.
- Relationship Effect: The unique feeling or behavior that emerges in a specific dyad, which cannot be explained by the actor's and partner's general tendencies. This captures the unique "chemistry" or dynamic of a pair.
A paper by Reifman, Eastwick, and Reis provides a wonderfully clear conceptual introduction to these components.
The Social Relations Model for Asymmetric-Block-Design ...
This tutorial lucidly explains the core logic of the SRM by posing a series of hypothetical questions about a simple behavior: smiling. This will help you intuitively grasp the different effects.
Please read the section titled The Social Relations Model (SRM). Focus on the five hypothetical questions about smiling. Notice how they cleanly separate the concepts of a personal trait (actor effect), the ability to elicit smiles (partner effect), and the unique dynamic between two people (relationship effect).
The Six Perspectives of the Triad
Now, let's map these SRM components onto your research context to formally define the six units of analysis. We have three individuals (Husband, Wife, Bull) and three unique dyads (H-W, W-B, H-B).
The 3 Individual Perspectives
Each person in the triad brings their own disposition to the system. The SRM captures this with the actor and partner effects. For any given individual (e.g., the husband), we can analyze:
- His Actor Effect: His general tendency to experience jealousy. Is he, by nature, a more jealous person compared to other husbands in similar situations? This is a property of the husband as an individual.
- His Partner Effect: His general tendency to elicit certain responses from others. For instance, does he consistently elicit feelings of guilt in his wife and a sense of caution in the bull, regardless of the specific situation? This is also a property of the husband as an individual.
Together, the actor and partner effects for the husband, wife, and bull constitute the three individual-level perspectives. They represent stable, person-specific contributions to the system's dynamics.
The 3 Dyadic Perspectives
The most interesting dynamics often lie in the interactions between pairs. The SRM captures this with the relationship effect. This effect is, by definition, dyadic. It represents the unique variance that emerges from the pairing of two specific individuals, after accounting for their individual actor and partner effects.
-
The Husband-Wife Dyad (H-W): The relationship effect here quantifies the unique emotional dynamic between them. For example, the husband might experience a level of jealousy with his wife that is higher (or lower) than what would be predicted based on his general jealous disposition (actor effect) and her general tendency to elicit jealousy (partner effect). This "extra" jealousy is a property of their specific dyadic bond.
-
The Wife-Bull Dyad (W-B): Similarly, the relationship effect captures the unique dynamic here. The wife might experience a level of guilt with a particular bull that is different from her interactions with other potential partners, even after controlling for her own disposition and the bull's general tendency to elicit guilt. This unique dynamic is the unit of analysis for this dyad.
-
The Husband-Bull Dyad (H-B): This often-overlooked dyad is also a critical unit of analysis. The relationship effect here could capture unique feelings of rivalry, camaraderie, or respect that are specific to the interaction between this particular husband and this particular bull.
Acknowledging Distinguishable Roles: The Asymmetric Block Design
In your triad, the members are not interchangeable. The husband, wife, and bull have fundamentally different roles, rights, and expectations. A simple SRM assumes "round-robin" data where everyone rates everyone on the same footing. This doesn't fit your context.
A more appropriate framework is the asymmetric block design. This SRM variant is designed for situations where individuals from different social categories interact. The speed-dating example in the Reifman et al. paper, where men and women are distinct categories, is a powerful analogy. We can think of your triad as being composed of several asymmetric interactions (e.g., Wife-Bull, Husband-Bull).
The Social Relations Model for Asymmetric-Block-Design ...
This section describes the asymmetric block design and illustrates its parameters using a clear, real-world example from speed-dating. This will help you see how the SRM handles interactions between people in distinct roles.
First, review the different SRM designs in Table 1 and the surrounding text, paying special attention to the definition of the asymmetric block design. Then, read the section "A pictorial and numerical example," focusing on the verbal formulas in Figure 2. See how the likelihood of a woman offering a second date is broken down into her general eagerness (actor), the man's general attractiveness (partner), and her unique reaction to him (relationship). This is precisely how we can decompose emotional responses in your triad.
The Formal Statistical Definition
Given your background in statistics, it is valuable to see the formal model that underpins these concepts. The decomposition of an observation into its constituent parts is not just a conceptual exercise; it is a mathematical one. The appendix of the Reifman et al. paper provides the formal equations.
The Social Relations Model for Asymmetric-Block-Design ...
This appendix provides the formal statistical model for the asymmetric block design. You don't need to memorize the equations, but seeing them will solidify your understanding of how an outcome is decomposed.
Please review the start of the Appendix. Focus on Equations 1 and 2. Notice how a given measure, Y_{ijk}, is expressed as a sum of several components: \mu: A grand mean. \alpha: The actor effect (the individual's deviation). \beta: The partner effect (the partner's elicitation deviation). \gamma: The relationship effect (the unique dyadic deviation). These equations are the formal definition of the units of analysis we have been discussing.
Conclusion
In this lesson, we have moved from viewing your research subjects as three individuals to seeing them as a six-faceted relational system. We have formally defined these six perspectives—the actor/partner effects for each of the three individuals and the relationship effects for each of the three dyads—using the robust framework of the Social Relations Model. This act of definition is a critical step in managing the analytical complexity of your data.
Key Takeaways:
- A triad is an interdependent system, not just a collection of three people.
- The Social Relations Model (SRM) provides a statistical method to formally define and analyze the different perspectives within this system.
- The six units of analysis are the three individual perspectives (captured by actor and partner effects for the husband, wife, and bull) and the three dyadic perspectives (captured by the relationship effects for the H-W, W-B, and H-B pairs).
- The asymmetric block design is an appropriate SRM variant for handling the distinct roles within the triad.
In our next lesson, we will build directly upon this foundation. We will explore the crucial concept of non-independence in your data, examining exactly how these individual (compositional) and dyadic (relational) effects create statistical dependencies that can bias your causal estimates if ignored.