Social Networks

29  Theories of Ego Network Homogeneity

  • Welcome
  • Introduction to Networks
    • 1  What Are Networks?
    • 2  What is A Social Network?
  • Graph Theory: The Basics
    • 3  Introduction to Graphs
    • 4  Graphs and their Subgraphs
    • 5  Types of Ties in Social Networks
    • 6  Types of Ties and Their Graphs
    • 7  Basic Graph Metrics
    • 8  Nodes and their Neighborhoods
    • 9  Nodes and their Degrees
    • 10  Degree-Based Graph Metrics
    • 11  Indirect Connections
    • 12  Directed Indirect Connections
    • 13  Graph Connectivity
    • 14  Tree Graphs
  • Matrices: The Basics
    • 15  Introduction to Matrices
    • 16  The Adjacency Matrix
    • 17  Matrix Operations: Row and Column Sums
    • 18  Basic Matrix Operations
    • 19  Matrix Multiplication
  • Motifs
    • 20  Triads
  • Centrality
    • 21  Centralities based on Degree
    • 22  Centralities based on the Geodesic Distance
    • 23  Centralities based on Shortest Paths
    • 24  The “Big Three” Centrality Metrics
    • 25  Getting Centrality from Others
  • Two-Mode Networks
    • 26  Affiliation Networks
  • Ego Networks
    • 27  Ego Network Metrics
    • 28  Collecting Ego-Network Data
    • 29  Theories of Ego Network Homogeneity
  • Subgroups and Blocks
    • 30  Clique Analysis
    • 31  Cohesive Subsets
    • 32  Equivalence and Similarity
    • 33  Local Node Similarities
  • Network Theory
    • 34  Dunbar’s Theory of Social Circles
    • 35  The Strength of Weak Ties
    • 36  Structural Holes and Brokerage
    • 37  Simmelian Tie Theory
    • 38  Dyadic Balance
    • 39  Triadic Balance
    • 40  Structural Balance
    • 41  Theories of Valenced Interactions
    • 42  Dominance Hierarchies
    • 43  The Diffusion of Innovations
    • 44  The Small World

Table of contents

  • 29.1 Blau’s Macrostructural Theory
  • 29.2 Core Tenets of Blau’s Macrostructural Theory
    • 29.2.1 Group Size Effects
    • 29.2.2 Multiform Heterogeneity Effects
    • 29.2.3 Types of Correlations across parameters of social structure
  • 29.3 Feld’s Theory of Social Foci
  • 29.4 Marsden’s Theory of Ego Network Diversity
  • References

View source

29  Theories of Ego Network Homogeneity

29.1 Blau’s Macrostructural Theory

Blau’s macrostructural theory provides a macro-level explanation for the composition of ego networks, specifically examining how sociodemographic characteristics influence the formation of ties and the overall homogeneity or heterogeneity of an individual’s connections (Blau 1977). It posits that the larger-scale structure of social distinctions significantly impacts the smaller-scale composition of personal networks.

Blau (1977) defines social structure as the distribution of people across various social positions, and these distributions directly influence the opportunities people have to associate with one another. These social positions are based on distinctions people use to differentiate themselves from one another, such as occupation, religion, educational rank, gender, age, and generation (e.g., “boomers” versus “Gen Z”), as well as race, ethnicity, and nationality (among many others).

The dimensions along which these social distinctions are made are referred to as parameters in Blau’s framework (Blau 1974). These can be categorical nominal parameters, like gender, race, or religion, which have no inherent ordering, or rank-ordered graduated parameters, such as education, age, or income, which do have inherent “higher” and “lower” values (e.g., some people are necessarily older or make more money than others). The degree of heterogeneity in a society is reflected by how equally or unequally people are distributed across these dimensions.

Blau’s theory includes several key principles, here referred to as Blau’s rules, which explain these macrostructural effects.

29.2 Core Tenets of Blau’s Macrostructural Theory

29.2.1 Group Size Effects

Blau’s theory highlights that the numerical size of social groups plays a crucial role in shaping who interacts with whom, often leading to observed patterns of homogeneity or heterogeneity. These effects operate independently of individual preferences for similarity, meaning that people may form certain types of ties simply because of the demographic landscape rather than through explicit choice.

Rule 1: Smaller Groups and Outgroup Ties.- Members of numerically smaller groups are proportionally more likely to form ties with individuals outside their own group (outgroup members). This means that individuals from minority groups might experience greater social diversity in their networks, not necessarily because they prefer it, but because the number of available ingroup members is limited, increasing the statistical probability of encountering and forming ties with outgroup individuals. Minorities, therefore, can benefit from outgroup ties by gaining social diversity and a diversity of intellectual or cultural outlook.

Rule 2: Larger Groups and Ingroup Ties.- Conversely, members of numerically larger groups are proportionally more likely to form ties with individuals within their own group (ingroup members), regardless of ingroup preference. As the inequality in group sizes increases such that one group comes to numerically dominate the others, members of the majority group tend to have almost exclusively ingroup relations with one another.

Some group size effects examples include:

  • Interracial Marriages: Data on interracial marriages in the U.S. illustrate these rules clearly. For instance, in 2013, a significantly higher percentage of individuals from numerically smaller racial or ethnic groups, such as American Indians (58%), were married to someone of a different race compared to White individuals (7%), who belong to a much larger group. This demonstrates how members of smaller groups are more likely to have outgroup relations, while members of larger groups are more likely to have ingroup relations.
  • Interracial Friendships in Schools: Similarly, in schools, a larger proportion of students from smaller “Other-Race” categories engage in friendships with individuals from other racial groups. In contrast, white students, typically a larger group in many schools, show a much higher percentage of ingroup friendships.

29.2.2 Multiform Heterogeneity Effects

This aspect of Blau’s theory addresses how multiple correlated dimensions of social differentiation interact to influence ego-network composition, leading to the observed patterns of bias in one direction or another.

Rule 3: Correlated Dimensions of Association.- Blau’s Rule 3 states that systematic correlations exist between different dimensions of association within a society, such as religion and race. For example, in the U.S., there are systematic correlations between religious preference and race, where a higher percentage of non-Hispanic whites are Protestant compared to Black or Hispanic individuals.

Blau’s Rule 3 implies that if an individual’s network exhibits bias based on one characteristic (e.g., race), it is highly likely to be biased by another correlated characteristic (e.g., religion). This can occur even if there is no explicit preference for homophily (the tendency to associate with similar others) along the second dimension (e.g., individuals are neutral about selecting contacts based on religious affiliation). Thus, if individuals have preferences for selecting contacts based on race, they will automatically show a bias based on religion (e.g., if your ego network is composed of mostly Black Americans, it will also be composed of majority Protestant individuals). The observed homogeneity in one dimension might simply be an outcome of the underlying correlations between various social attributes at the societal level, rather than a direct individual preference.

29.2.3 Types of Correlations across parameters of social structure

Three types of correlations can exist across various types of social distinctions:

  • Categorical-Categorical: For instance, between race and religion. If networks are biased by religion, they will also be biased by race, and vice versa, because there are systematic differences in the racial composition of religious groups at the macro level.
  • Categorical-Continuous: Such as between race and income. Networks biased by race might also be biased by income because there is a systematic correlation at the macro level between race and income in the U.S.
  • Continuous-Continuous: For example, between age and wealth. Networks biased by age might also be biased by income, since wealth increases with age (a trend that has increased over recent decades in the U.S.), creating systematic correlations between these two dimensions of association at the macro level.

Example: If individuals primarily choose contacts based on race, this can inadvertently lead to a biased representation of people from certain religious groups within their personal networks, even if they have no direct preference for or against specific religious affiliations. This is because race and religion are systematically correlated at the macro-structural level in society. Similarly, a preference for associating with people of a certain race might lead to a higher or lower representation of people from different social classes at the micro-structural level due to correlations between race and income.

In essence, Blau’s theory underscores that the composition of an individual’s social network is not solely a product of personal choices or preferences but is heavily shaped by the broader demographic structure of society and the complex interplay among various social distinctions within that structure.

29.3 Feld’s Theory of Social Foci

Feld’s Theory of Social Foci posits that social ties are organized and emerge from specific social contexts. These contexts, termed “foci” (singular: “focus”), are defined as any social, psychological, legal, or physical entity around which joint activities are organized (Feld 1981). Examples of such foci include workplaces, voluntary organizations, hangouts, or families.

Key Concepts:

  • Organization of Social Relations: The theory observes that social relations often form around “circles” or “common activities”. These foci act as a primary mechanism driving the formation of relationships within an individual’s network.
  • Mechanisms of Tie Formation: Feld’s theory suggests two primary mechanisms through which ties emerge from foci:
    • From circle to relation: Individuals who are part of the same “circle” or group are likely to form ties with each other.
    • From common activity to relation: Engaging in shared activities within a focus increases the probability of forming social connections. This highlights a “context-driven link formation” where the environment facilitates connections.
  • Dimensions of Foci: Foci can vary in several ways, including:
    • Constraint: The amount of time and energy they require from individuals, constraining foci require lots of your time and energy, while less constraining foci do not.
    • Size: The number of people involved in the focus. Some foci (e.g., UCLA) are very big, while others (your family) are small.

Foci themselves can also emerge from existing social relationships. For instance, friends might decide to form a club or participate in shared activities. Individuals typically belong to multiple foci that intersect. This phenomenon, referred to as “cross-cutting circles,” can lead to clustering within an ego-network due to overlapping memberships and activities.

In the context of segregation, Feld’s rule posits that segregated (or non-segregated) networks can arise from social foci, independent of individual preferences and group size differences. Social foci also play a role in maintaining ties. Ties embedded within constraining foci are more likely to be maintained, regardless of individual sentiments. Conversely, shared foci can contribute to the decay of ties by fostering reciprocity, which can stabilize ties or lead to their dissolution if reciprocity is not met.

Example: A clear example of Feld’s theory in action is seen with student clubs. These clubs function as “classic” foci, directly influencing the formation of ties among their members. Students who join a club are more likely to form friendships with other club members due to shared activities and regular interactions within that context.

29.4 Marsden’s Theory of Ego Network Diversity

Marsden’s (1987) work, particularly the one based on his analysis of the 1985 General Social Survey (GSS) data, contributes significantly to understanding the structural characteristics and diversity of ego networks. The 1985 GSS used a name generator approach to measure the ego networks of a representative sample of the American population (see Chapter 28), asking individuals to “Name all persons with whom you have discussed matters important to you,” thereby eliciting what are known as core discussion networks.

Key Propositions and Findings: Marsden’s propositions about ego networks provide insights into how their structural properties relate to diversity and homophily:

Network Size and Diversity:

  • Larger ego networks tend to be more diverse, less homophilous, and less constrained (contacts are less likely to know one another). This suggests that as individuals expand their network size, they are more likely to encounter and form ties with people who differ from them.
  • Smaller ego networks, conversely, tend to be more clustered and less diverse. This implies that smaller networks are often characterized by a higher concentration of similar individuals.

A general principle developed by Marsden is that greater network diversity is associated with less homophily. Homophily refers to the tendency for individuals to associate with others who are similar to themselves based on sociodemographic characteristics like age, race, gender, education, or nationality. Diversity, in this context, measures the variation in alter attributes within the ego network.

Additionally, Networks with a higher average tie strength tend to exhibit more clustering but less diversity. Strong ties often imply greater interdependence and shared contexts, leading to more interconnectedness among alters, but potentially limiting exposure to diverse perspectives. Similarly, when an ego network has a higher proportion of kin (relatives), it tends to exhibit greater homophily, less diversity, and greater clustering. Kinship ties often represent strong, enduring, and homophilous relationships that contribute to network closure and homogeneity.

In the context of segregation, Marsden’s theory suggests that larger networks are less segregated than smaller networks, independently of individual preferences. This rule is often contrasted with Blau’s rule (which focuses on numerical preponderance) and can be combined with it to explain patterns of diversity among minority groups.

Examples:

  • Core Discussion Networks: Marsden’s study using 1987 GSS data on “core discussion networks” (people with whom one discusses important matters) revealed these patterns. For instance, data from the GSS 1985 suggests variations in network size, with some people having many contacts and others being “isolated”. These networks also tend to be relatively homogeneous with respect to important social attributes like age, race, gender, and education.
  • Online Social Networks: Studies on online social networks, like Facebook, have also demonstrated Marsden’s rule. For example, analysis of Dutch, Turkish, and Moroccan Facebook users showed that the percentage of co-ethnic friends tends to decrease as the total number of friends increases, but only for members of minority groups (Hofstra et al. 2017), reflecting that larger networks can be less segregated but depend on the group size distribution of alters. This illustrates that regardless of preferences for “choice homophily,” larger networks may inherently lead to more diverse contacts because individuals, particularly members of minority groups, eventually “run out of ingroup alters”.

References

Blau, Peter M. 1974. “Presidential Address: Parameters of Social Structure.” American Sociological Review, 615–35.
———. 1977. “A Macrosociological Theory of Social Structure.” American Journal of Sociology 83 (1): 26–54.
Feld, Scott L. 1981. “The Focused Organization of Social Ties.” American Journal of Sociology 86 (5): 1015–35.
Hofstra, Bas, Rense Corten, Frank Van Tubergen, and Nicole B Ellison. 2017. “Sources of Segregation in Social Networks: A Novel Approach Using Facebook.” American Sociological Review 82 (3): 625–56.
Marsden, Peter V. 1987. “Core Discussion Networks of Americans.” American Sociological Review 52 (1): 122–31.
28  Collecting Ego-Network Data
30  Clique Analysis
Copyright 2023, Omar Lizardo & Isaac Jilbert