Social Capital & Network Characteristics Lecture 9
Social Capital & Network Characteristics
Lecture 9
Aims Lecture 9To understand:
To study characteristics of networks
Networks as social capital
The problem of only a structuralist approach+
+
+Homophily
Reciprocity
Centrality
Some Questions
Social network research requires general theories to answer:
In Network Research
Some Questions
Social network research requires general theories to answer:
Can the effects of networks (i.e., on behavior) be generalized across situations?
In Network Research
a.
Some Questions
Social network research requires general theories to answer:
Can the effects of networks (i.e., on behavior) be generalized across situations?
In Network Research
a.
Why certain network effects sometimes occur and sometimes not?b.
and if not,
i.e., Why is there more clustering in some networks than in others?
Structure & Social Capital
Structuralism
Structure overrides preferences
A first approach
You can explain people’s actions by only knowing the structure of their social network
Claims:
+
+
Structuralism
Structure overrides preferences
A first approach
You can explain people’s actions by only knowing the structure of their social network
Claims:
+
+
Give me the network & I will tell you what the actors will do
Selling Point
Labor markets (Granovetter, 1974)
Of this perspective
Illegal services: Abortion (Lee, 1969)
All markets:
+
+
Are socially organized in networks
Selling Point
Labor markets (Granovetter, 1974)
Of this perspective
Illegal services: Abortion (Lee, 1969)
All markets:
+
+
Are socially organized in networks
Role equivalence:Persons are tied not to the same persons but to similar persons
(Wasserman & Faust, 1994)
Selling Point
Labor markets (Granovetter, 1974)
Of this perspective
Illegal services: Abortion (Lee, 1969)
All markets:
+
+
Are socially organized in networks
Role equivalence:Persons are tied not to the same persons but to similar persons
(Wasserman & Faust, 1994)
Two positions in the aggregate: an elite person (well-connected) & a hanger-on (not well-connected)
Main ProblemsOf structuralism
It lacks a theory of individual behavior
Main Problems
Think about the micro-macro link
Of structuralism
It lacks a theory of individual behavior
+
Main Problems
Think about the micro-macro link
Of structuralism
It lacks a theory of individual behavior
+
Rational Choice Perspective:
Conceives networks as social resources
Main Problems
Think about the micro-macro link
Of structuralism
It lacks a theory of individual behavior
+
Rational Choice Perspective:
Personal networks can be treated as social capital that is instrumental in reaching our goals
Conceives networks as social resources
Not a new ideaSince Hobbe’s Leviathan:
Not a new idea
Thomas HobbesEnglish philosopher1588-1679
To have friends is to have power: for they are strengths united
Since Hobbe’s Leviathan:
Networks as Social Capital
Networks are treated as a specific resource important for most goals people have in life.
Networks as Social Capital
Two main propositions in S.C. Theory
Networks are treated as a specific resource important for most goals people have in life.
Networks as Social Capital
Two main propositions in S.C. Theory
Networks are treated as a specific resource important for most goals people have in life.
1 Social Resource Hypothesis: people better equipped with social capital will be better able to attain their goals
Networks as Social Capital
Two main propositions in S.C. Theory
Networks are treated as a specific resource important for most goals people have in life.
1 Social Resource Hypothesis: people better equipped with social capital will be better able to attain their goals
2 Investment Hypothesis: people will invest in social capital according to its instrumental value in producing their ends
Networks as S.C.It explains the emergence as well as the
effects of social networks
Networks as S.C.It explains the emergence as well as the
effects of social networks
A person’s social capital promotes her goal achievement
Networks as S.C.It explains the emergence as well as the
effects of social networks
She will invest in it depending on its instrumental value
&
A person’s social capital promotes her goal achievement
Networks as S.C.It explains the emergence as well as the
effects of social networks
Macro-micro link
She will invest in it depending on its instrumental value
&
A person’s social capital promotes her goal achievement
Networks as S.C.It explains the emergence as well as the
effects of social networks
Macro-micro link
She will invest in it depending on its instrumental valueMicro-macro link
&
A person’s social capital promotes her goal achievement
Homophily
Practical 11Choose links & Actions
Homophily
If instead of just looking at the network
We keep track of characteristics of the nodes (i.e., attributes)
Lazarsfeld & Merton (1954)
Homophily
If instead of just looking at the network
We keep track of characteristics of the nodes (i.e., attributes)
We tend to find that link nodes are similar to each other
Lazarsfeld & Merton (1954)
Homophily
If instead of just looking at the network
We keep track of characteristics of the nodes (i.e., attributes)
We tend to find that link nodes are similar to each other
Birds of a feather flock (will fly) together
Philemon Holland, 1960
Lazarsfeld & Merton (1954)
Real-life networksHomophily
Race & friendship networks US:
Interracial marriages US:
Gender & friendship networks
Real-life networksHomophily
Only 8% of people have any people of another race that they discuss important matters with (Marsden, 1987)
Race & friendship networks US:
Interracial marriages US:
Gender & friendship networks
Real-life networksHomophily
Only 8% of people have any people of another race that they discuss important matters with (Marsden, 1987)
Race & friendship networks US:
1% of white marriages, 5% of black marriages, 14% of asian marriages (Fryer, 2006)
Interracial marriages US:
Gender & friendship networks
Real-life networksHomophily
Only 8% of people have any people of another race that they discuss important matters with (Marsden, 1987)
Race & friendship networks US:
1% of white marriages, 5% of black marriages, 14% of asian marriages (Fryer, 2006)
Interracial marriages US:
Closest friends: 10% of men name a woman, 32% of women name a man (Verbrugge, 1977)
Gender & friendship networks
Real-life networksHomophily
Only 8% of people have any people of another race that they discuss important matters with (Marsden, 1987)
Race & friendship networks US:
1% of white marriages, 5% of black marriages, 14% of asian marriages (Fryer, 2006)
Interracial marriages US:
Closest friends: 10% of men name a woman, 32% of women name a man (Verbrugge, 1977)
Gender & friendship networks
In all cases lower than if ignoring attributes
Possible ExplanationsReasons for Homophily
Opportunity (Contact Theory):
Benefits/Costs:
Possible ExplanationsReasons for Homophily
The possibility that you meet people could be biased by attributes (i.e, race)
Opportunity (Contact Theory):
Benefits/Costs:
More of a chance of meeting your own type
Possible ExplanationsReasons for Homophily
The possibility that you meet people could be biased by attributes (i.e, race)
Opportunity (Contact Theory):
Benefits/Costs:
More of a chance of meeting your own type
Common attributes (i.e., language, culture, knowledge) make it easier
Possible ExplanationsReasons for Homophily
The possibility that you meet people could be biased by attributes (i.e, race)
Opportunity (Contact Theory):
Benefits/Costs:
Also social pressure or social competition
More of a chance of meeting your own type
Important:
Common attributes (i.e., language, culture, knowledge) make it easier
Possible ExplanationsReasons for Homophily
The possibility that you meet people could be biased by attributes (i.e, race)
Opportunity (Contact Theory):
Benefits/Costs:
Also social pressure or social competition
More of a chance of meeting your own type
Common attributes (i.e., language, culture, knowledge) make it easier
Important:
The structure of the network depends on the characteristics
i.e., why communication might circulate among one group and not another?
Quick Summarytwo points
From Structuralism:
From Homophily (Segregation Patterns):
Quick Summarytwo points
From Structuralism:
The characteristics of the network matter. They affect the individuals
From Homophily (Segregation Patterns):
Quick Summarytwo points
From Structuralism:
The characteristics of the network matter. They affect the individuals
From Homophily (Segregation Patterns):
The characteristics of the individuals matter. They affect the structure of the network
Reciprocity
ReciprocityLocal Patterns
ReciprocityLocal Patterns
Directed Networks
ReciprocityLocal Patterns
Directed Networks
A node can be linked to another without the second being linked to the first (i.e., webpages)
ReciprocityLocal Patterns
Directed Networks
A node can be linked to another without the second being linked to the first (i.e., webpages)
ij in g does not imply ji in g
ReciprocityLocal Patterns
Directed Networks
A node can be linked to another without the second being linked to the first (i.e., webpages)
Reciprocity
There is a tendency to dyadic reciprocation in most directed networks
ij in g does not imply ji in g
ReciprocityLocal Patterns
Directed Networks
A node can be linked to another without the second being linked to the first (i.e., webpages)
Reciprocity
There is a tendency to dyadic reciprocation in most directed networks
ij in g does not imply ji in g
if ij in g it is more likely ji in g
ReciprocityExplanation
ReciprocityExplanation
Mutual Dependence
ReciprocityExplanation
Mutual DependenceActors (i.e., players, people) depend on each other for valued outcomes, and benefits will be received from another actor only if they are also given in return
(Emerson, 1972)
ReciprocityExplanation
Mutual DependenceActors (i.e., players, people) depend on each other for valued outcomes, and benefits will be received from another actor only if they are also given in return
Think aboutCooperation: if relations are not reciprocated they are likely to be terminated more rapidly
(Emerson, 1972)
ReciprocityExplanation
Mutual DependenceActors (i.e., players, people) depend on each other for valued outcomes, and benefits will be received from another actor only if they are also given in return
Think aboutCooperation: if relations are not reciprocated they are likely to be terminated more rapidly
(Emerson, 1972)
Keeping a non-reciprocated relation implies status deference
tend to be eliminated
Node Centrality
Node CentralityPositions in Networks
Node CentralityPositions in Networks
Who are influential, powerful (Think of our Facebook Example with Ana)
How different nodes are positioned in the network?
Node CentralityPositions in Networks
Who are influential, powerful (Think of our Facebook Example with Ana)
How different nodes are positioned in the network?
Many social networks show a fair extent of centralization
Node CentralityPositions in Networks
Who are influential, powerful (Think of our Facebook Example with Ana)
How different nodes are positioned in the network?
Many social networks show a fair extent of centralizationdifferentiation between social actors with
respect to their centrality
Node CentralityExample
Ways of measuring centrality
Node Centrality
2
1 4
1 24
1 3
3 2
3 2
2
2
Example
Ways of measuring centrality
Node Centrality
2
1 4
1 24
1 3
3 2
3 2
2
2
Both white nodes have degree 2 (degree centrality)
The first seems more central - neighbors (3) & (4): (betweenness)
Better connected in another sense
Example
Ways of measuring centrality
Node Centrality
2
1 4
1 24
1 3
3 2
3 2
2
2
There are many other measures of centrality
Both white nodes have degree 2 (degree centrality)
The first seems more central - neighbors (3) & (4): (betweeness)
Better connected in another sense
Example
Ways of measuring centrality
Node CentralityWhy do we observe it?
It reflects social organization and opportunities
Node CentralityWhy do we observe it?
It reflects social organization and opportunitiesA strongly centralized network increases the likelihood of collective action in mobilizations - easier contact to others
(Marwell, Oliver & Prahl, 1988)
Node CentralityWhy do we observe it?
Result of feedback processesFavoring the creation of links to nodes that are already highly connected
It reflects social organization and opportunitiesA strongly centralized network increases the likelihood of collective action in mobilizations - easier contact to others
(Marwell, Oliver & Prahl, 1988)
Node CentralityWhy do we observe it?
Result of feedback processesFavoring the creation of links to nodes that are already highly connected
It reflects social organization and opportunitiesA strongly centralized network increases the likelihood of collective action in mobilizations - easier contact to others
(Marwell, Oliver & Prahl, 1988)
Unto him that hath is given and from him that hath not is taken away, even that which he hath
The Matthew effect (Merton, 1968)
Node CentralityWhy do we observe it?
Result of feedback processesFavoring the creation of links to nodes that are already highly connected
It reflects social organization and opportunitiesA strongly centralized network increases the likelihood of collective action in mobilizations - easier contact to others
(Marwell, Oliver & Prahl, 1988)
Unto him that hath is given and from him that hath not is taken away, even that which he hath
The Matthew effect (Merton, 1968)
However, centralization is most likely in physical networks: Internet hubs
Checklist
Both structure of the network & individual behavior (and characteristics) influence each other
Checklist
Both structure of the network & individual behavior (and characteristics) influence each other
Checklist
People use their social networks as a form of capital that helps them achieve what they want
Checklist
People use their social networks as a form of capital that helps them achieve what they want
Social networks portray different properties:
Both structure of the network & individual behavior (and characteristics) influence each other
Checklist
People use their social networks as a form of capital that helps them achieve what they want
Social networks portray different properties:
Individuals with common traits are likely to be related (Homophily)
Most relationships are reciprocal (both parts aim for it)
We can look locally at who is influential (centrality)Important for diffusion of information
Both structure of the network & individual behavior (and characteristics) influence each other
Questions?