Professional Documents
Culture Documents
Tom 0 0 1 1 0
Dick 0 0 1 1 0
Sally 1 1 0 0 1
Fred 1 1 0 0 0
Alice 0 0 1 0 0
Basic sociograms
Measurement of relational properties in
SNA
• Degree (in-degree, out-degree, average degree):
• Density:
• The actual number of connections in a network expressed as a proportion of
the total possible number of connections.
• High density should generate greater: trust, cultural homogeneity and
diffusion speed
• Not easy to make meaningful comparisons of density across networks of
different sizes (or involving different types of relations)
• Components:
• sub-graphs within a graph, any of whose vertices can be reached from any
other (by whatever route)
• Important because the holes that separate different components within a
graph prevent the flow of resources and might indicate tribal or ideological
divides.
• Cliques:
• a subset of vertices within a component, all of whose members are related to
one another (density = 1): fully connected component
• K-Cores:
• a subset of vertices within a component, all of whose members enjoy a
specified number of relations (=‘k’) with the others: e.g. a 6-core is a subset, all
of whose members enjoy relations with at least 6 of the other members
• Helps analysis of grouping and relative power
Measurement of relational properties in
SNA
• Centrality:
• Central vertices enjoy both advantages and disadvantages (potentially)
• Centralisation:
• Measures focused upon the network as a whole
• The most basic of these take variance measures for one of the centrality scores (or
something very similar)
• Centralised networks:
• may be more easy to organise
• are often robust in relation to random attack/damage
• But vulnerable to targeted attack
• Erdos/Bacon number(s):
• The concept of ‘degree’ is also used in this context (and the concept of ‘walk’).
An individual with a Bacon number of 4 is separated from Bacon by ‘four degrees of
separation’.
• The ‘small world’ phenomenon is sometimes referred to as ‘six degrees of separation’
• Diameter:
• The shortest possible path between the two most distant vertices (measured in
degrees)
• Or the (mean) average path distance between all possible pairs of vertices
• Relevant because resources, viruses etc. will take longer to travel across large
diameters.
• Information/gossip will be more liable to distortion (Chinese whispers) the more
mouths it passes through
Terminology cont.
• Network density
• Length and distance:
• Length of path is number of links
• Distance between two nodes is length of shortest path (aka geodesic)
• Component:
• Maximal sets of nodes in which every node can reach every other by some
path (no matter how long). E.g. connected has one
• Cut points:
• Nodes which, if deleted would disconnect net
• Bridge:
• A tie that if removed would disconnect network
• Core/periphery structure:
• Does the network consists of single group ( a core) together with hangers on
(periphery) or
• Are there multiple subgroups each with their own periphery
Terminology Cont.
• N-Clique:
• A set of nodes that are within distance n of each other
• K-Plex:
• A set of n nodes in which every node has tie to at least n-k others in the
set
• Faction:
• A set of mutually exclusive groups of actors such that density of ties
within group is greater than density of ties between groups
• Degree centrality:
• The number of nodes adjacent to given node
• Closeness centrality:
• Sum of geodesic distances to all other nodes, inverse measure of
centrality
• Betweenness centrality:
• Loosely: number of times that a node lies along the shortest path within
two others
• Types of relations:
• Bilateral: non directed nets.: Marriage, friendship aliance
• Unilateral: directed nets: advertising or links to web sites
Eigenvector Centrality
• Importance of an actor in a network
• Sociomatrix (Adjacency matrix)
• Aij = 1, if a link between i and j; 0 otherwise
• Centrality measure xi
Degree Centrality:
• In matrix form:
• Here λ is the eigenvalue
Closeness Centrality
• The measure focuses on how close an actor is to all the other actors in
the set of actors
• An actor is central if it can quickly interact with all others
• The geodesics, or shortest paths : minimum distance
• Actor Closeness Centrality : a function of geodesic distances
depends not only on direct ties but also on indirect ties
Closeness Centrality
• As actors who can reach i become closer, on average, then the ratio becomes larger
Terminology
• Structural equivalence:
• Actor i and j are structurally equivalent if actor i has a tie to k, iff j
also has a tie to k, and i has a tie from k iff j also has a tie from k
• Helps to reduce graph
• Measures of subgroup cohesion:
• A measure of degree to which strong ties are within rather than
outside is given by the ratio
• the average strength of the ties within and the denominator
is the average strength from subgroup members to
outsiders
Why Is Social Capital Important?
• Using an example of OSS
• OSS is developed by voluntary developers through individual
incremental efforts and collaboration.
• New contributions to the code often involve, to a large extent, a
recombination of known conceptual and physical materials (Narduzzo
and Rossi 2003, Fleming 2001).
• Developers with better access to and familiarity of such materials are
advantaged in their code development efforts.
• Because information about and knowledge of resources often lies spread
across developers in the community, social capital, i.e. a developer’s
access to resources from a network of relationships, may emerge as a
key factor that differentiates those who are more productive than others.
Network Relationships and Knowledge Benefits
• Where
• Mi → number of direct ties for developer i
• piq → proportion of i’s relations invested in the relationship with j
Social Capital and OSS Success
• Singh, Tan, and Mookerjee (2007)
• Data
• 5191 projects and 10973 developers
Data Collection
• The challenge is determining network boundary
• Two approaches
• Whole network
• Not easy
• Ego-centric
• Problematic
• Snowballing
Social Networks and Marketing.
Van den Bulte, Christophe and Stefan Wuyts (2007), Cambridge, MA: Marketing Science
Institute.
Related theories
• Social capital (7 pages):
• Resource of direct contacts
• Number/strength of ties
• Closure and local clustering
• Bridge positions, spanning, structural holes
• Strength of weak ties
• The dark side of social capital
• Network among customers (16 pages):
• Social contagion and new product
• Social influence and existing brands
• Transaction networks
Social capital
• Idea: how are the social networks tied to social capital?
• Resources of direct contacts;
• Number of ties
• Strength of ties
• Closure and local cluster
• Bridge positions spanning structural holes
• Strength of weak ties
• Resources of direct contacts:
• Resource level
• Resource diversity
• Alters’ social capital (“ego”; “alter”)
Social capital
• Number of ties
• Strength of ties
• Always better than weak ties
• Being strong:
• High frequency
• Long-standing
• High-valence
• Multiplicity
• Closure and local Clustering
• Bridge positions spanning structural holes
• Weak ties can be strong: e.g. job hunting
Network among customers
• Contagion through networks
• Products, technology, disease, gossip
• Reasons for contagion:
• Awareness and interest
• Belief updating (revision/confirmation)
• Normative pressure
• Competitive concerns
• Complementary network effects (number of prior adoptions)
• Social network structure and contagion:
• Macro level:
• Transitivity
• Clustering & “Bridges”– academic research in different fields
• Customer’s acceptance threshold
Networks among customers
• Social network structure & contagion (cont. )
• Micro-level:
• Direct ties matter? Types of ties: association with the incentives
(belief updating, normative pressure, competitive concern…)
• Overall network density and structure
• Networks among Customers:
• “Opinion leaders”: True or false?
• There are generalized opinion leaders
• They are a particular type of people
• One is either opinion leader or opinion seeker in a network
• They are up to date about new products
• Early adopters of new products
• Have central location in the network
• Engaging in many conversations about new products
Networks among customers
• Other issues:
• Buzz marketing – do not exploit network structure
• Viral marketing (WOM online)– considers contagion dynamics
while ignoring the network structure (Hotmail AD)
• Social influence and existing brands:
• How does information spread?
• Preference and belief on brands
• Transaction networks
Related theories
• Intra-organizational networks (6 pages):
• Power
• Disseminating market knowledge within the firm
• Employee recruitment and Turnover
• Industrial buying
• Inter-organizational networks (7 pages):
• Coordinating channels
• Accessing resources through alliances
• Gaining status from one’s network
• Managing competition
• Adopting new suppliers, technologies, and practices
Characteristics of Organizational Networks
• Intra-Organizational • Inter-Organizational
Networks Networks
• Power • Coordination of
• Dissemination of Channels
knowledge • Accessing ressources
• Employee recruitment • Gaining Status
and Turnover • Managing Competitions
Intra-Organizational Networks
• Power
• Associated with Central positions in workflow, advice, and
friendship neworks
• Related to measures of centrality: degree, closeness(related to
access, especially to dominant actors in the firm), betweenness
(information control)
• Research shows that closeness to important actors in the firm is
related to both power and promotion. However link of causality not
clear.
• Burkhardt and Brass (1992) show that centrality preceded
power
• Kilduff and Krackhardt (1994) concluded for the firm they
studied that the perception people have of the quality of your
network is associated with your reputation.
• In another study, Krackhardt (1990) shows that accurate
perception of the network structure is linked to power.
Intra-Organizational Networks
• Dissemination of Knowledge
• Market information must be shared
• How Marketers obtain information? Through their Network?
• Most rely on Market reports and Mkt information systems
• Little work is done on how marketers network affect their knowledge
of the market.
• Complexities are involved
• Under conditions of high uncertainties, Bankers relied on strong ties for
support on closing deals
• Transactions in which they used sparse approval networks more likely to
succeed than when they used dense approval network.
Intra-Organizational Networks
Disseminating Knowledge (ctd)
• Little work on the role of social networks
• Most work pertain on innovation:
• The Boundary-spanning actors
• Innovation comes from people holding a boundary-spanning
position. Allen (1997) Hutt, Reingen, and Ronchetto (1988)
• Actors with those positions have high betweeness centrality and
span structural holes between their formal unit and the rest of the
network. They are key actors for the flow of ideas.
• While they may not continually be involved in projects, They are very
important to projects because they act as mediators or bridges
between people during projects.
Intra-Organizational Networks
Disseminating Knowledge (ctd)
•Most work pertain on innovation:
• Strong ties vs weak ties
• Strong ties are related to the ability to mobilize
ressources to get project going. a
• Strong ties to many actors or access to many actors
through strong ties critical.
• May be less effective in locating novel information but
good in conveying complex information
• Weak ties good for search of useful information in
other subunits but bad in the transfer of complex
knowledge
Intra-Organizational Networks
Disseminating Knowledge (ctd)
• Closure or local clustering
• Affect willingness and motivation of individuals to
invent time, energy and effort in sharing kowledge
with others
• Teams with high closure among team members,
but with low closure among their contacts
tended to be more efficient.
• Lack of Redundancy
• Ability to access pool of nonoverlaping information
• Greater confidence in information received
• Results of Study:
• combining closure and non redundancy boost innovation. Reagans
and Zuckerman (2001), Uzzi and Spiro (2005).
Intra-Organizational Networks
Disseminating Knowledge (ctd)
• Organization of the company for creativity and innovation
• Adequate networking critical for success
• Example of PARC: Successful in generating new ideas, but
operated independendly of from corporate headquaters, and
from commercial divisions. Failed to get their new product
commercialized.
• Example of 3M: Created intermediary unit to link innovation to
market. OTC acted as a technology broker to find new
applications.
• Companies can organize themselves as constellation of
communities of similar activities. Insight will be obtained to know
who will benefit from interracting with whom. Companies will be
more efficient.
Intra-Organizational Networks
• Employee Recruitment and Turnover
• Metric important especially in the service industry
• Formal network referral programs
• Employers are provided with a wider pool of applicants
• Readily available support for newly hired employees
• Referrals are more productive, perform better, and longer
committment than non referrals. Castilla (2005)
• Cost savings. Fernandez, Castilla, and Moore (2005)
Intra-Organizational Networks
Employee Recruitment and Turnover(ctd)
• Employee Turnover
• Network effect: Social contagion. Krackhardt and Porter(1985,
1986). People will leave following people of similar position in the
advice network.
• However the friends of the leavers became more satisfied with the
company and remained longer than those who were not friends with
the leavers.
• Employee advancement
• Networks within the firms are used for advancement
• Minorities suffer disadvantage because network s structure makes it
less easy to use informal sources of information.
• Customer porfolio may affect career pattern. Beckman and Philips
(2005).
Inter-Organizational Networks
• H2a: When hubs adopt, the overall adoption process speeds up.
H2b: Innovator hub adoption has a larger correlation with speed of
adoption than follower hub adoption.
• More connections will be activated once hubs adopt → adoption rate ↑
• Since innovator hubs adopt earlier than follower hubs, they have
more time to influence the network.
• H3: The higher the relative out-degree of a hub,
the greater impact it has on adoption.
• in-degree: the number of people who convey information to hubs
→ related to when a hub adopts.
• out-degree: the number of people to whom hubs convey information
→ the influence of hubs on subsequent adoption
Hypotheses
• H4a: Hubs adoption increases the eventual size of a market.
H4b: Follower hubs have a stronger relation to market size
than
innovative hubs.
• Adoption by hubs → exposure of an innovation to the market ↑
→ market size ↑
• Homophily effect
: follower hubs are more similar to most of the population in terms of
innovativeness → more influence on the main market.