Networks I Flashcards
Collective intelligence
Impacts team performance
Improve CI
- Equal air time
- Social perceptiveness
Groups are just as prone to biases as individuals
Nodes
People, entities, orgs
Ties
Relationships
Information
Information you cannot get from any other person
Directed
Direction matters
Undirected
Direction does not matter
How do you get network data?
- Individual data through interviews, assessments, Outlook Exchange
- Full network data through surveys
- Full network data as by-product (email)
Unobtrusive sources of network data
- Public records
- Comapny records
- Observation
- Evil methods
What do we want to know from a network?
- Who’s important or “central”? Who are connectors, mavens, salesmen?
- Are there outliers or isolates?
- How dense (well-connected) is the network?
- Are there groups/clusters/cliques?
- Why does the network look like this?
Centrality
- Degree- how many people am I tied to?
- Betweenness - how often am I on the shortest path (“geodesic”) between two other people?
- Closeness: how easily can I reach every other actor in the system?
- Eigenvector: how well-connected are those I’m connected to?
Degree
Degree: # of contacts
Degree doesn’t give all info
Lois may be a connector where Bill is insular
In-degree
“received ties” number of people who name you
- high in-degree: you are a maven
- ofted used as a measure of prestige
Out-degree
“sent ties”: number of people whom you name
- high out degree: you are (a) clueless or (b) consultative
- Adv.: (out) degree can be assessed by individual
- Dis.: can mis-represent importance
Individual measures of centrality
Degree
Full network-based measures
Geodesic
Closeness centrality
Betweenness centrality