THE SMALL WORLD PROBLEM

'Social structure becomes actually visible in an anthill; the movements and contacts one sees are not random but patterned. We should also be able to see structure in the life of an American community if we had a sufficiently remote vantage point, a point from which persons would appear to be small moving dots... We should see that these dots do not randomly approach one another, that some are usually together, some meet often, some never... If one could get far enough away from it, human life would become pure pattern.'
-- Roger Brown, INSNA website.

SIX DEGREES OF SEPARATION

Many people first heard of this idea from the movie (based on the Broadway play by John Guare) of this name. But the original concept came from work in the 1960s by Stanley Milgram (yes, the 'electric shock research' guy!) on what's called the small world problem': how long a chain of people would it take to deliver a message from anyone in the world to any other person, using only the 'friends of friends'? Milgram found that there was a chain of only about 5.5 people (6 people if the 'sender' and 'receiver' are of different races) between any 2 people in the world.

ANOTHER SIX DEGREES OR SO

There's a party game called 'The Six Degrees of Kevin Bacon'. The object of the game is to determine, for any actor, how many 'Bacon degrees' (that is, how many links through co-stars in films they've been in) it takes to reach Kevin Bacon himself. For example, Tom Hanks is 'one Bacon degree'(they were in 'Apollo 13' together), while Mick Jagger is 'three Bacon degrees' away (anybody see Jagger's Australian Western?) The object of the game is to connect the entire 'Hollywood social system' via 'Bacon degrees'. (The most 'Bacon degrees' anyone's found is eight so far). There is also now a commercial site called "Six Degrees" that claims to link you with all your own personal network of contacts (assuming they sign up!)

SOCIAL NETWORK ANALYSIS CLAIMS THAT

Any individual person (point) is informally connected via family, friendship, occupation, etc. (ties) to other people through which the structure of any particular social system can be identified (networks). This social system can be a neighborhood, a village, a bowling team or any other (informal) group of people. A formal group (such as a company bureaucracy) can be studied this way, using the informal connections that people use to 'get things done around here.' Here's a story from CIO Magazine about a UCLA management professor who provides social network software and consulting services to help companies understand how they really work internally.

A DIFFERENT KIND OF DATA

Network data are not attributional but relational variables. Attributes are the intrinsic (and measurable) characteristics of individual people, objects or events. Relations are the linkages between individuals, organizations, nations. Relational variables can be:
transaction relations (e.g., who gives gifts to whom) communication relations (e.g., who talks to whom)
boundary penetration relations (e.g., who's on whose board of directors)
instrumental relations (e.g., who asks whom for expert advice)
sentiment relations (e.g., friendship cliques in high school)
power relations (e.g., who follows whom in informal group)
kinship relations (e.g., who's related to whom)

THE TIES THAT BIND

Network researchers are particularly interested in:
durability (how long ties last) reciprocity (whether ties look identical from either end)
intensity (whether ties are 'weak' or 'strong')
density (how many potential ties in a network actually exist)
reachability (how many ties it takes to go from one 'end'of a network to the opposite 'end')
centrality (whether a network has a 'center' point or points)

MORE ABOUT MULTIPLEX

One of the major research questions in social network analysis is the 'multiplexity of network ties.' Each 'tie' between two people can be made of several different 'strands':

they might go to the same school
AND be third cousins
AND be on the same softball team
AND hardly know anything about each other.
Alternatively, two other people might have those exact same connections and be 'best friends'. So, 'ties' can be operationalized with many different dimensions depending on the researcher's interests (just like regular variables!)

STRUCTURAL HOLES

Another area of growing interest has been popularized by well-known researcher Ronald Burt: 'structural holes' (gaps in and between networks and what they may mean.)
The analysis of industry networks is becoming particularly popular, focusing on possible areas of cooperation and/or competition among firms.

FROM SOCIOMETRY...

Originally called 'sociometry', it started with Jacob Moreno's 1934 book 'Who Shall Survive?', which developed a 'geometry of interpersonal relationships' by mapping people's friendship choices as a way of testing their psychological well-being.
This 'geometry' was formalized by Cartwright and Harary using 'graph theory' during the 1950s with the idea of representing groups of people as collections of points connected by:

'signed' (+ means 'likes' and - means 'dislikes') 'directed' (arrow from person A to person B and vice-versa)
ties (lines) which form a network structure.

TO SOCIAL NETWORK ANALYSIS

'Sociometry' evolved into 'social network analysis' during the 1960s due to the influence of Harrison White at Harvard and his use of set theory to develop algebraic models of kinship groups in anthropology. He also used multidimensional scaling techniques to represent 'social distance' in social network space. Famous examples of this are Lee's 1969 article, 'The Search for an Abortionist', in which she found that pregnant women had to approach an average of 5.8 people in order to find information about obtaining a (then-illegal) abortion, and Granovetter's 1973 article, 'The Strength of Weak Ties', in which he discovered that jobseekers found more new jobs by asking distant acquaintances (but not strangers) rather than close friends. Currently, social network analysis is very hot in computer-mediated communications environments, as it's an excellent way to trace the transfer of information among participants.

WHAT ARE THE TECHNIQUES

Researchers study social networks through:

observation interviews
surveys
'trace analysis' (of citations, letters, e-mail, or other written communications)
Researchers describe social networks through:

'sociograms' (visual displays of networks) 'block-modelling' (using a data matrix to show the units of analysis and their connections and

IF SOCIAL NETWORK ANALYSIS IS THE ANSWER, WHAT IS THE QUESTION?

Social network analysis can be either:
'micro'(focusing on the individual and his/her network)
'macro' (focusing on the network(s) and its members). It has been used to study:
diffusion of innovation spread of epidemics
scholarly citations
interlocking corporate directorates
hyperlinks among Web sites
criminal conspiracies...
in short, anything that may involve 'exchange relationships' among people or organizations

FOR MORE INFORMATION

See "Network Research" Section of Research Resources Page       http://web.archive.org/web/20041010110545/http:/web.syr.edu/~bvmarten/socialnet.html