Click to hide or show the feedback form

Bookmark and ShareBookmark


Network Exercise 

Designed to Enhance Students' Understanding

of Size, Strength of Relationships, and Diversity

in Social Networks


Developed by Caroline Hodges Persell

Department of Sociology, New York University

January 2008


With major reliance on:

Malcolm Gladwell. 2002. The Tipping Point: How Little Things Can Make a Big Difference. Boston: Little, Brown & Co.
David S. Hachen Jr. 2001. Sociology in Action: Cases for Critical and Sociological Thinking. Thousand Oaks, CA:  Pine Forge Press.


A social network is a set of relations, links, or ties among social actors.


The purpose of this exercise is to show that individuals vary in terms of their social networks, networks vary in terms of certain of their characteristics, and social networks serve various functions and have social consequences.  Three specific characteristics of networks will be analyzed: 1) their size, 2) the strength of relationships in them, and 3) the diversity of relationships in social networks.


Part I:


The first part of this exercise may be done in or out of class.  It is designed to illustrate variation in the size of social networks.


Have each person in the class take and self-score the following "test" devised by Malcolm Gladwell:


In the paragraph below is a list of around 250 surnames, all taken at random from the Manhattan phone book.  Go down the list and give yourself a point every time you see a surname that is shared by someone you know.  (The definition of 'know' here is very broad.  For example, if you sat down next to that person on a train, you would know their name if they introduced themselves to you and they would know your name.)  Multiple names count.  If the name is [Baptista], in other words, and you know three [Baptistas], you get three points.  The idea is that your score on this test should roughly represent how social you are.  It's a simple way of estimating how many friends and acquaintances you have.  [List of names]


I have given this test to at least a dozen groups of people.  One was a freshman World Civilizations class at City College in Manhattan.  The students were all in their late teens or early twenties, many of them recent immigrants to America, and of middle and lower income.  The average score in that class was 20.96, meaning that the average person in the class knew 21 people with the same last names as the people on my list.  I also gave the test to a group of health educators and academics at a conference in Princeton, New Jersey.  This group were mostly in their forties and fifties, largely white, highly educated-- many had Ph.D.'s-- and wealthy. Their average score was 39.  Then I gave the test to a relatively random sample of my friends and acquaintances, mostly journalists and professionals in their late twenties and thirties.  The average score was 41.  These results shouldn't be all that surprising.  College students don't have as wide a circle of acquaintances as people in their forties.  It makes sense that between the ages of twenty and forty the number of people you know should roughly double, and that upper-income professionals should know more people than lower-income immigrants.  In every group there was also quite a range between the highest and lowest scorers.  That makes sense too, I think.  Real estate salesmen know more people than computer hackers.  What was surprising, though, was how enormous that range was.  In the college class, the low score was 2 and the high score was 95.  In my random sample, the low score was 9 and the high score was 118.  Even at the conference in Princeton, which was a highly homogeneous group of people of similar age, education, and income-- who were all, with a few exceptions, in the same profession-- the range was enormous.  The lowest score was 16.  The highest score was 108.  All told, I have given the test to about 400 people.  Of those, there were two dozen or so scores under 20, eight over 90, and four more over 100.  The other surprising thing is that I found high scorers in every social group I looked at.  (Gladwell, 2002: p. 38-41)


After everyone in the class has taken this test, and scored it by totaling the number of people on the list they know, graph the results in class, either on the board or on a computer (for example, by using a spreadsheet program such as Excel).  Judging from Gladwell's results, you might want to set the X axis of the graph as 0-100.  You want to show the distribution (i.e., frequency) of the responses across this axis.  You also want to compute the average score for the class.


As you look at the distribution, try to formulate hypotheses about what social factors might be related to having above average or below average scores in the class and in Gladwell's examples.  For starters, you might reflect on why Gladwell mentions age, being recent immigrants, occupation, and income status as factors that might affect the number of network ties?  Other possibilities that you might think of are knowing many people in school (especially if the school is large) perhaps because of belonging to school-wide groups or organizations such as band, chorus, a big sports team, or a club; having moved more than once; belonging to multiple outside groups or organizations, being a member of a very large family; or having traveled a lot.  This is a chance to practice developing sociological (as opposed to dispositional) hypotheses to explain variation.  [See Unit I on "The Sociological Perspective" for an explanation of the distinction between individual and sociological explanations or hypotheses.]


            By the end of this part, you should understand that social actors vary in terms of the number of social ties they have, and be familiar with some possible explanations for that variation.


Part II: 


This section of the exercise is designed to increase your understanding of the strength of the links among social actors in social networks. In class discussion, you could reflect on what might make network ties stronger or weaker.  Some possibilities might include how often they see someone, whether the tie is with a family member or a friend, how long they have known each other, how "close" they feel to someone and how close they think that person feels to them, the degree to which they are interdependent, i.e., they help each other out, exchange favors, etc.


Sociologist David S. Hachen Jr. (2001) suggests that you can identify social ties and their strength by "looking for flows of resources between social actors" and he suggests six possible flows to consider:


Personal Evaluations:  Look at who likes whom, who is friends with whom, who avoids whom, and who dislikes or hates whom.  Friendship networks, social cliques at a church, and prestige hierarchies within communities are all examples of networks created by personal evaluations.


Transfers of Material Resources:  Look at how money, capital, commodities, services, and other valuable material resources flow.  Focus on exchanges in which A gives B something in exchange for B giving A something. Examples include who contributes to a local charity, whom banks lend to and where they borrow from, and buying and selling in markets.


Information: Who talks to whom?  Who communicates with whom?  Though networks, messages are sent and received, creating information networks.  Information networks are important in spreading gossip, learning about job openings, and diffusing innovations.


Movement of People:  Look for the flows of people between places, organizations, or occupations.  For example, some accounting firms recruit a large number of their new accountants from specific business schools.  Occupations often are linked by the flow of people such as the recruitment of principals from [the ranks of]) teachers.


Formal Roles:  Look at the rules and regulations that prescribe who can tell who what to do.  Command hierarchies in organizations depicted in organizational charts are examples of such networks.


Kinship:  Look at who is related to whom either by descent or by marriage.  (Hachen, 2001: p. 27-28).


By the end of this part, you should understand that the strength of ties between social actors may vary, and be familiar with some possible reasons for that variation.


Part III:


            This part of the exercise is designed to enhance your understanding of the diversity of relationships.  It is not just the number of links social actors have and the strength of those ties, but the social locations of the people to whom they are connected that are important.


            Malcolm Gladwell is helpful here as well, when he writes:


Perhaps the best way to understand this point is through the popular parlor game 'Six Degrees of Kevin Bacon.'  The idea behind the game is to try to link any actor or actress, through the movies they've been in, to the actor Kevin Bacon in less than six steps.  So, for example, O.J. Simpson was in Naked Gun with Priscilla Presley, who was in Ford Fairlane with Gilbert Gottfried, who was in Beverly Hills Cop II with Paul Reiser, who was in Diner with Kevin Bacon.  That's four steps.  Mary Pickford was in Screen Snapshots with Clark Gable, who was in Combat America with Tony Romano, who, thirty-five years later, was in Starting Over with Bacon.  That's three steps.  Recently, a computer scientist at the University of Virginia by the name of Brett Tjaden actually sat down and figured out what the average Bacon number is for the quarter million or so actors and actresses who have played in television films or major motion pictures and came up with 2.8312 steps.  Anyone who has ever acted, in other words, can be linked to Bacon in an average of under three steps.  That sounds impressive, except that Tjaden then went back and performed an even more heroic calculation, figuring out what the average degree of connectedness was for everyone in Hollywood who had ever acted in Hollywood.  For example, how many steps on average does it take to link everyone in Hollywood to Robert DeNiro or Shirley Temple or Adam Sandler?  Tjaden found that when he listed all Hollywood actors in order of their 'connectedness,' Bacon ranked only 669th.  Martin Sheen, by contrast, can be connected to every other actor in 2.63681 steps, which puts him almost 650 places higher than Bacon.  Elliot Gould can be connected even more quickly, in 2.63601.  Among the top fifteen are people like Robert Mitchum and Gene Hackman and Donald Sutherland and Shelly Winters and Burgess Meredith.  The best-connected actor of all time?  Rod Steiger.


Why is Kevin Bacon so far behind these actors?  One big factor is that Bacon is a lot younger than most of them and as a result has made fewer movies. But that explains only some of the difference.  There are lots of people, for example, who have made lots of movies and aren't particularly well connected.  John Wayne, for example, made an extraordinary 179 movies in his sixty-year career and still ranks only 116th, at 2.7173.  The problem is that more than half of John Wayne's movies were Westerns, meaning that he made the same kind of movie with the same kind of actors over and over again.


But take someone like Steiger: he has made great movies like the Oscar-winning On the Waterfront and dreadful movies like Car Pool.  He won an Oscar for his role in In the Heat of the Night and also made "B" movies so bad they went straight to video.  He's played Mussolini, Napoleon, Pontius Pilate, and Al Capone.  He's been in thirty-eight dramas, twelve crime pictures and comedies, eleven thrillers, eight action films, seven Westerns, six war movies, four documentaries, three horror flicks, two sci-fi films, and a musical, among others.  Rod Steiger is the best-connected actor in history because he has managed to move up and down and back and forth among all the different worlds and subcultures and niches and levels that the acting profession has to offer."  (Gladwell, 2002: p. 46-48)


            In the case of Rod Steiger, the diversity of his network was related to its size and to the social locations (in his case the varied types of films) he had occupied.  Some people, like John Wayne, spend most of their time in one social location (i.e., type of film).  As a result, they know fewer other people and less diverse ones.  Some social actors have links with many different social locations, while others have links with relatively few.  The diversity of a social actor's network may have important social consequences.  For example, social actors whose networks are diverse may serve as "bridgers" between social worlds that are not always connected.  For example, individuals who are in interfaith or interracial marriages have the potential of having links with people in more than one faith or race.  The presence of such bridgers reduces the chances that a society will develop sharp cleavages between social groups that might otherwise be socially isolated.


Part IV:


What do you think are the social consequences of variations in the size, strength, and diversity of social networks?  Hachen discusses four things that happen through social networks:


Diffusion:  Networks can facilitate the diffusion of ideas and influence.  A good example is rumors…..


Exchanges:  Through networks, people make exchanges with others.  Nobody is totally self-sufficient.  People exchange with others to obtain things they need to do what they want to do…. It is important when mapping out a network to look at where resources are located within the network. If strategic resources are located in only a few places in a network, others within the network will be dependent on those social actors.  In the extreme, you can have networks with monopolies in which only one social actor controls an important resource that others within the network need.


Social Support:  Social networks can provide social support. When people move to a new town or country, they often have a difficult time because they have lost their support network, that is, social ties that can provide assistance and information in times of need….  Support networks can be very important in finding a job (Granovetter 1995) and, for those who are really having a difficult time, finding shelter.


Exclusion:  Becoming connected to a network can open doors, whereas exclusion from a network can perpetuate inequalities.  The most obvious case of such exclusion are "old boy" networks in which men, who have traditionally dominated a given line of work, inform only other men of hiring and promotion opportunities. Because networks are created through social ties, the presence of a tie can include people, but the absence of a social tie can exclude.  Looking at networks requires, therefore, looking at who is included and who is excluded through the network.  The clearest cases of exclusionary networks are social cliques and exclusive clubs such as country clubs. (Hachen, 2001: p. 31-32). 


You might reflect on how the size, strength, and diversity of social networks might affect the extent or nature of these four types of processes.


Overall, the goal of this exercise is for you to become aware of the existence and nature of social networks and to understand that they have important social consequences.