Social: Theories and Models - Foundations for Designing User-Centered Systems - page 255

about one another, but eBay itself will help teach reciprocity—if you violate their
rules they will remove you from the system. And, finally, it certainly cares about
and works towards improving the ability to recognize buyers and sellers, serving as
a type of recognition for sellers.
9.2.4 Networks
Another way to analyze social interactions is by using network theory (see, for
example, Rainie et al. 2012 ; Watts 2003 ). This approach emphasizes how the
actors are associated across multiple interactions, including how well connected
they are, and who they are connected to.
People are linked to other people in many different ways. They can have family
connections, or they may be work colleagues, for example. These relationships are
a type of social capital. They can be represented using networks, which can be
drawn graphically, showing people as nodes, and the relationships as links. An
example of a network is shown in Fig. 9.3 . Node 13 is the least well connected;
node 14 has the most connections; nodes 6, 7, and 14 have the strongest
connections. A more complex example is included as Sect. 9.2.5 . The way that
networks function can have some implications for the system that you are
designing.
Leadership in networks. It is often useful to be aware of which nodes in a
network are the most important, because all the nodes may not be equally
important. If you know which nodes link groups within the network, you can make
use of them to beneficial effect. This is what advertisers attempt to do when they
plan campaigns. If they target the nodes (people, locations, etc.) which link groups
together, this can lead to campaigns going viral, rather than just being of interest to
a local group in a network. A more complex example is included in the breakout
box explaining how connections influence work in an electronics firm.
The distance between two nodes can be calculated by identifying all the other
nodes you need to pass through to get from the start node to the end node, and then
adding up all the pairwise path lengths (between the adjacent nodes). If the two
nodes are important (to the operation of the network, or the organization, and so
on), then you may want to make sure there is more than one way to get from the
start node to the end node, so that you can cope with failures of nodes or links. It
turns out that the number of links that you need to get from one person to another
can be quite small. This so-called small world problem is the basis for the film The
Six Degrees of Kevin Bacon. Essentially, you know somebody who knows
somebody who… who knows somebody who knows Kevin Bacon.
The small world phenomena. Perhaps the most common idea is based on
connectivity between nodes, that is, what is the shortest path between two nodes?
How many people would a letter from one person to another have to pass through,
or how many introductions would you have to have to meet Brad Pitt? In a
company or even the world, what is the longest path between any two people?