A comparison of on and offline networks through the Facebook API

  • warning: Illegal string offset 'data' in /var/www/nir/skabber.nir.com.ua/includes/tablesort.inc on line 110.
  • warning: Illegal string offset 'data' in /var/www/nir/skabber.nir.com.ua/includes/tablesort.inc on line 110.

SOCIAL Network Sites (SNS) are a part of everyday
social activity for millions around the world (boyd &
Ellison, 2007). Sites vary in their affordances, audiences
and scope. While some sites cater to specific interest
group audiences and “niche” networks, others, such as
Facebook, LinkedIn and MySpace are designed to be
general purpose “social utilities”. Are these sites in fact
general purpose social utilities? If so, we should expect
to find not only an overlap between the pre-existing
personal network and the Facebook network, but also
a logic that can connect the two. At present, a host of
studies are emerging that take Facebook networks as
a stand-in for the ‘real’ social networks of individuals.
The study of community structure through Facebook
is meant to signify actual community structure (Traud,
Kelsic, Mucha, & Porter, 2008). The study of taste on
Facebook is meant to signify actual differences and
clusters in taste (Lewis, Kaufman, Gonzalez, Wimmer,
& Christakis, 2008). These studies, by virtue of their
research design, do not actually measure or examine
the personal network. Instead, they capture a slice of
Facebook’s database and assert that for the group in
question (generally college students), it is a reliable
proxy for the network of active ties. However, if we are
to continue down this road, it is important to connect
these networks to the pre-existing studies of personal
networks and networking. This enables researchers to
leverage past insights on personal networks (such as
theories of foci, closure, multiplexity and so forth, c.f.,
Fischer, 1982) and because it is an important validity
check on these online networks.
This paper is a research outline and a preliminary
quantitative autoanalysis of the relationship between the
personal network and one’s Facebook network. Insofar
as Facebook is the most popular social utility for
individuals in the United Kingdom and Canada, it is
an ideal candidate to examine how personal networks
manifest themselves online (or to the extent that they
do so). The results should not be taken as exclusively
generalizable to Facebook, but to any social utility that is
sufficiently diffused in a population to be considered as
the dominant platform for managing one’s online social
network. Accordingly, one might consider a similar analysis
of personal networks and Orkut networks in Brazil,
or personal networks and QQ networks in China. That
said, it is likely that there is only one or two dominant
social utilities in any given country that are sufficiently
diffused to be considered a viable candidate for such an
analysis.
Should this analysis be confined to SNSs or should it
be broadened to consider other media, such as email,
mobile phones and so forth? There are several reasons
to suggest that while it is possible to replicate this
analysis with a single medium, SNSs offer something
qualitatively different from specific media and also from
personal networks.
As an example of digital communication, SNS users
leave traces of their activities that can be reconstructed at
a later time. This allows for a high-fidelity representation
of activity. Indeed, this was the case for email, instant
messenger and other digital communication devices.
However, SNS are built around social relationships that
are a means to communication of various kinds, including
status updates, photo sharing, event participation,
message distribution and real-time chat. By contrast,
prior ICTS are communication media foremost from
which one can extract and illustrate social relationships
in a specific modality. In this regard, SNS are similar to
offline relationships in that individuals have a mental
model of who they are communicating with, apart from
any specific mode of communication. For this reason
boyd has termed SNS sites as “publicly articulated social
networks” (2004). The emphasis here is that the network
is articulated or specified by the user, rather than merely
constructed from past interaction.
Additionally, SNS are now broadly diffused through
many national populations. According to Alexa.com,
Facebook is the number 2 site in Britain behind google.
In Canada it is number 3 and it is number 5 in the U.S.,
France, and Italy. This broad diffusion means that the
pool of possible individuals for social relationships is not
a niche market of specialized interests, but approaches
the pool of possible relationships from everyday life.
This paper introduces a Facebook application that
leverages the Facebook API (Application Programming
Interface) to create a personal network from the data
available on Facebook. The basics of this technique is
not very complicated. However, beyond simply asking
for a list of friends and their mutual ties there are a
host of persnickety pitfalls to which the researcher must
attend. After a discussion of alternative data, history of
personal network methods and the Facebook API, I give
an overview of the present situation and expected results
of the application under development. In the penultiElectronic
copy available at: http://ssrn.com/abstract=1331029
mate section, I provide a cursory analysis using the
application on my own Facebook and personal networks.
I conclude by reiterating the specific challenges in this
approach as well as directions for future research.

AttachmentSize
SSRN-id1331029.pdf373.98 KB