This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
         If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line, 
send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").

(Jan 2006)
Thank you.
Elliott Parker
====================================================================

Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Introduction
The second annual State of the Media report1 produced by the Project for
Excellence in Journalism identified as one of the big trends in 2005 
the changing model
of journalism which is now "faster, looser, and cheaper." Cautioning 
against the current
wave of citizen journalism, the authors position blogs as part and 
parcel of "unchecked
media" which they feared "expands this culture of assertion 
exponentially." Side by side
with this view of the blogosphere as threatening the foundation of 
responsible ethical
journalism is the embrace of blogs within traditional mainstream media. Though
mainstream media adopted an initial suspicious view of blogs 
evidenced in one case with
the firing of Stephen Olafson of the Houston Chronicle (Olafson, 
2003), ABC.com's The
Note2 and National Review Online's The Corner3 provide evidence of a 
new embrace of
the bloggers within the newsroom. Television has been the latest to 
incorporate blogs
within their news programs. MSNBC has begun a show, "Connected: Coast 
to Coast"4
which tracks blog stories through the news program. CNN's Judy 
Woodruff's program
Inside Politics5 now includes a two minute segment called, "Inside 
the Blogs.". Flexing
their political muscle as fact checkers on mainstream media, blogs 
have played an active
role picking up where the media left off through being instrumental 
in the fall of Trent
Lott, Dan Rather, Eason Jordon, and James Guckert alias Jeff Gannon. 
Most recently,
the White House admitted the first blogger to a news briefing.6
Recent 2005 reports continue to show an ever expanding blogosphere. Founder
and CEO of Technorati, David Sifry's March edition of the State of 
the Blogosphere's
report (2005) reports Technorati's tracking of 7.8 million Weblogs 
and 937 million links,
with the blogosphere doubling in size every 5 months. According to 
one of the most
recent Pew Reports (2005), blog readership shot up 58 percent at the 
end of 2004, an
increase of 17 percent since February 2004. Much of this growth in 
readership can be
attributed to the active role that blogs played in the 2004 election. 
Howard Dean, then
1
2
3
4
5
6
State of the Media report available at 
http://www.stateofthemedia.org/2005/index.asp
ABC's The Note available at 
http://abcnews.go.com/Politics/TheNote/story?id=156238
National Review's Weblog available at 
http://www.nationalreview.com/thecorner/corner.asp
Connected: Coast to Coast. Program description available at 
http://www.msnbc.msn.com/id/6862172/
Inside Politics with Judy Woodruff available at 
http://www.cnn.com/CNN/Programs/inside.politics/
Garett M Graff writes for the Weblog Fishbowl. DC. Details of his 
credentialing can be found at
http://www1.cnn.com/2005/ALLPOLITICS/03/07/w.h.blogger.ap/
1
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
governor of Vermont, inspired the use of candidate blogs7 in March 2003 as an
unmediated citizen communication channel on his candidate Web site. 
The ripple effect
of his usage of the blog was evident in its quick pick up by other 
candidates such as John
Edwards, Wesley Clarke, John Kerry, and George Bush.
A watershed moment for political blogs in 2004 was its usage in the 2004
Democratic National Convention (DNC) and Republican National Convention (DNC).
Defined by Journalism Professor and Blogger Jay Rosen blogger as a "crossover
moment" (Glaser, 2004) for both non media and media bloggers, the Democratic
National Convention issued press credentials to 35 non media bloggers 
in July 2004 to
share media space with mainstream media journalists, who were also 
event blogging the
convention. Technorati partnered with CNN to deliver feeds of blogger 
stories, and
special RSS gatekeeper aggregate feeds from Feedster and 
ConventionBloggers.com
were developed to help citizens and fellow bloggers tame the 
information tide of blog
positings. The RNC followed suit in August 2004 credentialing 22 non 
media bloggers to
cover the August 2004. For both conventions, mainstream media such as CNN and
MSNBC developed specialized group Weblogs and Knight Ridder permitted 
its bloggers
to host their Weblogs on Blogspot. Some non media bloggers also 
crossed over into
mainstream media space for the DNC: Jay Rosen of pressthink blogged for Knight
Ridder, David Weinberger of hypermedia.org worked for Boston.com, and 
Marie Cox
a.k.a. Wonkette was picked up by MTV to blog the event.
This paper assesses of the contribution of the bloggers through examining the
nature of coverage by both the non media and the media event bloggers 
in terms of link
analysis. Harvesting a total of 5,225 links on 104 Weblogs, this 
study seeks to see how
both media and non media bloggers used the blog form to spread 
influence and to form
connections among various networks. Though non media and media 
bloggers were forced
to share media space symbolically and physically, this paper examines 
whether there
were truly any crossover moments in terms of linked 
interrelationships among the
different four different networks of bloggers. Using network theory, 
homophily/echo
chamber concepts and gatekeeping theory to predict certain states 
within the networks,
this paper probes the nature of these bloggers networks through an 
assessment of blogger
7 Howard Dean's candidate Weblog is held at www.democracyforamerica.com
2
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
linking practices both among each other and outside each other to the 
mainstream media .
More importantly, this paper uses link analysis to get at the nature 
of the relationships
among the blogger networks in an effort to see how this shared media 
space was used and
how it was negotiated both between mainstream media bloggers and non 
media bloggers
as well as between the blogger networks of the democratic and the republican
conventions.
Theoretical Review
A useful theory that can be applied to the study of blog communities 
is network
theory. Network theory has been useful in the study of such 
real-world networks as social
networks, information networks, technological networks, and 
biological networks. The
World Wide Web has been one area within information networks in which network
theory has been readily applied to explain the nature of connectivity 
in cyberspace.
According to network theory, a network is comprised of vertices or 
nodes, with the
connections between them referred to as edges (Newman, 2003). In the 
world of the
Web, Web pages can be viewed as nodes or vertices, while the 
hyperlinks between Web
pages can be seen as edges. Edges can run in one direction, termed 
arcs or unidirectional
edges, while undirected edges are bidirectional connecting two 
vertices. Degree or
connectivity relates to the quantity of edges that connect to a 
vertex—Web pages with
many in-links or edges are considered vertices with high degrees or 
strong connectivity
and both vertices and edges, or Web pages and links, have variant 
characteristics. It is
these characteristics that determine the shape and character of the 
World Wide Web
network topology.
Network theory has its foundation in mathematical graph theory and discrete
mathematics with Euler's 1735 solution to the Konigsberg bridge 
problem as the first true
proof in the theory of networks. Most of the current theory 
underlying real world
networks gained footing in the random graph models and theory of 
Erdos and Renyi
(1960). Commonly called ER networks, random networks are exponential, mostly
homogeneous networks, with each node or vertex having approximately 
the same number
of links, contributing equally to the network's structure. There is 
equal probability that
each edge is present or absent, and the degree distribution is 
binomial or Poisson. Local
3
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
connectivity in random networks is free of scale and vertices of high 
connectivity are
generally absent. As Babarasi and Albert (1999) point out, random 
networks assume that
there are a fixed number of vertices that are randomly connected without any
modification. Also, the probability that two vertices are connected 
is also random and
uniform. Both of these characteristics are absent in real world 
networks where the
network is constantly expanding and where there is an evidenced 
preference and choice
in the way vertices connect with each other.
Networks such as the World Wide Web are commonly called scale free networks.
Scale free networks commonly gain distinction from random networks 
due to two factors:
growth and preferential attachment. Barabasi and Albert (1999) 
highlighted that real
world networks like the World Wide Web are not fixed but are open and 
grow through
the continuous addition of vertices that connect to the vertices 
already present in the
system. Also, newly added vertices tend to connect to those vertices 
that already display a
high degree or connectivity. The result of this growth and 
preferential attachment is a
power law distribution where the probability P(k) that a vertex in 
the network is
connected to k other vertices decays as a power law following P(k) ~ 
k-y (Barabasi,
Albert and Jeong, 2000). This leads to a power law distribution with 
a heavy tail which
drops off more slowly than the tail of a Gaussian distribution.
Applying this theory to the World Wide Web, a small number of Web pages have
a great majority of the inbound links while the majority of the sites 
suffer from low
connectivity. Networks can be said to display clustering or 
transitivity, which deviates
from the random graph theory, which states that if A is connected to 
B, and B to C, then
there is a heightened probability that A connects to C. It shows the 
probability that two
vertices that are network neighbors of the same other vertex will 
themselves be
neighbors. The probability that two vertices in a directed network 
point to each other is
called reciprocity. Related to this is the study of which vertices 
decide to pair up in a
network. In social networks, this kind of selective linking is termed 
homophily or
assortative mixing, with classic examples being mixing by race, or by 
other scalar
characteristics such as age or income. There is also mixing according 
to vector
characteristics such as geographic location. Degree correlations is a 
special case of
4
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
assortative mixing according to a scalar vertex property called 
vertex degree. Do high
degree vertices in a network associate preferentially with other high 
degree vertices?
The homophily principle posits that people are attracted to 
similarity instead of
difference—a concept which has ramifications in the discussion of the 
structure of
networks on the World Wide Web. The first homophily study has been traced to
Lazersfeld and Merton's (1954) study on friendship processes in 
Hilltown and Craftown.
In this study, the authors identified two types of homophily: status 
homophily due to
demographic factors and value homophily resulting from shared beliefs 
and values.
Homophily theory, also known as the "birds of a feather flock 
together" principle, cites
geographical, organizational, and role foci as powerful structural 
forces that induce
homophily (McPhearson, et al, 2001).
Viewing the Web as a network of Web pages connected by links, a concept that
has gained traction in discussing social networks on the Web is the 
echo chamber effect.
Related to groupthink philosophy, Sustein (2000, 2002) cited that the 
Internet could be
harmful to democracy because of the role it can play in fragmenting, 
insulating, and
encouraging enclave deliberation through filtering both what is 
produced and what is
read. Resulting in daily editions of the "Daily Me" (Sustein, 2001), 
Sustein highlighted
the most extreme result of cyberbaulkanization—group polarization and 
the movement to
more extreme positions than that held at predeliberation due to the 
effects of limited
argument pools and perspectives, and because of the powerful hold of 
social influence.
This fracture of the public sphere into several "counterpublics or 
enclaves" is viewed as a
threat to democracy because of negative informational cascades that 
can result when the
group remains insulated and homogenous in both perspective and 
composition. Balkin
(2004)'s espoused that the blog's promotion of cybalkanization and 
group polarization is
overstated and hyped. Blogs gain their foundation on criticism, 
making it difficult to cite
things that are not agreed with. According to Balkin, the Internet 
allows a wider
distribution of ideas as opposed to the unidirectional mode of 
delivery of mainstream
media, resulting in a continuous bringing together of people who disagree.
Gatekeeping is another theory that can be readily applied to the 
study of networks
that transmit news. The term gatekeeper was coined from the 
sociologist Lewin (1947) in
his study of group dynamics, but since then gatekeeping has been used 
to explain how
5
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
news travels through news organizations to news audiences through the 
news selection
process. Shoemaker et al, (2001) defined gatekeeping as "the process 
by which the vast
array of potential news messages are winnowed, shaped, and prodded 
into these few that
are actually transmitted by the news media." The classic study of 
gatekeeping was
conducted by White (1950) who found that a wire editor, Mr. Gates, 
seemingly made
subjective decisions in choosing the news for the day based on his 
value judgments,
personal experiences and attitudes, and expectations. The theory of 
gatekeeping was
expanded by Shoemaker and Reese (1996) who cited five different 
factors that affect
news selection: ideological factors, extramedia factors, 
organizational factors, media
professional routines, and individuality. Most of these factors speak 
to the traditional
notion of gatekeeping as performed within the hierarchial, vertical 
structure of the
newsroom. With the advent of online journalism, Weblogs, citizen 
journalism sites, and
other independent news services, the idea that the consumer could 
play an active role in
both the production and selection of news is enabled by new Internet 
technologies.
Trench and Quinn (2003) highlighted the defining practice of online news as
horizontal communication versus the vertical communication from 
authoritative sources.
Houston (2000) pointed out that the choices available to news 
consumer are now endless
and diverse with the burgeoning of news outlets on the Internet. With 
the advent of the
World Wide Web, more journalists are using the Web as a source for 
materials to create
the news. According to Singer (1998), in the new environment of the 
World Wide Web,
gatekeeping theory has become redefined as journalists refashion the 
meaning of their
craft towards more sense making and quality control. Of specific 
interest in this paper is
the intersection of gatekeeping and network theory through the device 
of hyperlinking.
Trench and Quinn (2003) identified hyperlinking as the mode through 
which the audience
is allowed to work through different routes through the story 
according to previous
knowledge. Examining the perspective of the news producer, Dimitrova 
et al, (2003)
identified the act of gatekeeping in the very function of selecting 
which hyperlinks to
include in Web stories. Taking this idea a step further, the choice 
of hyperlinks in the
Web environment can also signal which sources are trusted, valued, 
and included in the
network that creates the online news. In the words of Dimitrova et 
al, "Gatekeeping may
be as noticeable in its absence as in its presence."
6
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Literature Review
Networks and the World Wide Web
Examining some of the literature in social network theory is essential to an
understanding of how the theory applies to real world analysis of Web 
networks. Price
(1965) was one of the first to describe the power law distribution 
underlying scale free
networks through examining citations between scientific papers. 
Several studies have
shown that patterns of link structure within networks on the World 
Wide Web network
exhibit a power law distribution. Huberman et al, (1998) found that 
the amount of Web
pages surfed among 107 staff and faculty at the National Center for 
Supercomputing was
inversely proportional to the Web site's depth. Nielsen (1997) found 
similar trends in Sun
Microsystem's Web pages. Power law trends have also been found in the 
growth of Web
sites (Huberman and Adamic, 1999; Huberman and Adamic, 2001). Web 
site traffic has
also been found to exhibit power law characteristics. Examining 
market share on the Web
through analysis of the AOL usage logs of 120,000 Web sites, the 
authors found that the
distribution of visitors to the sites followed a power law 
distribution. Rating popularity as
the number of unique visitors to a site per month, the authors found 
that the top 119 sites
captured 32.36 of the user volume. These findings were replicated for 
adult sites, with the
top 10 percent capturing 60 percent of the traffic, while with .edu 
sites, the top 5 percent
accounted for 60 percent of the Web traffic. Examining Web blog 
traffic, Kottke (2003)
found power law trends in the examining the links to the top 100 
Weblogs as derived
from Technorati data in terms of inbound links. Shirky (2003) found 
power laws in
LiveJournal listings regarding number of friends listed, with a few 
users having the
majority of the friends.
The majority of past studies have examined Web power law relationships through
conducting Web crawls and using link data as the form of analysis. 
Barabasi, Albert, and
Jeong (2000) found power laws for in-links and out-links on Web 
documents based on
constructing a robot to track URLs from the Web. Broder et al, (2000) 
confirmed a power
law for in and out degree distributions after a Web crawl of 200 
million pages and 1.5
billion links. Link data has been mined to examine the strength of 
the social connections
and bonds between different communities primarily through the 
methodology of Web
crawls. Many of these studies examine the concept of clustering and 
transitivity among
7
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
nodes that have preferential similarity. Through the analysis of the 
co-occurrence of text,
and links within the personal homepages of Stanford and MIT 
communities, friendship
was accurately predicted within these communities by similarity 
(Adamic and Adar,
2003). Deriving a sense of the social structure of a university 
community, the authors
found evidence of the power law in that only a few pages are linked 
to the most, while for
most authors there was a limited number of in-coming links. Eckmann 
and Moses (2002)
used solely hyperlink data to glean hidden thematic Web relationships 
by using reciprocal
hyperlinking as evidence of mutual relationships. Kumar, Raghavan, 
Rajagopalan, and
Tomkins (1999) used a process dubbed "trawling" to crawl the Web for 
evidence of
interlinked Web pages in order to identify over 100,000 emerging 
communities. The
authors mined the Web through link analysis of over 200 million Web 
pages archived
from Alexa in order to derive hidden or emerging community structure 
patterns. The
authors used cocitation, a notion derived from bibliometrics which 
espouses the idea that
pages that are related are frequently referenced together. Applying 
18 month old crawl
data to the then present Web, the authors found that 70% of the 
communities were still
alive.
Some studies have sought to explicate more closely how the notion of
preferential attachment relates to scale free networks such as the 
World Wide Web.
Dorogovtev et al., (2000) hypothesized that before the operation of 
preferential
attachment, initial attractiveness governs the probability that 
"young" sites attract new
links. After that initial attractiveness and in keeping with the 
concept of preferential
attachment, the number of incoming links or the Web site's 
connectivity becomes the
prime determinant influencing whether a new link points to a given 
site. Questioning the
notion of the "winners take all" concept of preferential attachment, 
Pennock (2002) found
that the degree of preferential attachment on Web sites was dependent 
on categoryspecific
Web pages. These authors found unimodal versus power law relationships such
that losing and mediocre sites attracted a high proportion of links 
among Web sites of the
same type. Company homepages and newspaper pages were dominated by 
preferential
attachment while scientist and university homepages showed a more 
balanced mixture of
preferential and uniform attachment.
8
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
An interesting idea advanced by Klienberg (1997) was the discovery of
authoritative sources and its relationship to hub pages. According to 
Klienberg, the
strongest authorities tend not to link to each other because of 
competition. Often, a
network of interrelated pages must exist to point or link to these 
various authorities,
termed "hub pages." For a page to qualify as an authoritative page, 
there must be
considerable overlap in the sets of pages that point to it. Hub pages 
exist to pull together
the authorities on a common subject such that a good hub must point 
to good authorities,
and a good authority is pointed to from many hubs. Much of the power 
dynamics among
Web pages can also be determined by the linkage patterns.
Applying network theory to journalism is a new emerging area. Tremayne (2004)
found that linking practices within online news organizations is 
predicted by network
theory. The author found that the number of links in news stories 
increased in a
longitudinal assessment of digital archives in Web stories. 
Preferential attachment
predicted which stories contained more links, as well as the increase 
in the gap between
sites over time. Using links as a sign of community relationships, 
this paper sought to use
network theory as a basis to help explain some of the linking 
practices adopted by
bloggers reporting on the 2004 conventions in both mainstream and non 
media networks.
Blogs, Bloggers, and Mainstream Media
Work on cybalkanization as it relates to the differences between 
conservative and
liberal bloggers in terms of their network affiliations is now 
emerging in academic
discourse. Examining a single day of linking patterns between the 
left-leaning blogger
Atrios and the right leaning Instapundit, Welsch (2005) found that 
Instapundit cited more
URLs, with no overlap when compared with the URLs mentioned by 
Atrios. Adamic and
Glance (2005) examined the relationship between the 2004 presidential 
election and the
political blogosphere by taking a single day's snapshot of over 1000 
blogs, as well as
charting 40 influential blogs for two months preceding the November 
2004 presidential
election. Examining both the blogroll links and page citation links, 
the authors found that
91 percent of links originating within either the conservative or the 
liberal communities
stayed within their communities. Conservative blogs had a higher 
tendency to link with
an average of 15.1 links versus 13.6 links from liberal blogs and the 
top 20 conservative
9
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
bloggers fell within the top 44 most cited in comparison to the top 
20 liberal bloggers
who fell in the top 77. Conservative blogs not only had more links 
than liberal blogs but
tended to have stronger connections when bidirectionality or 
reciprocation was factored
into the analysis.
Few studies have as yet examined the actual relationships forged between the
blogosphere and mainstream media. Drezner and Farrel (2004) posited 
that mainstream
media, or the mediasphere, uses the blogosphere for interpretative 
frames; additionally,
blogs are ideally positioned to support media because many blogs have 
preexisting social
ties with some of the earliest bloggers. Bloggers have first-mover 
advantage because of
the low cost of the real-time updates to the Web, and many bloggers 
possess expertise
that supersedes the generalist knowledge of a journalist, resulting 
in dependence on the
blogosphere for material. The authors conducted an online survey 
resulting in responses
from 140 editors, reporters, columnists, and publishers. The 
respondents spanned the
gamut of the New York Times to rural publications with less than 
100,000 readers. Asked
to name the three blogs they frequent, the authors found that 
collectively over 125 blogs
were cited, with the top 10 blogs accounting for 54 percent of all 
the citations. Among
elite media respondents, the skewness was more pronounced with the 
top ten blogs
responsible for over 74% of the citations, the top five being 
responsible for 56%.
Sullivan, Kaus, and Marshall were among the top five of all bloggers.
A study conducted by Singer (2004) examined the efforts of mainstream
journalist bloggers, the "j-bloggers" during the Super Tuesday time 
frame of the 2004
presidential election. Examining ten national and ten regional or 
local blogs by journalists
for one constructed week between February 15 and March 15, Singer 
analyzed 1,136
political blog postings. The author found that inclusion of personal 
opinion varied: 61%
of the political j-blogs contained expressions of personal opinion at 
least occasionally,
with blogs provided by columnists more likely to contain opinions 
than reporters.
Examining gatekeeping in this new environment, Singer found that the 
j-bloggers
restricted user involvement, and overwhelmingly the mainstream media 
tended to link to
either other mainstream media sites or themselves. Of the 2,614 links 
contained in the
1,136 blog items, 2,015 links are to mainstream media sites with 360 
links to the
blogger's parent medium. Singer concluded that journalists are 
"normalizing the blog
10
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
format" in maintaining control over the information they disseminate; 
through there is a
seeming commitment to expanding transparency through hyperlinking, 
accountability is
second hand through the primary linking to mass mediated political views.
Using citation as a measure of authority in an attempt to understand 
the social
implications of linking, Marlow (2004), creator of Blogdex, examined 
referential
information for a 3-year period through analysis of hyperlinks. 
Authority was comprised
of two measures: popularity as measured by the public affiliations of 
the Weblogs
(blogroll links) and influence as measured by page citation links 
(permalinks). Marlow
found the fall off of authority in terms of the power law 
relationship to be steeper in
blogroll links than permalinks. Examining 310 known URLs from 4,728 
articles derived
through querying the Lexus Nexis database, the author found that most 
of the top cited
were among the higher ranking sites for both blogroll and permalinks. 
Journalists are
prone to read those weblogs that have greater authority as determined 
by popularity and
influence.
Gatekeeping and Participatory Journalism
A few studies have examined the changing role of mainstream media gatekeeping
within the new media environment. Deuze (2002) cautioned of the 
tendency of news sites
to avoid external linking in favor of exclusive on-site linking. 
These statements are
empirically supported in a study by Dimitrova et al, (2003) of 473 
stories from15 Big
Media newspapers. The authors noted that there was a decline in the 
use of external
hyperlinks, suggesting that hyperlinks were used as a gatekeeping 
device to prevent users
from leaving the Big Media Web sites. This study supported previous 
findings (Peng et
al, 1999; Kovarick, 2002; Tremayne, 2004) that news Web sites are 
shirking away from
openly exploiting the open infrastructure of the Web. In Dimitrova's 
et al (2003) study,
94.8 percent of stories took readers to related material within the 
newspaper's own Web
site. Over 80 percent of the hyperlinks were repeated or duplicate 
links. The authors
conclude that news organizations may be unwilling to diminish their 
gatekeeping strength
by preventing outside links. Will journalist exercise their 
gatekeeping powers in their
blogging of the 2004 conventions? This study seeks to find out the 
nature of the media
11
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
blogger's hyperlinking practices when compared to that of the non 
media political
blogger.
Research Questions
In an effort to apply many of the ideas of network theory and 
gatekeeping to the
examination of the 2004 blog networks of the democratic and 
republican national
conventions, this study put forward a series of hypotheses and 
research questions
designed to get at the nature of the linking patterns as well as the crossover
interrelationships among the media bloggers, and the credentialed 
non-media bloggers
covering the 2004 conventions. Essentially, this study examined four 
sets of networks:
1. credentialed DNC non media bloggers
2. DNC media bloggers working for official news sites
3. credentialed RNC non media bloggers, and
4. RNC media bloggers working for official news sites
As has been previously mentioned, the networks were not distinct: Jay 
Rosen and David
Weinberger were examples of two non-media bloggers that were hired to work as
journalist bloggers for media companies. Thus, this study presented 
an ideal opportunity
to question whether the networks of media and non media bloggers 
intersected, and in
what ways, as mainstream media openly co-opted the blog format through sending
bloggers to cover the both conventions.
As many past studies have shown, power law relationships exist such that a few
nodes in the network attract more edges than the majority of the 
network nodes. Within
this blog network of credentialed non media and journalist bloggers, 
this study
hypothesized that power law relationships would exist in the linking 
patterns for the
entire network and for all four networks of bloggers such that a few 
sites would
command most of the links as opposed to the majority of the network:
Hypotheses 1: All four networks will show classic power law 
relationships in their
external linking patterns such that only a few sites would be more 
widely cited in links
when compared to the majority of the sites cited by bloggers.
12
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
As network theory predicts, the power law relationships arise as a 
result of both growth
and preferential attachment in a network, which causes some nodes to 
attract more edges
than others. Examining the latter construct, this study hypothesized 
that network
familiarity, social position, and prior bonds due to status and 
shared ideology in the
network all leverage powerful forces in determining who non media 
bloggers prefer to
link to. As a result, this study made the following hypothesis and 
posed the following
research question:
Hypothesis 2: Non media bloggers would show a preference to link more 
to other non
media bloggers than media bloggers.
Related to hypothesis 2 would be to ask the question of who do media 
bloggers point to
in their links. Network theory suggests preferential attachment as 
one of the main reasons
for high node connectivity in the network, but who would media 
bloggers consider as the
preferred node? As Klienberg (1997) found, authority sources often 
avoid linking to other
authorities because of the competition. Authority sources are often 
determined by a high
degree of in-linking. In terms of connectivity, high degree vertices 
prefer to connect to
other high degree vertices. However, it is questionable whether media 
bloggers would
view non media bloggers as the preferred alliance to forge. This 
paper questioned the
nature of the media's bloggers connection to other non media bloggers 
within the two
conventions:
Research Question 1: To which network of bloggers would the media bloggers
demonstrate preferential attachment towards: fellow media bloggers or 
non media
bloggers?
Past studies have shown that the media exercises gatekeeping in the 
online environment
through limiting the external linking and increasing internal linking 
within their site as a
form of preferential attachment. This study proported that, even 
though journalists
embrace the blog format, old practices will continue to pervade the 
blog format such that
13
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
media bloggers will show a greater propensity to externally link less 
than non media
bloggers.
Hypothesis 3: Media bloggers would have less external links than non 
media bloggers
covering the convention.
A major part of this study resided in the harvesting and tagging of 
URLs to form greater
impressions of the connections developed among media and non media bloggers. A
related idea of network theory is that vector characteristics can 
also determine the
preferential attachment of nodes within a network. Not only can the 
composition of the
network determine attachment, but such other factors as network 
neighborhood and
geographical proximity to other similar nodes. Since the network of 
media bloggers was
so small and defined, one of the study's research questions looked to 
the non media
blogger network, questioning whether the selection and credentialing 
of non media
bloggers would act as a powerful signal of network neighborhood such 
that both non
media and media bloggers would be encouraged to link more to 
credentialed bloggers:
Research Question 2: To what extent did the media credential define 
the parameters of
the network boundaries and create preferential attachment such that 
both media and non
media bloggers were more prone to link to non media bloggers 
credentialed to cover the
conventions as opposed to those not credentialed?
Many question whether World Wide Web networks create greater fragmentation,
cybalkanization, or cocooning such that there is an echo chamber 
effect, particularly in
networks comprised of similar ideologies or partisan inclinations. 
Examination of the
networks of non media bloggers for both conventions showed a heavy 
weighting to the
favorable partisan ideology in the choices to credential certain non 
media bloggers. For
both conventions, the major percentage of credentialed non media 
bloggers covering the
convention could be easily identified as sharing the convention's 
partisan perspectives
through analysis of the blogger's blogroll links. Harvesting page 
citation links as opposed
to blogroll links to provide some analysis of the network 
composition, crossover, and
14
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
diversity of opinion sharing among the bloggers in the blog 
ecosystem, this study also
proposed the following research questions in an attempt to examine 
the diversity of
choice that was presented in the networks in terms of media choice 
and voices within the
network blogosphere:
Research Question 3: Was there a difference in the type of links that 
were the most
popular among the different networks of bloggers?
Research Question 4: Was there a difference in the quantity and type 
of traditional
media linked to by both media bloggers and non media bloggers in the different
conventions?
Research Question 5: To what extent was there a reflection of diverse 
opinion selection
through linking to bloggers with heterogeneous ideologies in the 
different conventions
among bloggers?
Methodology
To examine the three hypothesis and four research questions, this 
study attempted
to examine all of the blogs that were credentialed to cover the 
democratic and republican
convention. Cyberjournalist.net recorded 22 credentialed non media 
individuals and 28
journalist blogging for news sites for the Republican conventions and 
37 independent
credentialed individuals and 30 journalists blogging for news sites 
covering the
Democratic conventions. Because this study examined the data some 
months after the
event, some data was lost or blocked behind media companies' 
archives. In sum total, 34
independent credentialed individuals were examined for the DNC non 
media bloggers
and 25 for the media bloggers. For the Republican conventions, a 
total of 21 credentialed
non media bloggers and 24 sites for the media bloggers were examined. 
A full listing of
these blogs can be found in Appendix A.
In order to extract the URLs for analysis, blog posting DNC 
convention coverage
for the period of July 25 to July 30 period and blog posting RNC 
convention coverage for
the period August 30 to September 4 period were downloaded to Word 
files for each of
15
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
the four network of bloggers. Differing from past studies which 
depended on both
blogroll and page citation links (Adamic and Glance, 2005), this 
paper confined itself to
an analysis of page citation links, or links embedded within posts, 
as opposed to blogroll
links, or links held on the left and right columns of blog pages. 
Blogroll links are more
static and represent affiliations that do not change as opposed to 
page citation links which
capture the changing dynamics of a community and influence. 
Downloaded files were
also cleaned by removing all self referencing links to the said blog 
posting from
permalinks, comments, and hyperlinked headlines. Removing these links 
stripped the
pages of excessive internal linking or recursive pointing that could 
have skewed the
results.
To extract URLs from the blog postings, cleaned files were run 
through a desktop
software tool called the URL Extractor program. This automated 
extraction yielded a
total of 5,225 link mentions from all four networks of bloggers. To 
prepare the URLs for
coding, URLs were reduced to their main domain addresses to afford 
easy comparisons
and summary statistics. For example, if specific documents were 
pointed to at CNN, the
URL became cnn.com. Links to media bloggers became, for example,
weblog.herald.com. Reducing the URLs to their main domain meant that 
URLs were not
examined for specificity as much as for broader relationships and higher level
connections. Care was taken to ensure that enough identifying 
information was left in the
domain of the URL to ensure proper tagging, particularly since many 
media blogs were
held behind the main parent domain. To allow analysis of the URLs, a 
computer program
was written to remove trailing slashes as well as the http://www 
before all of the URLs.
Many sites hold several URLs as pointers to their sites, such that 
http://dailykos.com and
http://www.dailykos.com point to the same place. Removing identifying 
information
(http://www:) before the URL allowed all versions of the URL that 
differed by the www
prefix to be automatically resolved.
Doing an inter-coder reliability sample of approximately 522 URLs or 
10 percent
of the codes through the entire codebook resulted in an intercoder 
reliability of 92 percent
using Holsti's coefficient of reliability formula. As opposed to hand 
coding through paper
sheets or Excel, a customized Web-based application was built to 
facilitate easy coding of
the 5,225 URLs, permitting them all to be entered through a front-end 
application into a
16
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
backend relational MS SQL database. Several variables were assessed 
in relation to each
URL. Each URL was tagged for its network affiliation as cited by 
either a non media or
media blogger covering either the democratic or the republican 
convention. URLs were
also tied to specific bloggers to enable identification of who cited 
the URL, as well as tag
whether the URL was external or internal to the person's blog site. 
Each URL was then
manually coded as a link fitting one of 23 distinct categories: 
television, newspapers,
radio, newsmagazines, blended media (mix of television, radio, 
newspaper) media parent,
web-based news, news portal, search engine, media blogger, nonmedia blogger,
government, internal link, candidate/politician/campaign Web site,
activism/grassroots/partisan, food/recreation/tourism/entertainment, 
other .com, other
.edu., other non .com, encyclopedia/dictionaries, .orgs, software 
sites, and media
watchdog/polling/media analysis sites.
To analyze the URLs, the researcher used two methods. The actual URLs were
primarily analyzed directly out of the database by writing structured 
query language
(SQL) queries. As a flexible database programming language, SQL 
permitted analysis
tailored to assessing the actual URL strings outside set categories 
as opposed to only
numeric precoded categories as handled by SPSS. For example, URL data 
was examined
through the database by writing queries on the popularity of URLs 
through domain
querying (cnn.com, msnbc.msn.com), as well as allowing outputs of 
each blogger's links
with frequency counts and sum. Without a database, this would not 
have been possible
Data was also exported as a CSV text file to SPSS to permit frequency 
and significance
analysis within the numeric codes assigned to each URL as determined 
by the codebook.
Results
Table 1 (Appendix A) presents an overview of all of the 104 blogs analyzed
during the entire 5-day period of the DNC and RNC conventions with 
their link totals
through the period under investigation. The table presents the blogs 
within their networks
as defined by their status as either a media or a non media blogger 
covering either of the
two conventions.
To test the first hypothesis, a broad summary of the top 20 links across the
network is presented in Table 2 (Appendix A). Because there were 100s 
of links,
17
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
representations of the power law relationships are confined to the 
first 100 sites and are
presented in Figures 1 through 5 (Appendix B). From the tables, it 
can readily be seen
how dramatically sites lose connectivity: by the 20th site in each 
network, particularly
with the media networks, connectivity markedly decreased for sites 
moving away from
the prized top positions. As the figures illustrate, both the summary 
of all four networks
and the summary of links through each network display power law 
relationships such that
only a few links in the network command the most attention while the 
majority of the
links are poorly connected. Strong support is provided for hypothesis 
one when looking
at both the most linked and the least connected. For the summary of 
links across the
network, out of 4,588 external links, the top 10 percent of sites 
account for 52% or 2,364
of the total links. Over 80% of the network receives 2 links or less 
with 66% of the
network only receiving one link. The picture is mirrored in the four 
networks that
comprise this big network. The top 10% account for 50% of the links 
of the DNC media
bloggers, 55% of the links among the DNC non media bloggers, 44% of 
the links of the
RNC media bloggers, and 60% of the links for the RNC non media 
bloggers. Similarly,
sites that receive two links or less are much larger: 85% of the 
links mentioned by DNC
media bloggers have two links or less, 81% for the DNC non media 
bloggers, 88% for the
RNC media bloggers, and 80% for the RNC non media bloggers. Sites 
that receive one
link comprise 75% of the links of the RNC media bloggers and 72% for 
those covering
the DNC, higher than the both non media bloggers networks which hover at
approximately 60% of the links mentioned.
These results show that power law relationships existed across the 
entire network
and across each of the four blogger networks for both conventions. 
Essentially, each
network provides another replication and test of the hypothesis that 
there are a few sites
within the network that receive the majority of the links, with the 
majority of sites
receiving few links or being poorly connected. There are some 
differences across the
networks in terms of the appearance of the power law relationships: 
both media blogger
networks have longer tails, showing that they tend to link to the 
majority of sites less
intensively when compared to the non media blogger networks. For the 
media blogger
networks, more sites within their network receive less links. Another 
noteworthy
difference in the power law relationships is in the network of the 
RNC non media
18
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
bloggers. Within this network, the top 10 percent of sites receive 
the most traffic at 60%
such that within this network of bloggers, the top few sites were 
linked to much more
intensively than the rest of the three networks.
Data from this first table provides some answers to research question 
three, which
asks about the differences in the types of links that gain top 
prominence across the four
networks. As can be seen in all four networks, elite traditional 
media still gain top
prominence across all of the networks. However, there are some clear 
differences in the
linking styles of media and non media bloggers. Both media blog 
networks cite more
traditional media in their links. Additionally, their preference for 
more formal sources of
information is evidenced in the presence of government sources, the 
formal party
campaign Web sites, as well as the presidential candidate Web sites. 
The high value of
entertainment in the news is seen in the presence of the Internet 
Movie Database Link
(imdb.com) in both media networks. In terms of blog mentions, three 
of the networks
share a characteristic of allowing blog voices entrance into their 
top links list. Both non
media blogger networks have a high presence of blog voices: 6 blogs 
for DNC non media
bloggers and 7 for RNC non media bloggers. Interestingly, blogs found 
it difficult to get
attention among RNC media bloggers: only three blogs were mentioned 
in the top 20.
Amazingly, only one media blog is mentioned in the four networks, 
reason magazines'
blog cited by the RNC media bloggers.
Hypotheses two states that due to preferential attachment, non media bloggers
would chose to link more to other non media bloggers as opposed to 
media bloggers. A
general summary of links allocation throughout the different 
categories in both time
periods is presented in Table 3 (Appendix A). The results highlight 
that traditional media,
aggregated to include newspapers, television, radio, and 
newsmagazines, account for the
greatest percentage of links in the networks culling 28 percent. The 
second highest
category of links is links to non media bloggers, which accounts for 
over 24 percent of
links. Internal linking is the next highest category claiming over 12 
percent of the links.
 From this table, it is apparent that almost no links were allocated 
to linking to media
bloggers with approximately 2 percent of links pointed to them.
This finding is solidified in Table 4 (Appendix A), which presents 
the results of a
cross tabulation between the different networks and the categories of 
links. As the table's
19
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
results highlight, non media bloggers are significantly more likely 
to link to other non
media bloggers as opposed to media bloggers. According to table 4, 
approximately 26
percent of DNC non media bloggers linked to other non media bloggers 
as compared
with less than 3 percent links to other media bloggers. Similarly, 
over 30 percent of DNC
non media bloggers linked to other non media bloggers, while linking 
to media bloggers
in slightly over 2 percent of their links. The findings are magnified 
when the links to non
media bloggers are augmented with the non media bloggers' internal 
links to their own
sites. This addition means that approximately 40 percent of DNC non 
media bloggers
linked to non bloggers' sites, with their site included, while 
approximately 47 percent of
links from RNC non media bloggers linked to other non media bloggers' 
sites, with links
to their own site included. When non media blogger links to other non 
media bloggers are
aggregated with non media bloggers' own internal links, this 
aggregated figure for non
media blogger links makes links to non media bloggers the highest 
category in terms of
links within the networks, surpassing even the amount of links 
pointing to the aggregated
traditional media category.
Data from this table also answers research question one which 
questions whether
media bloggers tended to link more to other media bloggers as opposed 
to non media
bloggers. Like non media bloggers, media bloggers were also 
significantly more likely to
link to other non media bloggers as opposed to media bloggers. DNC 
media bloggers
allocated over 19 percent of their links to non media bloggers' sites 
in comparison to
slightly over 2 percent to other media bloggers' sites. Similarly, 
RNC media bloggers
dedicated a higher percentage of their links to other non media 
bloggers: over 9 percent
of their links went to non media blogger sites in comparison to less 
than 2 percent for
other media bloggers. These results highlight that there was a high degree of
interrelationships in terms of nodes directing links or edges to 
other non media bloggers
from both non media bloggers and media bloggers. Notably, this 
relationship was more
pronounced for the DNC non media bloggers who were able to gain over 
twice as much
links from the DNC media bloggers versus the RNC non media blogger 
connection to the
RNC media bloggers. Another significant finding from this table is 
that media bloggers
tend to direct more traffic to their blog versus other media blogs. 
Examining the table's
data for internal linking, DNC media bloggers direct 8 percent of 
traffic to their blog,
20
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
four times more than to other media blogs. Similarly, the RNC media 
blogs directed over
4 times the amount of traffic to their blog when compared to their 
link traffic directed to
other media blogs. It was apparent that media bloggers were more 
prone to link inward if
they were referencing a media blog.
Based on past studies that have found a tendency in mainstream media to limit
external linking practices, this study hypothesized that like big 
media, media bloggers
will also tend to link less externally when compared to the non media 
blogger networks
covering the democratic and the republican conventions. Table (5) 
presents a summary of
link totals across the four networks, together with the numbers of 
site bloggers. Using
central tendency statistics to find the mean links for each of the 
four blogger networks
provides strong support for the hypothesis that media bloggers linked 
less than non media
bloggers as both networks of non media bloggers had higher link means 
per blog than the
media bloggers. With a sumtotal of 1090 links divided over 21 blogs, 
RNC non media
bloggers had the highest mean links per blog, with 52 links per blog 
for the five day
period. DNC non media bloggers covering the Democratic National 
Convention also had
higher mean links per blog than the media bloggers for both 
conventions as well. With a
total of 1602 external links spread over 34 blogs, the non media 
bloggers had a mean of
47 mean links. When compared with these figures, the media bloggers for both
conventions had lower external links per blog. The DNC media bloggers 
totaled 862 links
spread over an assessed 25 blogs resulting in an average of 34 links 
per blog. The RNC
media bloggers produced a total of 1034 links over an assessed 24 
blogs yielding an
average of 43 links per blog. The fourth hypothesis is proven in that 
media bloggers for
both conventions continued the trend of the traditional online media 
in reducing the
percentage of links that point externally from their sites.
By harvesting the page citation URLs for a 5-day period in both 
conventions, it
was possible to supply richer analysis on the type of URLs cited by 
each network in an
attempt to compare not only the cross linkages between the different 
networks, but the
level of agreement as well as diversity in the types of URLs cited 
within the different
networks. The fourth research question questioned whether media and non media
bloggers would be encouraged to point more to the credentialed non 
media bloggers than
other non media bloggers. This hypothesis can be tested by reference 
to the top non
21
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
media bloggers linked to from the four networks of bloggers to see if 
there is greater
mention or higher weighting given to those non media bloggers who 
were at the specific
conventions in question as opposed to bloggers not credentialed for 
the event. Since the
network is much larger for non media bloggers, this finding would 
provide evidence that
bloggers do seek to define their network neighborhood by familiarity; 
in this case, could
a possible boundary condition be credentialed versus noncredentialed blogger?
Results for the top mentions of media and non media bloggers by each 
of the four
networks of bloggers are presented Table 7. According to the results, 
75% of the
mentions of non media bloggers by DNC non media bloggers were to 
those present at the
convention. DNC media bloggers mentioned DNC non media bloggers 
credentialed at the
convention in 80% of the top 20 mentions. DNC media and non media 
bloggers agreed
on nine of the bloggers that gained top spots in their highest links 
list, overall showing a
high level of agreement between DNC media and non media blogger networks.
Examining the networks that covered the Republican conventions showed less
agreement. RNC media and non media bloggers agreed on only 4 of the 
blog mentions,
and only three of those were present at the convention. RNC non media bloggers
mentioned other credentialed non media bloggers in 55% of the 20 
cases. For RNC media
bloggers, only 6 out of the 20 blogs were credentialed RNC bloggers. RNC media
bloggers made mention of many bloggers outside the RNC convention, 
including four
former DNC non media bloggers at the DNC convention. This additional finding
solidifies an earlier finding that RNC non media bloggers found 
greater difficulty in
getting mentioned among RNC media bloggers, while DNC non media bloggers were
able to gain more attention throughout both time periods.
Examination of which non media bloggers are mentioned in each network reveals
findings that provide some evidence of a shared partisan ideology 
among non media
blogger networks. Among RNC non media bloggers, five of the nine sites
(instapundit.com, michellemalkin.com, cayankee.blogs.com, 
drudgereport.com, and
alarmingnews.com) not credentialed as RNC non media convention 
bloggers could be
easily deemed as conservative because of their blogrolls and blog 
postings. However,
other pegged liberal blogs such as dailykos.com, wonkette.com, and 
pandagon are also
mentioned by the RNC non media bloggers. Closer inspection of the 
other networks
22
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
reveals that elite blogs, regardless of partisan ideology, stand a 
better chance of being
included as links when being cited by both media and non media 
bloggers, regardless of
the network's primary composition. For example, dailykos and 
instapundit are both
mentioned in all networks. Wonkette is mentioned in three of the four 
networks though
she is widely considered to be a liberal blogger. The media's choices 
to link to non media
bloggers outside the RNC non media convention bloggers can also be said to be
determined by the level of popularity and degree identification with 
the said blogger. For
example, RNC non media bloggers link to andrewsullivan.com, 
buzzmachine.com and
advance.net (both owned by Jeff Jarvis), and talkingpointsmemo.com, all former
journalists. These findings support some recent studies on the 
tendency of journalists to
connect to more popular bloggers who were once ex-journalists. In 
terms of network
theory, preferential attachment would suggest that vertices connect 
to those of the same
degree or status, and it can help to explain why journalists were 
more prone to link to
bloggers who were once journalists. In sum total, these findings 
provide some central
contradiction to the echo chamber theory in the form of the elite 
blogger who tends to be
cited regardless of partisan ideology. As such, there is survival of 
diversity of opinion
regardless of the partisan leaning of the blog.
Turning to the media cited among the four networks, was there agreement in
terms of which media was the most popular? Table 8 presents 
Spearman's rho correlation
analysis of the top 12 media across the four networks correlated 
against its placement in
each of the four networks. A correlation coefficient of .899 for DNC 
non media, 705 for
RNC non media bloggers, and .795 for RNC media bloggers was obtained. 
A very low
coefficient of .46 was obtained between the DNC media bloggers media 
agenda and the
general network agenda, mainly due to the low mention of the New York 
Times in the
DNC media agenda. The top media mentions for each network is shown in 
Table 9. All
of the blogger networks included the New York Times, the Washington Post, and
CNN.com, three elite media establishments, within their top media 
mention. Notably
absent in any of the DNC blogger networks is fox news, which is 
present in both the
RNC blogger networks. The DNC non media blogger network mentioned one news
source not mentioned by any of the others in their top 12—NPR. 
Interestingly, the RNC
non media network of bloggers have the most differences in their 
choice of media, with
23
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
some media not mentioned by any of the other three networks in the 
top 12 choices—the
Weekly Standard, the Washington Times, and the Kansas City Star. The 
tendency for
conservative bloggers to link to the Weekly Standard and the 
Washington Times has been
noted in a recent study (Adamic and Glance, 2005). Overall, it 
appears that the DNC non
media bloggers included more elite, traditional media, in their blog 
links as opposed to
the network of RNC non media bloggers who were more varied than the 
other three
networks, yet seemingly in agreement within their own network on some 
media choices
not shared by the other three networks.
Discussion
Cooper of Cnet news reported of convention blogging that "blogging 
blew its big
chance in Beantown" coming off like "cyberhayseeds in the big town." 
Bricklin (2004)
rebutted the notion that journalist and blogging was in competition, 
instead preferring to
concentrate on the unique qualities, benefits, and differences in 
style that both mediums
can bring to each other. Weinberger, non media blogger who crossover 
to be a media
blogger for Boston.com during the DNC conventions had this to say: 
"Read my pixels, I
am not a journalist" (Weinberger, 2004). Blogging enthusiasts and 
outspoken elite
bloggers see the blogging phenomenon as part and parcel of a new form of
journalism—dubbed many names including open source journalism, participatory
journalism, grassroots reporting, personal media (Lassica, 2003; 
Gillmor, 2004).
This study examined the nature of the connections and interrelationships among
non media and media bloggers covering the 2004 democratic national 
conventions and
the republican national conventions. A total of 104 blogs were 
examined across the four
networks of bloggers, and using primarily page citation link 
analysis, a sum total of 5,225
links were extracted from the blog postings for the total 10 day 
period under examination
in this study. Primarily, one of the main findings of this study was 
that of a disconnected
media blogger network throughout both convention periods. Media 
bloggers were rarely
cited by their fellow mainstream media bloggers, as well as by the 
other non media
bloggers. The non media bloggers gained more attention from the media 
bloggers than
they chose to give to their network. Instead, non media bloggers 
continued to turn their
attention to traditional media. Network theory correctly predicted 
that non media
24
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
bloggers would tend to link more to other non media bloggers than 
media bloggers
because of preferential attachment as a result of former social 
bonds, familiarity with the
said network neighborhood, and partisan shared ideologies.
Of great interest was the choice that media bloggers made to direct 
more links to
non media bloggers than media bloggers, and to point links to their 
own media blogs,
while limiting external links. It could be possible that many 
journalists were sent on
assignment to cover the story of the non media bloggers. An 
interesting finding was that
RNC non media bloggers were unable to get media attention with as 
much frequency as
DNC non media bloggers. No data exists to test why this finding 
arose. Similar studies
would need to be conducted to see whether there is a difference in 
the influence of
democratic and conservative networks on the media's agenda.
Was there evidence of an echo chamber in terms of choice of sources? Most
bloggers credentialed for the convention could be identified as 
sharing consistent
ideologies and linked to other bloggers who shared their viewpoint. 
Yet, elite bloggers of
different partisan viewpoints gained easy entrance into non media 
blogs as well as media
blogs. Traditional media gained high positions in each of the four 
networks, once again
identifying that diversity that exists in the network in that 
bloggers depend on the media,
even in situations of eye witness accounts and first hand event 
reporting, for material to
comment upon. However, there was media mentioned among the non media networks
that did not gain precedence among either the media network or the 
general network.
This study found many differences in the media and non media blogger networks
in terms of their linking practices. Indeed, both can be said to 
contribute different
strengths to the news making process. It now remains to be seen 
whether the closer
alliances forged between media and non media networks will result in 
a change in
practices within either of these networks. With the movement of some 
elite non media
bloggers into mainstream newsrooms8, and with more and more 
journalists joining the
ranks of bloggers, it is possible that we are beginning to move 
towards a more open form
of conversation among mainstream media and citizen journalists.
8 InstaPundit blogger Glenn Reynolds has a blog/column for MSNBC.com, 
CalPundit Kevin Drum's blog is at Washington Monthly's
Web site, and Mickey Kaus' Kausfiles is at Slate.
25
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Bibliography
Adamic, L.A. and N. Glance. (2005). The political blogosphere and the 
2004 U.S.
Election: Divided they blog. H.P. Labs. Available online at
http://www.hpl.hp.com/research/idl/papers/politicalblogs/
Adamic, L.A. and E. Adar. (2003). Friends and neighbors on the Web. 
Social Networks
Available on the Web at http://www.cs.man.ac.uk/~rizos/web10.pdf
Adamic, L.A. and B.A. Huberman. (2001). The Web's hidden order. 
Communications of
the ACM, 44(9).
Adamic, L.A. and B.A. Huberman. (2000). Power-law distribution of the 
world wide
web. Science, 287.
Barabasi, A.L., R. Albert, H. Jeong, and G. Bianconi. (2000). Power 
law distribution of
the World Wide Web. Science, 287.
Barabasi, A-L.and R. Albert. (1999). Emergence of scaling in random 
networks. Science,
286.
Barabasi, A.L. (2002). Linked: The new science of networks. Perseus: 
Cambridge, MA.
Bricklin, D. (2004). Blogging breakthrough in Boston. CNET news. 
Available online at
http://news.com.com/Blogging+test+pilots+in+Boston/2010-1025_3-5293461.html
Broder, A., R. Kumar, F. Maghoul, P. Raghaven, S. Rajagopalan, R. 
Stata, A. Tompkins,
and J. Winer. (2000). Graph structure in the Web. Computer Networks. 
Available on the
Web at
http://www.almaden.ibm.com/webfountain/resources/GraphStructureintheWeb.pdf
Cooper, C. (2000). Cybertourists in Boston. CNET news. Available online at
http://news.com.com/Cybertourists+in+Boston/2010-1028_3-5289475.html?tag=nl
Deuze, M. (2002). The Internet and its journalisms. Available at
http://www.ojr.org/ojr/future/10264077729.php.
Dimitrova, D.V.(2003). Hyperlinking as gatekeeping: online newspaper 
coverage of the
execution of an American terrorist. Journalism Studies, 4(3).
Dorogovtsev, S.N., J.F.F. Mendes, and A.N. Samukhin. (2000). 
Structure of growing
networks with preferential linking. Physical Review Letters, 85 (21).
Drezner, D.W. and H. Farrell. (2004). The power and politics of 
blogs. Presented at the
American Political Science Association.
26
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Eckman, J.P. and E. Moses. (2002). Curvature of colinks uncovers 
hidden thematic layers
Blogspace. Communications of the ACM, 47(12).
Hindman, H., K. Tsioutsiouliklis, J.A. Johnson. (2003). Googlearchy. 
How a few heavily
linked sites dominate politics on the Web. Presented at the Annual 
Meetings of the
Midwest Political Science Association.
Houston, F. (2000). Enjoy the ride while it lasts. Columbia Journalism Review,
July/August. Available on the Web 
http://archives.cjr.org/year/00/2/houston.asp
Huberman, B.A and L.A.Adamic. (1999). Growth dynamics of the World Wide Web.
Nature, 40.
Huberman, B.A, P.L.T. Pirolli, J.E.Pitkow, and R.M.Lukose. (1998). 
Strong regularities
in World Wide Web surfing. Science, 280(3).
Kottke, J,. (2003). Weblogs and power laws. Available at
http://www.kottke.org/03/02/weblogs-and-power-laws
Kovarick, B. (2002). Web design for the mass media. Boston: Allyn & Bacon.
Lazersfeld, P.F. and R.K.Merton. (1954). Friendship as a social 
process: a substantive
and methodological analysis. In Freedom and control in modern 
society, ed. M. Berger,
pp. 18-66. New York: Van Norstrand.
Lewin, K. (1947). Frontiers in group dynamics. Channels of group 
life: Social planning
and action research. Human Relations.
Marlow, C. (2004). Audience structure in the weblog community. 
Presented at the
Association for Education in Journalism and Mass Communication, New Orleans.
McPhearson, M., L. Smith-Lovin, and J.M. Cook. (2001). Birds of a 
feather: homophily
in social networks. Annual Review of Sociology, 27: 415-44.
in the World Wide Web. Proceedings of the National Academy of 
Sciences of the United
States of America, 99(9).
Erdos, P. and R. Renyi. (1960). On the evolution of random graphs. 
Magyar Tud. Akad.
Mat. Kutato. Int. Kozl.
Gillmor, D. (2004). We the media: Grassroots journalism by the 
people, for the people.
Available online at http://wethemedia.oreilly.com/.
Glaser, M. (2004). Bloggers, Citizen media and rather's fall -- 
little people rise up in
2004. Available at http://www.ojr.org/ojr/stories/041221Glaser/ .
Kumar, R., J.Novak, P. Raghavan, and A. Tompkins. (2004). Structure 
and Evolution of
27
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Neuman, M.E.J (2003). The structure and function of complex networks. 
SIAM, 45(2).
Nielson, J. (2003). Diversity is power for specialized sites. Available at
http://www.useit.com/alertbox/20030616.html.
Journalism.org. (2005). State of the media report. Available at
http://www.stateofthenewsmedia.org/2005/index.asp.
Kleinberg, J.M. (1999). Authorative sources in a hyperlinked 
environment. Journal of the
ACM, 46(3).
Olafson, S. (2003) A reporter is fired for writing a weblog: Nieman 
Reports, Fall.
Peng, F.Y., Tham, I. Naphtali, and H. Xiaoming. (1999). Trends in 
online newspapers: a
look at the U.S. Web. Newspaper Research Journal, 20(2).
Pew Internet Reports (2004). The state of blogging. Available at
http://www.pewinternet.org/PPF/r/144/report_display.asp.
Pennock, D.M., G.W. Flake, S. Lawrence, E.J.Glover, and C. L. Giles. 
(2002). Winners
don't take it all: Characterizing the competiton for links on the 
Web. Proceedings of the
National Academy of Sciences of the United States of America, 99(8).
Price, D. (1965). Networks of scientific papers. Science, 149.
Shirky, C. (2003). Power laws, weblogs, and inequality. Available online at
http://www.shirky.com/writings/powerlaw_weblog.html.
Sifry, D. (2005). The state of the blogosphere. Available at
http://www.sifry.com/alerts/archives/000298.html.
Singer, J.B. (2004). The political j-blogger. Normalizing a new media 
form to fit old
norms and practices. Presented at the Association for Education in 
Journalism and Mass
Communication Conference, New Orleans, 2004.
Singer, J.B. (2001). The metro world wide web: changes in newspapers' 
gatekeeping role
online. Journalism & Mass Communication Quarterly, 78, pp. 65-80.
Singer, J.B. (1998). Online journalists: foundations for research 
into their changing roles.
Journal of Computer Mediated Communication, 4. Available online at
http://www.ascusc.org/jcmc/vol4/issue1/singer.html.
Shoemaker, P.J. and S.D.Reese. (1996). Mediating the message: 
theories of influences on
mass media content. London: Longman Publishers.
28
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Sustein, C. (2002). The law of group polarization. The Journal of 
Political Philosophy.
10(2).
Sustein, C. (2001). Republic.com. New Jersey: Princeton University Press.
Sustein, C. (2000). Deliberative trouble? Why groups go to extremes. 
Yale Law Journal,
110(1): 71-119.
Trench, B. and G. Quinn. (2003). Online news and changing models of 
journalism. Irish
Communications Review, 9.
Tremayne, M. (2004). The Web of Content: applying network theory to the use of
hyperlinks in journalism on the Web. Journalism and Mass 
Communication Quarterly,
81(1).
Welsch, P. (2005). Revolutionary vanguard or echo chamber? Political 
blogs and the
mainstream media. Available online at http://www.blogninja.com/index.php.
White, D.M.(1950). The gatekeeper: a case study in the selection of 
news. Journalism
Quarterly, 27.
29
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 1: List of Bloggers Covering the Democratic and Republican Conventions
Non Media
Alan Nelson
Jay Rosen
Kevin Aylward
Scott Sala
Brian Reich
Matt Margolis
Captain's Quarters
BoiFromTroy
Power Line
Dean's World
Spot On
INDC Journal
Real Clear Politics
Red State
Tacitus
Irish Lass
Roger Simon
OxBlog
Taegan Goddard
Interplanetary
Travis Lenkner
* Entire column reflects total of page citation links exclusive of 
blogroll links for the entire 5-day period
under observation.
Appendix A: Tables
RNC
Media
138 * Jay Rosen
50 Chris Matthews
50 Charles Madigan
64 Ellen Warren
105 Rick Neudorff
81 Wireless Election
24 Kevin Anderson
37 Joshua Clover
88 Anya Kamenetz
32 Pamela Brown
82 Tony Cobb
120 Alan Bjerga
27 Steve Love
76 Adam Smeltz
37 Tom Webb
24 Dave Barry
17 Walter Mears
119 Dan Kennedy
119 Marc Cooper
0 Joshua Bearman
92 Ben Ehrenreich
Steve Mikulan
Matt Welch
Non Media
49* Alan Nelson
78 Dave Winer
13 Taegan Goddard
0 Jay Rosen
0 Markos Moulitsas
325 Jerome Armstrong
21 Aldon Hynes
20 Jeralyn Merritt
16 Tom Burka
0 Paul McCullum
0 OxBlog
34 Rick Heller
83 Matthew Gross
41 Byron LaMasters
2 Jessamyn Charity
30 Dave Pell
105 Natasha C
79 Jesse Taylor
21 Michael Feldman
26 Christopher Rabb
0 Kirk Johnson
14 Bill Scher
178 Joe Rospars
Matt Stoller
Alison Teal
Nathan Paxton
Christian Crumlish
Dave Johnson
John Burns
Sean Munison
Jusiper
Steve Soto
DNC
Media
57*
103
71
35
0
34
95
168
27
27
0
23
106
0
8
6
0
5
0
0
0
135
47
0
168
191* Dave Barry
151 Matt Welch
138 Gotham Gazette
41 Eric Zorn
71 Ellen Warren
57 Charles Madigan
94 Chris Matthews
75 Convention Blog
4 Stewart Powell
0 Adam Smeltz
58 Daniel Rubin
107 Michael Klein
43 Alan Bjerga
93 Tom Webb
27 Ruben Navarrrette
58 Poynter
33 Timm Herdt
44 Terry Neal
20 Collin McNickle
42 Dave Cook
20 Peter Robinson
47 Newsweek
19 Dave Weinberger
96 Convention Confid.
0 Coonvention Blog
55
89
78
11
39
5
32
30
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 2: Top 20 Citation Links Through The Blogger Networks
Top Links in All Networks
152*
104
102
90
82
45
43
41
39
35
31
30
30
30
29
28
28
27
25
25
nytimes.com
msnbc.msn.com
washingtonpost.com
boston.com
cnn.com
dems2004.org
dailykos.com
usatoday.com
news.yahoo.com
latimes.com
akamai.com
imdb.com
instapundit.com
talkleft.com
news.bostonherald.com
online.wsj.com
pandagon.net
bopnews.com
amazon.com
command-post.org
Top Links in DNC Media
23*
18
17
14
13
13
12
12
11
11
10
10
9
9
8
8
8
8
7
7
cnn.com
boston.com
reason.com
kansas.com
chicagotribune.com
washingtonpost.com
talkleft.com
latimes.com
msnbc.msn.com
dems2004.org
command-post.org
johnkerry.com
imdb.com
dailykos.com
chron.com
amazon.com
whitehouse.gov
pandagon.net
usatoday.com
logicalrealism.org
* Represents the links through the 10-day period under observation
Top Links Among DNC Non Media
67*
32
31
24
23
20
20
19
18
17
17
16
15
15
14
13
10
10
10
10
nytimes.com
washingtonpost.com
boston.com
dailykos.com
news.yahoo.com
msnbc.msn.com
usatoday.com
bopnews.com
mydd.com
online.wsj.com
dems2004.org
talkleft.com
scripting.com
pandagon.net
cnn.com
latimes.com
npr.org
news.bbc.co.uk
pollingreport.com
conventions.nationaljournal.com
Top Links in RNC Non Media
40*
36
23
21
20
16
16
13
13
12
12
12
11
11
11
10
9
8
8
8
washingtonpost.com
nytimes.com
msnbc.msn.com
cnn.com
wizbangblog.com
powerlineblog.com
captainsquartersblog.com
flickr.com
foxnews.com
instapundit.com
redstate.org
weeklystandard.com
usatoday.com
news.yahoo.com
blogsforbush.com
nationalreview.com
startribune.com
strategicvision.biz
slantpoint.com
washtimes.com
31
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Top Links in RNC Media
50*
43
37
30
24
17
17
16
15
12
9
9
7
7
7
7
7
7
7
6
msnbc.msn.com
nytimes.com
boston.com
akamai.com
cnn.com
imdb.com
washingtonpost.com
dems2004.org
news.bostonherald.com
reasonmag.buzznet.com
ohio.com
georgewbush.com
house.gov
mbta.com
latimes.com
bostonphoenix.com
reason.com
protestwarrior.com
slate.com
redstate.org
* Represents the links through the 10-day period under observation
32
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 3: Links Across Categories for DNC and RNC Conventions
Category
Traditional Media
Non Media Blogger
Internal Link
Recreation
Activism/Grassroots
.Orgs
Candidate/Campaign Web sites
Government
Other .com
Software sites
Media Blogger
All others
TOTAL
Table 4: A Comparison of Media and Non Media Blogger Networks and Link
Category Frequency
Category Name
Traditional Media
Media Blogger
Non Media Blogger
Internal Link
Government
d.f. = 88, p <.001
* Columns not meant to total 100
DNC Media
(%)
24*
DNC Non
Media (%)
27*
3
26
13
2
Links (%)
28
24
12
5
4
4
4
3
3
2
2
9
100
RNC Media
(%)
RNC Non Media
(%)
31* 30*
2
19
8
5
2
9
9
6
2
31
16
2
33
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 7: Top Mentions of Non Media Blogs by Media and Non Media Bloggers for
the DNC and RNC Conventions
Top Non Media Blogs for DNC Non Media Bloggers
Blog
dailykos.com *
bopnews.com *
mydd.com *
talkleft.com *
scripting.com *
pandagon.net *
reachm.com/amstreet *
instapundit.com
burntorangereport.com*
atrios.blogspot.com#
journalism.nyu.edu/pubzone/weblogs/pressthink*
blogs.law.harvard.edu/dowbrigade*
command-post.org*
blogger.com
buzzmachine.com
seetheforest.blogspot.com*
talkingpointsmemo.com#
tomburka.com*
theleftcoaster.com*
electablog.com*
Top Non Media Blogs for DNC Media Bloggers
Blog
talkleft.com*
command-post.org*
dailykos.com*
pandagon.net*
logicalrealism.org*
mydd.com*
electablog.com*
blogs.law.harvard.edu/dowbrigade*
wonkette.com#
blog01.kintera.com/dnccblog#
mathewgross.com*
greaterdemocracy.org*
instapundit.com
mattwelch.com*
oxblog.blogspot.com*
pacificviews.org*
atrios.blogspot.com#
burntorangereport.com*
bopnews.com*
centristcoalition.com*
* blogs present at the convention
# blogs not credentialed, but present at the convention, liveblogging 
the event
Page Citation Links
24
19
18
16
15
15
9
9
9
8
8
7
7
6
6
6
6
6
5
5
Page Citation Links
12
10
9
8
7
5
5
5
5
4
4
4
3
3
3
3
3
3
3
2
34
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Top Non Media Blogs for RNC Non Media Bloggers
Blog
wizbangblog.com*
captainsquartersblog.com*
powerlineblog.com*
redstate.org*
instapundit.com
blogsforbush.com*
michellemalkin.com
slantpoint.com*
rogerlsimon.com*
wonkette.com
hughhewitt.com*
command-post.org*
dailykos.com
cayankee.blogs.com
deanesmay.com*
drudgereport.com
realclearpolitics.com*
pandagon.net
celluloid-wisdom.com
alarmingnews.com
Top Non Media Blogs for RNC Media Bloggers
Blog
blogger.com
instapundit.com
redstate.org*
journalism.nyu.edu/pubzone/weblogs/pressthink*
dailykos.com
buzzmachine.com
andrewsullivan.com
mattwelch.com*
thenationaldebate.com
atrios.blogspot.com
advance.net
blog.johnkerry.come
commonsensej.blogspot.com
kenlayne.com
hughhewitt.com*
powerlineblog.com*
bopnews.com
realclearpolitics.com*
wonkette.com
talkingpointsmemo.com
* blogs present at the convention
# blogs not credentialed, but present at the convention, liveblogging 
the event
Page Citation Links
20
16
16
12
12
11
8
8
7
7
7
7
6
6
6
6
6
5
5
5
Page Citation Links
6
6
6
5
4
4
4
4
3
3
2
2
2
2
2
2
2
2
2
2
35
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 8: Media and Non Media Blogger Rankings of Traditional Media
DNC Non Media DNC Media Media
Mentions Rank Mentions Rank Mentions Rank Mentions Rank Overall
Rank
1 67* 1 NYTimes
2 32 2 Washington Post
3 31 3 Boston.com
7 14 4 CNN
5 19 5 MSNBC
4 20 6 USA Today
8 13 7 LA Times
10 7 News.BostonHerald 8
6 17 9 Online.WSJ.com
12 0 10 Reason Magazine
11 4 11 Fox News
9 10 12 News.BBC.co.uk
*Each mention represents one link citation.
RNC Non Media RNC Media
36* 8 8*
40 6 9
4 2.5 17
21 1 23
23 5 10
11 7 7
3 4 12
3 9.5 4
8 11.5 0
0 2.5 17
13 11.5 0
1 9.5 4
1 43* 2
5 17 1
2 37 8
4 24 4
3 29 3
11 3 6
7.5 7 9.5
6 15 9.5
11 3 7
7.5 7 12
9 6 5
11 3 11
36
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Table 9: Top 12 Media Mentions Across the Four Networks
Top Media Mentioned by | DNC Non Media Bloggers
Blog
nytimes.com
washingtonpost.com
boston.com
usatoday.com
msnbc.msn.com
online.wsj.com
cnn.com
latimes.com
news.bbc.co.uk
npr.org
cbsnews.com
news.bostonherald.com
Top Media Mentioned by DNC Media Bloggers
cnn.com
boston.com
reason.com
kansas.com
latimes.com
msnbc.msn.com
washingtonpost.com
usatoday.com
nytimes.com
nationalreview.com
news.bbc.co.uk
news.bostonherald.com
Top Media Mentioned by RNC Non Media Bloggers
washingtonpost.com
nytimes.com
msnbc.msn.com
cnn.com
foxnews.com
weeklystandard.com
usatoday.com
nationalreview.com
startribune.com
washtimes.com
online.wsj.com
kansascity.com
Top Media Mentioned by RNC Non Media Bloggers
nytimes.com
boston.com
msnbc.msn.com
cnn.com
washingtonpost.com
news.bostonherald.com
latimes.com
reason.com
bostonphoenix.com
slate.com
foxnews.com
abcnews.go.com
Page Citation Links
67
32
31
20
19
17
14
13
10
10
8
7
23
17
17
14
12
10
9
7
6
4
4
4
40
36
23
21
13
12
11
10
9
8
8
7
43
37
29
24
17
15
7
7
7
7
6
5
37
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Figure 1
Figure 2
25
20
15
10
5
0
Frequency
Power Law Relationships Across DNC Media
Bloggers
Top 100 Links
news.yahoo.com
Appendix B: Figures
cbc.ca
nationaljournal.co
weblog.herald.com
siliconvalley.com
opinionjournal.com
cnn.com
msnbc.msn.com
whitehouse.gov
blogs.law.harvard.e
comedycentral.co
news.bostonherald
Mentions
38
carville.info
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
msnbc.msn.com
Figure 3
Power Law Relationships Across RNC Media
Bloggers
Figure 4
60
50
40
30
20
10
0
Frequency
news.bostonherald
reason.com
foxnews.com
andrewsullivan.co
Top 100 Links
michaelmoore.co
campaigndesk.org
rushlimbaugh.com
societyfortruthandj
rockthevote.com
opinionjournal.com
39
wired.com
amazon.com
Mentions
Event Blogging the 2004 Conventions:
Media Bloggers, Non Media Bloggers, and Their Network Connections
Figure 5
40