8/15/2019 Social News and Politics: How to digg an election
1/26
Content Analysis: How to Digg an election
By. Gabe McGuinness
Introduction:
The internet is the guiding force in the development of the new media .
Recent technologies are beginning to revolutionize every aspect of the news cycle .
As the 2008 presidential campaigns are ramping up it is clear that rich online media
such as video, blogs, podcasts, social networks, and even massively multiplayer
online role playing games (MMORPGs) will have a dramatic impact on the elections
outcome . Given the new medias recent ascension, and its ever-expanding reach, it
is obvious that it its role in shaping the publics political opinions will continue to
grow . Traditional means of measuring the medias impact on public opinion
(circulation, demographics, polling etc .) have proven less reliable in the era of new
media . Accordingly, new tools are beginning to emerge .
Using Yahoo! Pipes, an advanced RSS-feed manipulation tool, I will analyze
content appearing on leading social news website Digg .com . Digg has over a
million registered users, and is the leader among a new wave of social news
websites who all relinquish (in varying degrees) editorial control of content to their
user base . Users are permitted to submit stories from any source (new media or
old), which are then voted on by other users to increase or decrease their visibility
on the site (digging up, or burying), with the most popular stories eventually
reaching the front page . I cannot think of a better way to observe the impact of
8/15/2019 Social News and Politics: How to digg an election
2/26
new media on the democratic election process, than analyzing only democratically
selected stories . Because not all stories will be from new media sources, I will have
the added advantage of being able to gauge the old medias presence and
relationship to new media within the social news construct .
The general lens through which Ill be analyzing my data may be more
contextual than most content analyses, yet I intend to focus specifically on
how candidates fare within the greater context of social news. Early on in
my research, I realized that given the non-traditional nature of my research,
traditional means and methods would not always be applicable. Rather than
coding stories all relating to a specific topic, my data collection required me
to develop a coding system capable of analyzing data on a wide variety of
subjects, and from a wide variety of sources. The fragmented nature of new
media described by Doris A. Graber in her Mass Media and American Politics
(7 th edition) is echoed in the variety of stories produced by my research.
In chapter eight, Elections in the Internet Age (p.218-245) of
Grabbers aforementioned text, the large majority of the chapter is devoted
to cable, and broadcast television coverage of elections. Very little (about 2-
3 pages) of the chapter on the Internet Age is actually devoted to the
internet. I was surprised to find such a lack of information about new
technology, specifically the internet in a book bearing a 2006 copyright date.
I cant in good conscious, however blame her for this oversight. Indeed, she
disclaims her discussion of the internets impact on page 220:
8/15/2019 Social News and Politics: How to digg an election
3/26
Internet messages of all types are a new andconstantly changing territory where much analysisremains to be done.
Grabber does go into slightly more detail in chapter twelve where she
specifically discusses the Internet as a trend in media policy (p.362-365). It
is fascinating how quickly things she discusses change.
There is, as yet, no widely available solution to theproblem of finding ones way through the Internetslush jungles of information where search engines likeGoogle and Yahoo provide guidance, but often
perspectives skewed to business interests.Moreover, the stock of information that requiressearching doubles every few months. (p.363)
Although Graber alludes to RSS-feed technologys ability to automatically
deliver relevant information in her discussion of Blogs (p.364), her 7 th edition
of Mass Media and American Politics (2006) bears no discussion of the
phenomenon of Social News, or Web 2.0 services that are poised to, and
have indeed begun to drastically alter the relationship between old and new
media. The impact of YouTube.com alone can be seen in its $1.65 Billion
acquisition by Google, and its subsequent legal battles, and strategic
alliances with many mainstream media outlets and parent companies.
Given that this previous research is, as grabber points out still in its
infancy, I feel it has little to contribute in guiding my research. I feel that my
base of knowledge, and personal interest in being familiar with most of the
cutting edge developments in new media provide a good foundation for my
research. Indeed I feel that my membership in the demographic traditionally
8/15/2019 Social News and Politics: How to digg an election
4/26
8/15/2019 Social News and Politics: How to digg an election
5/26
to allow the sharing and proliferation of content across different websites.
Very simply, its feeds are streams of data listing titles and descriptions of
data, that link to that particular document. Another term Ill use throughout
my discussion is Web 2.0 described by Wikipedia.com as, a perceived
second-generation of Web-based servicessuch as social networking sites,
wikis, communication tools, and folksonomiesthat emphasize online
collaboration and sharing among users. The term has become a bit of a
buzz word, but essentially refers to a recent, noticeable shift to service
oriented web sites many of which act like programs that would traditionally
run on a computers desktop, and almost all of which involve a social or
collaborative element.
In the spirit of innovation with witch much of the new online media
presents itself, Ive decided to use online tools and techniques to guide my
research and data collection wherever possible. Ive constructed a programusing Yahoo!s new Pipes customizable web based RSS-feed manipulation
service that takes data from four different Digg.com categories, and filters
the results for popularity, and for the names of the 2008 candidates. The
result is a custom RSS-feed that displays only the most popular stories
featuring one or more of the candidates in the title and/or the description of
the story. The custom RSS-Feed draws popular stories from the main page of
Digg.com, as well as from Digg categories specific to the 2008 election,
politics, and political opinion. Only stories with more than 25 diggs will
make it through the filter. Stories not bearing one or more of the candidates
8/15/2019 Social News and Politics: How to digg an election
6/26
names in the title and/or description are filtered out. I use Googles web-
based Reader app to subscribe to the custom feed, which allows me to
save all stories the feed produces for later analysis.
For the purposes of this analysis, the media outlet Ill be analyzing is
Digg.com. It is important to note, however, that none of the stories analyzed
are published by digg.com, and that it simply acts as a means of connecting
users with interesting content elsewhere on the web. As a result, the news
stories Ive aggregated for analysis are from a variety of media outlets
including blogs, YouTube, and old media websites.
I began aggregating data on March 13 th around 12:00p.m. I stopped
aggregation April 16 th , also around 12:00p.m. Initially I included in my
customized feed stories from other social news sites such as Netscape.com,
Reddit.com, and Newsvine.com. I let the feed continue to automatically
aggregate data for over a month. Upon completion, I realized that Id
aggregated over 500 stories. Looking closer, I realized that the vast majority
of stories had come from Newsvine.com. I realized that the feeds from
Newsvine were not being restricted by the popularity filter I had in place, and
the several hundred stories I received as a result were basically a direct feed
of Associated Press stories that had passed my candidates name filter, the
majority of which had not reached popularity by community voting. I
decided to keep only stories obtained from Digg.com, because they
consistently featured a high number of votes, ensuring a respectable sample
8/15/2019 Social News and Politics: How to digg an election
7/26
of people viewed them. In total, I received 41 qualified stories, of which I
omitted 3 because the candidates they concerned have yet to announce
their candidacy, leaving me 38 stories to code and analyze.
The unit of analysis for this project is the paragraph. Overall, coding
by paragraph was effective, with a few caveats. It was hard to code for
transcripts of videos, and for interviews. I did not code actual videos that
were embedded in the stories (or were the stories themselves) I chose to
analyze their content only if it was transcribed within the article (as it was in
several cases). In the event that there was no transcription, I coded the
paragraphs detailing the video, or captions for it. When transcripts of
interviews were involved, I treated the first paragraph of the question, and
first paragraph of the answer as one coding unit.
I decided to use the classic four content categories for coding political
election content, with very little modification. They are, The Horse Race, The
Campaign Trail, Personality Traits, and Policy Issues. As a rule, I decided to
allow for contextual judgment to place difficult units into a proper category,
the few units to which this applied were easy to place when taking into
account the source, surrounding units, and overall thrust of the story. The
rules for units coded as part of, The Horse Race required they dealt
explicitly with issues of whos ahead and whos behind, polling numbers, and
comparative economic data. An example from a story on
WashingtonPost.com:
8/15/2019 Social News and Politics: How to digg an election
8/26
Clinton, of New York, continues to lead Obama andother rivals in the Democratic contest, according tothe latest Washington Post-ABC News poll. But heronce-sizeable margin over the freshman senatorfrom Illinois was sliced in half during the past month
largely because of Obamas growing support amongblack voters.
The rules for units coded for, The Campaign Trail required they
include problems, controversies, strategies, and political maneuvering
associated with the running of the candidates political campaign. For
example, when John McCain fumbled a question on H.I.V. Prevention in Africa:
What followed was a long series of awkward pauses,glances up to the ceiling and the image of one of Mr.McCains aides, standing off to the back, urgentlymotioning his press secretary to come to Mr.McCains side. (thecaucus.blogs.nytimes.com)
The rules for units coded for, Personality Traits required that units
expressed information regarding who a particular candidate is as a person,
including their religious, moral, and ethical beliefs, as well as likes and
dislikes. One such article, entitled, Ten Things You Didnt Know about Barak
Obama from the U.S. News and World Report website was quite popular,
and included gems such as:
10. His heroes are Martin Luther King Jr., MohandasGandhi, Pablo Picasso, and John Coltrane.
The rules regarding units Coded for, Policy Issues were that they be
directly indicative of a candidates position on a particular issue or issues,
8/15/2019 Social News and Politics: How to digg an election
9/26
including their voting record, and their personal views and statements on a
political issue. For instance, the SmallGovTimes.com piece on Ron Pauls
announcement of candidacy details his position on many issues:
Paul stands as one of the last remaining believers instrict enforcement of the Constitution and a limitedfederal government in Washington D.C. Paul ranunsuccessfully for the White House in 1988 under theLibertarian ticket, but now caucuses with theRepublican Party. His political platform includes lowtaxes, individual liberties, and a principled belief inthe right to life.
I found the process of coding for content to be much easier than that of
tone. When coding for content, it didnt matter that the specific topics of the
stories being coded were all different. When it came time to code for tone, I
had some issues to resolve. The only thing that all of the stories had in
common was that they featured at least one of the 2008 presidential
candidates, and that they were popular on Digg between March 13 th and April
16 th . How could I code a unit that was equally positive for Barak Obama,
and Negative for Hillary Clinton? Because not all stories were about a
particular topic, or candidate, I had to develop a system of coding for tone
that would effectively quantify each candidates treatment by the Digg
community. To do so, I made coding rules that allowed for each unit
(paragraph) to be coded positive, negative, or neutral for whichever
candidate was mentioned in that paragraph.
Rules for coding tone:
8/15/2019 Social News and Politics: How to digg an election
10/26
1. Each unit can have only one code per candidate mentioned. Positive,
negative, or neutral.
2. If no candidate is mentioned in a particular unit, the code goes to the
candidate or candidates mentioned in the Digg.com title or summary
of the story.
3. Each storys Digg.com title and summary are to be coded separately
from the actual story, with one code for the title and one for the
description. If more than one candidate is mentioned in the title and
summary, each candidate can be coded as per rule 1.
4. If a units code is not easily discernable from the first 3 rules, a
judgment call based on contextual analysis may be permitted.
Results:
I was a bit surprised, and not a little overwhelmed by the sheer volume
of data I was able to obtain for this analysis. Making sense of it all was no
small task. The coding process proved both challenging initially, and
invaluable ultimately. I anticipated a large variety of stories, from at least as
large a variety of sources, and was not disappointed in that regard. I was
pleased to see that the majority of stories were sourced from blogs in some
fashion. Whether liberal, conservative, personal, or professional, the sheer
volume of blogs vs. old media was encouraging a new media evangelist such
as me. Nearly two thirds of the stories were sourced from new media outlets
8/15/2019 Social News and Politics: How to digg an election
11/26
in the form of blogs, many with videos embedded from YouTube. Informally, I
did notice that on issues of fact, bloggers often linked, quoted, or referred to
old media sources most likely to lend authenticity to their stories, and gain
the readers trust.
The content categories I chose proved to be an effective means of
analyzing the data, and did not have a very hard time placing units into their
categories, with a minimal number of units leaving me on the fence. I
attribute this to simple coding rules that allowed me to make judgments
based on context when necessary. I was a bit surprised by the relatively few
units that coded for Personality Traits (5.14%). I expected that given the
wide open republican field, and the general lack of knowledge about all but
the most well established candidates (Clinton, McCain) that there would be
more information regarding candidates as people.
The Largest category coded was Policy Issues (38.11%), followed closely by
Campaign Issues (33.24%). The data suggests that Digg users are far
8/15/2019 Social News and Politics: How to digg an election
12/26
more interested in policy or campaign issues, than the personal lives of
candidates. That is not to imply that Diggers are less interested in scandal,
or intrigue than the rest of us. Many of the stories that dealt primarily with
campaign issues were centered on problems/controversies, and gaffes made
by candidates. For instance, several stories relating to Senator McCains
Baghdad Stoll and his comments on the safety of some Baghdad
neighborhoods were among the top coded for Campaign Issues because
they were framed as a campaign blunder on the part of McCain. Articles
concerning the, Horse Race made up nearly a quarter of coded units
(23.51%) and were for the most part from old media outlets. I attribute this
to the fact that old media conducts sponsored polls, and has the first access
to polling data.
I found the coding of each unit for its tone to be more difficult than it
was for content. There are obviously fewer coding options, and it is a moresubjective classification. Compounding the difficulties was the fact that my
coding rules for tone allowed me to code each unit multiple times, should the
unit involve multiple candidates. While there were not a lot of these
paragraphs, there were enough to necessitate a rule. One article I found
particularly difficult to code was from The Washington Posts website, the
Digg headline read. New Poll: Blacks Shift To Barack Obama, McCain Falls.
The articles content was coded primarily for Horse Race yet most units
featured both positive and negative statements for different candidates. The
unit below was coded positive for Giuliani, and negative for McCain:
8/15/2019 Social News and Politics: How to digg an election
13/26
In the Republican race, former New York mayorRudolph W. Giuliani, who recently made clear hisintentions to seek the presidency, has expanded hislead over Sen. John McCain of Arizona. Giuliani holdsa 2 to 1 advantage over McCain among Republicans,
according to the poll, more than tripling his margin of a month ago.
This type of code was by far the exception, rather than the rule. Most units
were easy to code, with a large majority of them (62.38%) falling into neutral
territory, as was expected. Several big stories made the cycle of news, and
had a large impact on the balance of positive vs. negative codes. These
stories got a lot of coverage, and were promoted on Digg multiple times in
different capacities; one was McCains aforementioned Baghdad issue,
bearing a decidedly negative tone and frame, Giulianis uneasy relationship
with the firefighters union, also negative, and bigoted Comments Made by
Ann Coulter about John Edwards.
8/15/2019 Social News and Politics: How to digg an election
14/26
It was interesting to note, that across all the articles relating to Ann
Coulters bigoted remarks about John Edwards, there was a much more
neutral net effect than the other major issues that played negatively. I
attribute this to the fact that most of the stories reporting the comments
were links to blogs that provided little or no commentary on the video clip
they were displaying, and that republican candidates were quick to distance
themselves from Coulter, and condemn the remarks. One of the stories
ended up being coded negatively for John Edwards, because the negative
remarks were transcribed multiple times. The Coulter Case if you will goes
to show that despite an apparent liberal bias in Diggs readership, a well
handled controversy can end up being a positive. Still, negative units
outweighed the positive ones 24.27% to 13.35%.
Coding for positive/negative/neutral tone by candidate allowed me to
assess the popularity of individual candidates, and political parties within theDigg community. It became obvious upon doing so that my initial suspicion
of a liberal bias was confirmed, and then some. I was actually surprised by
how liberal the coding results indicated the stories were. While most stories
were not excessively liberal or conservatively biased (save for a few), in
total, the division and preferences are clear.
8/15/2019 Social News and Politics: How to digg an election
15/26
8/15/2019 Social News and Politics: How to digg an election
16/26
43.02% of all units coded republican were coded negative as well. Even more
shocking is that only 7.26% were coded positive, about half were neutral. It
seems that not only do Diggers tend to promote positive stories about
democrats; they also seem to actively promote negative material on
republicans. It is clear that if an unbiased look at politics is your goal, Digg
should not be the first place you turn.
Having the benefit of automatic story selection for this analysis
afforded me to analyze other sets of data, related to the coding. To further
illustrate this division between Liberal and Conservative issues on Digg, I
tabulated the total number of diggs (votes) each of the positively,
negatively, and neutrally coded stories from both parties received. As you
can see, negative stories about republicans received about 1300 more diggs
than positive stories about democrats. It is also interesting to note that the
stories that were expressly positive for democrats received more diggs than
8/15/2019 Social News and Politics: How to digg an election
17/26
the negative and neutral stories combined, despite the fact that they coded
much more neutral than positive or negative. The same effect is seen on the
republican side of the coin.
While it is unlikely that these figures would change significantly over time, it
should be noted that they represent a snapshot in time of the digg count.
Because once a storys popularity has peaked, it can still receive diggs, the
more recent stories Ive analyzed may not yet have reached their full
potential.
8/15/2019 Social News and Politics: How to digg an election
18/26
8/15/2019 Social News and Politics: How to digg an election
19/26
While Hillary narrowly edged out Paul as the third most positively
coded candidate (behind Richardson) her negative ratings were the highest
of all the democrats. I attribute this to the fact that many of the units
mentioning her were in a negative comparison to Obama.
Rounding out the democratic field of candidates are John Edwards, and
Dennis Kucinich, who both, along with Ron Paul enjoyed an outrageouslylarge neutral percentage. This is indicative of a lack of both stories, and
strong feelings on these candidates.
8/15/2019 Social News and Politics: How to digg an election
20/26
Interestingly enough, Edwards is the candidate who has put forth the
greatest effort to embrace new media, announcing his candidacy on the
video blog Rocketboom, and then on YouTube the following morning. He was
also the candidate who seemed to be part of the most bizarre stories;
including the vandalism of his virtual campaign headquarters in the
massively multiplayer online role-playing game Second Life, the evacuation
of his actual campaign headquarters upon receipt of an envelope full of white
powder, and the below the belt blow by conservative pundit Ann Coulter
calling him a, faggot. This hodgepodge of stories earned him a nearly 2:1
positive-negative ratio.
The republican field of candidates looks about how you would expect.
The candidate with the highest positive rating is also the one least likely to
earn the nomination. Mitt Romney fared almost as positive as Bill
Richardson, yet with a 46.15% negative code rating. I can attribute thisresult to a lack of stories mentioning Romney as well.
8/15/2019 Social News and Politics: How to digg an election
21/26
The remaining two Republican candidates only truly seemed to be
competitive in which could have the most negative stories about them
promoted on Digg. Ultimately Ill have to crown Giuliani the winner in that
respect, as it seems reports couldnt help but mention repeatedly how his
liberal views are hurting him, and how big a hurdle his ties to 9/11 are for his
campaign. He came in with a whopping 58.82% negative, and only 7.35%
positive thus earning him the title of most polarizing candidate, with only
38.46% of codes neutral.
Despite Giulianis strong showing at sucking in the eyes of diggers,
John McCain was able to save face, with the lowest positive rating among
major candidates (3.39%) Kucinich barely edged him out for the absolute
title by 0.16%. McCains various blunders have already been mentioned, and
8/15/2019 Social News and Politics: How to digg an election
22/26
contributed primarily to the fact that he received only 2 units coded positive
for him.
Summary & Discussion
While my initial suspicions on the liberal slant of Digg users was
confirmed, I was surprised at how one sided the results proved to be.
Perhaps over a longer time period, or different news cycles the results would
have been slightly more balanced. It does seem informally that republicans
have been getting a lot of negative press lately, and perhaps that is echoedand amplified in my findings.
While the data Ive gathered provides a good exploration of how social
news can be used as a means of information gathering, and analysis much
more work remains to be done. The readership of Digg is growing at a strong
rate, having recently passed the one million user mark. While early adopters
have proven to be younger and more liberal, it will be interesting to see how
those demographics shift as the idea catches on. Already weve begun to
see spin offs of digg, that while not quite ready to be included in this report,
are capable of providing the same sort of user-driven editorial control as
digg, to perhaps different demographics. It will be interesting to watch the
progress of social news as a medium in the new media.
While the methods Ive used for this analysis have yet to be proven
against the test of time, and repetition. I hope Ive provided a base for
continuing research to build upon when studying new media. I encourage
8/15/2019 Social News and Politics: How to digg an election
23/26
those that will follow to embrace the newest tools to study the newest
mediums, and further their progression. With an investment in
understanding of all Web 2.0 media has to offer, political parties, as well as
corporations, and organizations will better be able to manage their image in
the eyes of the cutting edge public, a valued demographic.
Bibliography
1. Grabber, Doris A . Mass Media & American Politics . 7th ed . Washington, DC:CQ P, 2006 .
2. "RSS ." Wikipedia, The Free Encyclopedia . 18 Apr 2007, 05:23 UTC . WikimediaFoundation, Inc . 19 Apr 2007< http://en .wikipedia .org/w/index .php?title=RSS&oldid=123736511 > .
3. "Web 2 .0 ." Wikipedia, The Free Encyclopedia . 18 Apr 2007, 16:31 UTC .Wikimedia Foundation, Inc . 19 Apr 2007< http://en .wikipedia .org/w/index .php?title=Web_2 .0&oldid=123842365 > .
http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=123842365http://en.wikipedia.org/w/index.php?title=RSS&oldid=123736511http://en.wikipedia.org/w/index.php?title=Web_2.0&oldid=1238423658/15/2019 Social News and Politics: How to digg an election
24/26
8/15/2019 Social News and Politics: How to digg an election
25/26
Coding Tables
ContentCategories Units
Percentage
Horse Race 87 23.51%Campaign Issues 123 33.24%
Personal Traits 19 5.14%Policy Issues 141 38.11%
# %Old Media Stories 14 36.8%
New Media 24 63.2%
StoriesCoded: 38
Candidate:
Richardson
Kucinich
Edwards
Obama
Clinton
Giuliani
McCain Paul
Romney
Totals
PositiveUnits 3 1 2 30 6 5 2 4 2 55
NegativeUnits 0 3 1 5 14 40 31 0 6 100
NeutralUnits 16 27 28 61 36 23 26 35 5 257
% positive 15.79% 3.23% 6.45%31.25
%10.71
%7.35
%3.39
%10.26
%15.38
%13.35
%
%negative 0.00% 9.68% 3.23%5.21
%25.00
%58.82
%52.54
%0.00
%46.15
%24.27
%
%neutral 84.21%87.10
%90.32
%63.54
%64.29
%33.82
%44.07
%89.74
%38.46
%62.38
%
8/15/2019 Social News and Politics: How to digg an election
26/26
Stories
Units Tone: By party
Democrats Republicans Democrats % Republicans %Positive Units 42 13 18.03% 7.26%Negative Units 23 77 9.87% 43.02%Neutral Units 168 89 72.10% 49.72%Total Units 233 179
Number of Diggs By Tone/Party:
Storytype/party
Positive/D
Negative/D
Neutral/D
Positive/R
Negative/R
Neutral/R
Number of diggs: 11802 5088 4826 2453 13182 2618Number of stories: 12 4 5 2 12 3Avgdiggs/story 983.5 1272 965.2 1226.5 1098.5
872.66667
DemocraticStories 21RepublicanStories 17