1 .uz . .-

....m..

, hfifizfiar?
I .rts.

”mumngn. . .

.1
.6. rxtl.
II a»: a

‘13:.
1:53.121}...

,1:
E.
u..- .flfl‘urfim Hula
.. . _
1. .. :il
Ii... $55!"..h‘

 

\
.4. .2
2.1.3:! “I...
:3v1.i.§:.$~.o 5v
£354..“1yt ,
. 7:335 3531.! «:1
.9?..2.§t‘\W{si. a.
1.14.33)? 1. {£31313
, .11..
5‘...- «#:13va
3.‘r3¢‘§i..}
‘3‘!!!
fia"..r.\53
38$!

.53.. 2|:
2‘25341...
.33 .0

l

:1 £23.} l

O. 9! it
1.,

 

a“ .QMfi. ‘ . . , .. . . , fir:

m, 3 33.3%,»...si ,_

2.3.5:. _ 3 L: I.

.. .1:
. V

 

 

 

This is to certify that the
Dissertation entitled

THE SELF AND ONLINE POLITICS: A SOCIAL
PSYCHOLOGICAL APPROACH TO POLITICAL
COMPUTER-MEDIATED COMMUNICATION IN THE
UNITED STATES

presented by

Benjamin Gross

has been accepted towards fulfillment
of the requirements for the

Doctorate degree in Sociology:

 

 

7 Major Professor? Signature

April 28, 2010

Date

MSU is an Affirmative Actbn/Equal Opportunity Employer

i-.--—--I----u-.--o—

- —.—

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:/Proleoc&Pros/ClRC/DateDuo.mdd

 

THE SELF AND ONLINE POLITICS: A SOCIAL
PSYCHOLOGICAL APPROACH TO POLITICAL COMPUTER-
MEDIATED COMMUNICATION IN THE UNITED STATES
By

Benjamin Gross

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Sociology

2010

ABSTRACT

THE SELF AND ONLINE POLITICS: A SOCIAL
PSYCHOLOGICAL APPROACH TO POLITICAL COMPUTER-
MEDIATED COMMUNICATION IN THE UNITED STATES

By

Benjamin Gross

As computers have become a more prominent part in the daily lives of people in
the United States, social researchers have begun to grapple with the impact of this new
technology on our everyday lives. Foremost among these concerns is how people
understand and connect to politics, and how these new forms of interaction have an
impact on political discourse in the United States. In this dissertation, my focus is placed
upon how an Internet user will seek information about politics, which kinds of other
people will the user likely prefer interacting with online, and whether the offline social
living environment of the user has a meaningful effect on these behaviors. Drawing on
quantitative analysis, I attempt to explain and illustrate how the social attitudes, offline
living environments, and political identities of the users impact how they connect to

politics in cyberspace.

This dissertation is dedicated to Brenda Sue and William T. Gross, Sr., my mom and dad,
who are the most wonderful and inspirational people that I have ever known.

iii

ACKNOWLEDGEMENTS

There are many who have helped me reach this point, and deserve recognition
before moving any further. I wish to thank the Department of Sociology for providing me
with a strong education, and the graduate school for funding this dissertation through a
fellowship. Stan Kaplowitz has been a supportive advisor from my first moments in this
program, and I am very grateful for all of his efforts. I consider his expertise in Social
Psychology and Statistics second to none, and I appreciate all the guidance I have
received from him over the years. The Social Attitudes course I took with Stan was
perhaps the most interesting one I have ever taken, and the knowledge I gained there is
invaluable to me. Toby Ten Eyck has also been a superb advisor and friend, whose
guidance through my second year paper and comprehensive exams was exemplary. Other
than being a great professor and an impressively rounded scholar, he is also a tremendous
painter and author as well. Not too shabby, eh? Ralph Pyle has been remarkably
accessible and supportive of me, and his advice has always been excellent. He is an
exceptional instructor and communicator, whose teaching serves as a model for how I
instruct my own Sociology courses. Frank Boster has given my insightful feedback,
particularly during the dissertation proposal defense, which turned out greatly improve
the work which follows. I am very grateful for the course in Persuasion he allowed me to
take as an independent study, and found this course to be incredibly informative and
insightful, which I expect to guide me through my professional career.

I also wish to thank former committee members Cliff Broman and Steve Gold, for
the guidance they have given and the expertise they have shared. Both gave me excellent

iv

advice and feedback during the writing of my second year paper. Additionally, the
support I have received from my colleagues has been priceless, and I am particularly
appreciative of the attention and advice of three great friends: Chris Oliver, Josh Woods,
and the late Kimberly Perez. All are these people are wonderful scholars and dear friends.
My family and friends all deserve heartfelt thanks, especially my mom and dad.
Both have been a continual source of love, support, sacrifice and inspiration to me my
entire life. I love them both, and I hope this dissertation makes them feel proud of their

youngest son.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ viii

CHAPTER 1: An introduction and literature review of political computer-mediated
communication in the United States.

Abstract ............................................................................................ 1
Introduction ...................................................................................... 1
A Review ofPrevious Online US Presidential RacesZ
Important Questions Still Unresolved ......................................................... 5
An Overview of this Dissertation .............................................................. 8
Summary ......................................................................................... 21

CHAPTER 2: Political News Sources on the Internet: Similarities and Differences.
during the US Presidential Election of 2008.

Abstract ........................................................................................... 23
Introduction ...................................................................................... 23
Democracy and the Public Sphere ............................................................. 24
The Journalists ................................................................................... 30
The Issue of Bias ................................................................................ 35
The Audience ..................................................................................... 37
Blogging: the good, the bad, and the ugly ...................................................... 40
The Newsmakers ................................................................................ 44
Presidential Candidates and the Presentation of Self47
Media Frames and Political Discourse ....................................................... 49
Methodology: the 2008 Presidential race and online news coverage .................... 51
Analysis and Findings .......................................................................... 54
Discussion and Conclusion .................................................................... 68

CHAPTER 3: The Self Goes Online: How Personal Attributes Mediate Political Activity
on the Internet.

Abstract ......................................................................................... 73
Introduction .................................................................................... 73
Community and the Internet .................................................................... 74
The Self and Identities ......................................................................... 83
Uses and Gratifications: The Self goes online .............................................. 93
Hypotheses ...................................................... 95
Methods ......................................................................................... 97
Variable Construction and Measurement .................................................. 97
Measuring Self-Construal ..................................................................... 98
Measuring Strength of Political Identity .................................................. 100
Measuring Political Interest Levels .......................................................... 101

vi

Measuring Political Activity Online ......................................................... 102

Measuring Online Political Information Seeking .......................................... 103
Data Analysis and Results ..................................................................... 104
Other Relevant Findings Coming from the Data Analysrs ............1 11
Discussion and Conclusions ................................................................... 113

CHAPTER 4: What Makes Someone a Cyber Balkan? Finding the Linkages Between
Social Psychology and Self-Selectivity in US Politics Online.

Abstract .......................................................................................... 116
Introduction .................................................................................... 1 16

Like-Minded Selected Communities and Attitude Polarization... . .. .. .. .............118
SIDE Theory and Online Interaction” 122
The Daily Me: Consumer Ideology and Selective Exposure 126
Selective Exposure ................................................................................. 128
User Characteristics Potentially Tied to Cyber Balkanization. . . . .. . .. . . . . .. . . .............132
Offline Environments .............................................................................. 132
Trust ............................................................................................. 134
Dimensions of Political Identity: Direction, Strength, Emotion & Certainty. . . .135
Personality, Personal Motivations, and Demographic Factors .......................... 136
Emotions ........................................................................................ 138

Methodology ................................................................................... 139
Dependent Variable Construction and Measurement ...................................... 140
Predicted Relationships to Cyber Balkanization ........................................... 142
Regression Results: “CB Actions” as the Dependent Variable ......................... 143

Regression Results: “CB Community” as the Dependent Variable ..................... 146
Discussion and a Summary of the Findings ................................................ 149

Findings Summary: Cyber Balkanization defined by Actions of the Internet User. . .151
Findings Summary: Cyber Balkanization as defined by self-selective community. . . 153

CHAPTER 5: Findings, Limitations, and Directions for Future Research

Introduction ..................................................................................... 155
Summary of Main Findings ................................................................... 155
Limitations of the Dissertation ............................................................... 159
Directions for Future Research ............................................................... 163
APPENDICES

Appendix A: Content Analysis Coding Sheet used in Chapter Two ..................... 166
Appendix B: Alphabetical Listing of Each Survey Item used ............................ 168
BIBLIOGRAPHY .............................................................................. 176

vii

LIST OF TABLES
Table 1: Candidate Framing by Different Sources ............................................. 55

Table 2: Does Source Allow Barack Obama to assert his stance in his own words? ....... 56

Table 3: Does Source allow McCain to assert himself in his own words? ...................... 57
Table 4: Which Sources are allowing others to define John McCain? ............................. 59
Table 5: Does the Article allow others to define where Barack Obama stands? ............. 60
Table 6: Which Sources are criticizing the media coverage? ........................................... 62
Table 7: "Horserace" Coverage by News Source ............................................... 63
Table 8: Topics Covered During the Presidential Race by Source .......................... 65

Table 9: Does the Writer Express Hostility Towards the Candidate? ............................. 66

Table 10: Does the Article Accuse the Candidate of Dishonesty? .................................. 67

Table 11: Factor Analysis of Self-Construal Scale ............................................ 99
Table 12: Component Loadings Matrix for Self Construal ................................... 99
Table 13: Factor Analysis Results for Strength of Political Identity ...................... 100

Table 14: Component Loadings Matrix for Strength of Political Identity ................ 101

Table 15: Correlated Measures for the Variable Political Interest .......................... 102
Table 16: Factor Analysis Results for Political Activity Online ............................ 103
Table 17: Component Loadings Matrix for Political Activity Online.......................103
Table 18: Factor Analysis Results for Political Information Seeking ..................... 104
Table 19: Component Loadings Matrix for Political Information Seeking ............... 104
Table 20: Regression Results for Predicting Political Activity Online .................... 105
Table 21: Which Model Best Predicts Political Activity Online? .................................. 107

viii

Table 22: Regression Results for Predicting Online Political Information Seeking. . ...108
Table 23: Which Model Best Predicts Online Political Information Seeking? .............. 109
Table 24: Regression Results: Self-Construal and Political Interest ....................... 110

Table 25: Regression Results: Self—Construal and Strength of Political Identity ........ 111

Table 26: Correlation Matrix of Important Chapter 3 Variables ............................ 112
Table 27: Factor Analysis Results for the Variable "Cyber Balkanization" ............... 140
Table 28: Rotated Component Matrix for Cyber Balkanization ............................. 141
Table 29: Descriptive Statistics for both measures of Cyber Balkanization. . . . . .....141

Table 30: Predicted Independent Variable Relationships to Cyber Balkanization ....... 142
Table 31: Regression Results: Predicting Cyber Balkanization as defined by Actions..145

Table 32: Regression Results: Predicting Cyber Balkanization as idea of Community..148

ix

M
An introduction and literature review of political computer-mediated

communication in the United States.

Abstract

This chapter is intended to illustrate the overall layout of this dissertation and
discuss the important issues which will be addressed. The author will explore what
important questions have not been properly addressed in the current literature regarding
online politics. An appropriate research plan in order to fill many of these gaps in the
study of computer—mediated political communication is then outlined.
Introduction

Technology and the media are having profound effects on the social worlds of
people in a wide variety of ways, including new information technologies such as the
intemet. Whether one encounters its presence at work, in the home, at school, or in a
local coffeehouse, virtually everyone seems to be connected to the intemet to the point
where escaping its grasp for just one day seems increasingly improbable. People have
come to use the internet for almost every facet of daily life, whether the task involves
conducting everyday business transactions, finding a date, keeping in touch with family
and fiiends, or to seek out news in order to keep up with what is happening in the world.

As the intemet becomes more common to everyday life, questions about its effect
on the social life of its users have surfaced. These concerns serve as the starting points for
the main research questions I intend to address in my dissertation. While there are an
endless number of areas which the intemet might affect interpersonal relationships and
intrapersonal attitudes, I will focus on the area of US politics. My main concern will be

on the relationship between a person’s political attitudes and self-concept, and how those

factors may influence how a person may approach the intemet to generate, support, or
possibly challenge these characteristics of the self.

This study will incorporate the social psychological literature on self and identity
with what others in political science and computer-mediated communication have said
about the effects of the intemet on contemporary political life in the US. In particular, my
objective is to address three core issues:

1. How does the self-concept mediate the ways in which someone uses

the intemet, and what implications do these findings hold in regards to

political activity and discourse in the US?

2. Are people willing to use the intemet as a communications vehicle

for challenging their viewpoints and gaining exposure to a greater

variety of informational sources, or do people use it to reinforce

previously held beliefs and make them more hostile to opposing

viewpoints?

3. Do politically partisan blogs and candidate e-mail messages differ

significantly from the content of mainstream network news coverage

in terms of their frameworks and definitions of the candidates
themselves, and if so, in what ways?

A Review of Previous Online US Presidential Races

The last years of the twentieth century saw the intemet grow in importance within
US politics. In the 1996 US presidential election, both President Bill Clinton and
Republican challenger Bob Dole launched their own websites aimed at disseminating
campaign and candidate information directly to intemet users. This was the first US
presidential election in which both major party candidates used an official website (Klotz
2004), and the volume of interactions was extraordinary. Since then, the utility and
popularity of campaign websites has expanded exponentially, as every major party
politician running for the US Presidency now uses websites hoping to accomplish

multiple aims: To raise money, to gain volunteers, and of course to offer information to

visitors. During the 2004 Presidential Campaign, 63 million Americans consumed news
online, which also includes 4 million people donating money to a campaign website, 7
million signing up to receive a candidate listserv email, and 34 million who used the
intemet to research where candidates stand on the issues (Rainie, Cornfield, & Horrigan,
2005). Politicians understood the value of being able to transcend the limitations of
coverage by media organizations and embraced the chance to speak in an unedited
fashion to a large audience. Research has found that Presidential candidates generally use
the intemet to serve four main functions: to inform the public, to connect with voters, to
get people involved in the campaign, and to mobilize their supporters (Schneider & Foot
2006).

However, the politicians themselves are certainly not the only ones who seek
these advantages. In the years since the 1996 US election, the rise of political blogs, chat
rooms, list-serve e-mails and message boards have given any US citizen with access to
the intemet a chance to both send and receive enormous amounts of political information
to other intemet users. Even the media organizations themselves run websites which are
updated constantly, continuing to have a relevant voice in the political discussions taking
place both on-line and off-line.

In the years since, candidates have learned how to meet an array of different
campaign needs via the Internet through a variety of trial-and-error methods (Kamarck
1999; Foot & Schneider 2002; D’Alessio 2000). However, the works of Schneider and
Foot (2006) indicate that by the 2004 Presidential election, all the presidential candidates
were actively using the Internet for the purposes of informing the public and involving

people with the campaign. Others have shown how politicians such as Howard Dean used

a combination of online communication techniques to develop communities based on
shared support of his candidacy, helping him quickly emerge as a leading candidate
during the Democratic Presidential primaries of 2004 (Davis 2005); practices that were
quickly adopted by his competitors in order to match his ability to mobilize volunteers
and raise money (Hull 2006; Davis 2005). In particular, he had found a new way to make
others feel a strong personal connection to their candidate, stunning political pundits with
the high levels of personal connection and commitment made by his volunteers (Stromer-
Galley and Baker 2006), which is the kind of interactivity all politicians seek in their
online campaigns (Klotz 2004)

Before the 2004 general election started, it was clear to all that the Internet was a
space where candidates should be constructing identities and meaning while developing
frameworks around personal character and issues, in addition to meeting basic
organizational functions such as raising money. For better or worse, the Internet had
become a space where attacks (accurate or not) were launched, daily blogging had
enhanced feelings of personal attachment to candidates, and sophisticated forms of
influence became major strategic focal points in a highly combative election (Williams
2006; Trammell 2006; Stokes 2006). Today, all serious candidates for the US Presidency
implement campaign strategies via the Internet while their supporters (affiliated and
unaffiliated with the official campaign) circulate online to give their candidate a web
presence.

The latter point above is an important one; that is, the politicians themselves are
not the only ones who have access to online communication devices. The Internet is a

place where politicians, media organizations, various political organizations of all types,

 

and millions of individual users simultaneously co-exist with one another. At their best,
these users engage in well-meaning political discourse with others, seeking to influence
the beliefs of others and learning from others to become better informed citizens. Users
even find the intemet a place where they can volunteer for political activism with others
who share a common identity or interest. At their worst, users embark hate-filled rhetoric,
intolerance of opposing viewpoints, and spread misinformation, retarding political
discourse.

Not surprisingly, the existence of such an environment has drawn the interests of
social scientists of all kinds over the past decade. In sum, researchers have disagreed
upon two major questions. First, does the Internet make people more or less politically
active and informed? Second, does the Internet integrate people in society into an
increasingly unified ‘commonality’ global village of perspectives, or does the Internet
lead people to become bitterly divided into partisan camps where they only interact with
like-minded people?

Important Questions Still Unresolved

As debates about how the Internet affects political discourse have surfaced in
recent years, relatively less attention has been paid to the social psychology of the
Internet user. Theorists and researchers have mainly attempted to categorize the Internet
as ‘good’ or ‘bad’ for the public at the aggregate level. However, the domain of analysis
in this dissertation mainly takes place at the individual level, focusing on which traits and
beliefs lead people the use the Internet in differing ways.

The work here recognizes the Internet user as capable of selecting and rejecting

the messages they receive. Additionally, the online users are capable of challenging and

transforming the meanings of messages that they receive, making discussion of overall
aggregate effects much more complicated. However, the Internet offers the unique ability
for people to create their own messages and distribute them to a large audience
themselves, bypassing most forms of gate-keeping and censorship from others. Within
this dissertation, attention will focus upon which ways dzfikrent sources of information
described the 2008 presidential contest, while illustrating specific examples of how
meaning was constructed in dWring ways. Additionally, statistical analysis will be used
to determine whether these dzflerences in campaign coverage by source were significant
or not. The results will be tied into theoretical perspectives about online politics that have
developed in recent years which often speculate why these differences may be taking
place.

By incorporating what social psychologists have uncovered regarding the self-
concept and social identities. Researchers have often overlooked how Internet users have
many important self-definitions and motivations that they bring with them when the sit
down at their computer monitor, and there are likely to be great variations between one
user and the next in regards to how and why the Internet is used. As this literature is
further incorporated into the study of Computer-Mediated Communication, discussion
can turn to which characteristics are important motivates that explain why people act as
they do online, and how this overlaps to what the self experiences in their ofiline social
life.

Furthermore, the relationship between the public and politicians has been
transformed by new communicative technologies. Political discourse today takes place

within more two-way communicative processes, with more interaction between message

senders and receivers taking place than in previous eras. The presence of the Internet
creates the opportunity for users to feel an increased sense of personal connection to the
candidates, and immerse themselves in political news from sources of their own choosing
at high rates of consumption. Additionally, the Internet user can even become a producer
of political news and experience daily contacts with the political candidates at all times of
the day or night, with few restrictions. Continuing to focus on individuals, it is important
to see which specific identities and social attitudes may correlate with dijflrent patterns
of news exposure and interpersonal communication when one uses the Internet.

Moreover, it may be that some people rarely or never use the Internet to keep up
with political issues, using this technology as an enabler which helps them ‘drop out’ of
political discussion in the pursuit of endeavors that he/she finds more interesting and
important. The Internet did not make a person politically apathetic, but the user’s apathy
caused him/her to avoid politics altogether, investing their personal time and energy
elsewhere. While some people may feel more informed and personally connected to
politics because of the Internet, others are likely to become more distant and less
informed than ever before as well, retreating into their own personalized life-world when
they go online. This dissertation aims to uncover which people are likely to become more
involved and which ones become less involved, and identifi» important attitudes and
identities which are responsible for these difi'krences.

Those with Internet access in their homes now have the opportunity to read an
enormous amount of information about political candidates and issues, which would
seem to be a welcome device for any functioning democracy. However, these benefits

may come with unforeseen costs; as this massive expansion of political communication

may motivate the self to engage in selective exposure and self-selection in order to avoid
feeling overwhelmed and confused. Additionally, the public run a risk of over-exposure
to politics, which can cause people to lose a sense of efficacy and trust in government
institutions and the news media themselves. Within this dissertation, statistical
comparisons between difirerent {mes of people will be made in order to find if high levels
of Internet news consumption correlates with feelings of distrust towards these
institutions.

These premises and research questions are the primary focus of this dissertation,
which will address all of these concerns in a five chapter format. The remainder of this
chapter will illustrate how I chose to test and measure these questions, and validate my
choices of methodologies. Chapters two through four will be devoted to resolving these
questions through empirical testing and discussion. Finally, chapter five will provide a
summary of key findings, discuss any relevant shortcoming which may occur, and
provide insight and direction in regards to where future research should be directed.

An Overview of this Dissertation

I plan to address a number of different research questions in three separate
chapters within this dissertation. Each chapter will focus on the effects of computer-
mediated political communication on the 2008 US presidential election. All chapters will
seek to incorporate self-concept literature, personal identities, media flames, social
psychology, and social attitude theories into the study of intemet usage and its relation to
political involvement and activity.

My second chapter will focus upon the content of intemet news sources and the

way they shape political discourse during the 2008 presidential election in the US. This

chapter is intended to both describe online political discourse as well as finding whether
different news sources qualitatively cover major political news topics and candidates in
significantly different ways. The main hypothesis of this chapter is to find out whether
various online news sources differ in content, and if so, in what ways? What kind of
assumptions do they make about the US political landscape? What do these writers think
should be of primary concern to their readers which helps to set the agenda for the
presidential election?

Chapter two will address how media frames are constructed and maintained
during the election: What specific words/phrases became popular during the most heated
weeks of the campaign, what labels were attached to the candidates and their policies,
and to what degree did the political blogs echo what was being reported on the websites
of independent news sources? Also, were the political parties and candidates effective at
defrning themselves and the public agenda on their own terms and phrases, or did other
sources of information on the web reject and/or modify them, leading to a redefinition of
the candidates, their policies, and the national agenda?

Overall, this chapter will employ both quantitative and qualitative analyses to
resolve my research questions. In regards to the former, the recording of data into
categorical measures will permit the usage of chi-square analysis in order to determine
whether different online news venues vary significantly in terms of how they cover the
presidential election and define the candidates. These values can be used to illustrate
whether different intemet news providers are showing meaningful differences in what

frameworks (Lakoff 2004; Gamson & Modigliani 1989) they employ, whether partisan

sites use more inflammatory language than other sites, and whether those who read these
sites engage in productive political dialogue with each other or not.

While an exhaustive analysis of all online political communication content is
virtually impossible to do (Jones 1999), my plan is to focus on a variety of highly
relevant and heavily-viewed sources that are the most likely to impact large groups of
people. The next section contains a list of forty specific intemet sites that will be
randomly selected on a daily basis for content analysis. These websites include a variety
of different writer and user characteristics, which should add validity and make the
findings more applicable to a larger group of intemet message users. The findings in
chapter four will incorporate the intemet postings of all types of message producers: the
candidates themselves. their political party, other major figures within the respective
parties, various independent media organizations, partisan political writers, and even the
general public itself to see what kinds of messages are being produced and how they had
a material impact upon the 2008 US Presidential campaign.

Chapter three will integrate important self-construal and identity theories from to
examine whether the internet makes people more or less politically informed and active.
While some have suggested that the presence of the intemet increases political activity
and knowledge among its users, positively benefitting communities (Rheingold 1995;
Haythomthwaite & Welhnan 2002; Weinberger 2005). Others have been concerned
about the negative effects of the intemet (and other communicative technologies) on
groups of people at a societal level (Putnam 2000; Margolis & Resnick 2000; Galston

2005; Keen 2007).

My hypotheses are similar to what Eric Uslaner (2004) has claimed: the Internet

10

does not make a person more or less community oriented or political all by itself. Rather,
people bring their offline selves to the intemet, and then engage in user-specific driven
uses and gratifications online (McLeod & Becker 1981). More specifically, how a person
uses the internet in terms of political activity and information is influenced by both their
political identities and their overall self-concept.

Within this chapter, my aim is to show that levels of political activity and political
knowledge are not dependent upon the amount of time spent online, but rather by a
mixture of the user’s identities and self-construal. I am interested in how people have a
tendency to see themselves overall: in individualist terms or interdependently to others.
My interest is to see whether people tend to think of themselves within mainly self-
referent individualist attributes (such as thinking that I am smart or frmny, for example)
or within some collective relationship to others (such as being an American or a
Republican, for example). In order to do this, I will use elements from the twenty-four
question Singelis (1994) self-construal scale. Twelve questions from this scale will be
used to determine how the survey respondents predominantly see themselves, and then
classifying the responses as being either self-referent or other-referent in their nature.

By using this instrument, I will be able to classify respondents along a continuous
individualist-collectivist measure of the self, where the respondents range between
holding an exclusive preponderance to see oneself as a member of larger social groups (a
“we” orientation) versus those who see themselves wholly within idiosyncratic personal
qualities (a “me” orientation). These kinds of approaches have been theoretically
supported by other scholars of the self-concept, who have noted that the self operates

within a continuum of possible identities (Hewitt 1989) and locates social and personal

11

identities between polar opposites which pull in competing directions (Brewer and
Gardner 1996), yet producing a measurable and relatively stable self-concept amenable to
quantitative measurement techniques (Kuhn and McPartland 1954; Stryker 1980; Serpe
1987; Sedikedes and Brewer 2001; Prentice 2001; Rosenberg 1979; Wright, Aron &
Tropp 2002). Previous research has explored, for example, how this balancing act can be
affected by culture (Markus and Kitayama 1991; Kashima et al 1997; Roberts & Helson
1997) and by time period (Zurcher 1977; Baumeister 1986; Gergen 1991; Demo 1992).

This dissertation embraces the self-construal scale (Singelis 1994) as an
admittedly crude but useful tool in measuring tendencies of the self to think and act in a
way that either leans towards individuality or through relational/collectivist aspects of the
self. This instrument is not designed to explore the vast complexities of the self or how it
changes from one situation to the next, nor does it refirte that people contain a virtually
endless amount of identities within their self-concept. Instead, it is a pragmatic tool
where respondents describe themselves mainly in general tendencies and understandings.
Relationships between self-construal and online politics are not expected to explain all
the complexities between competing identities and political behavior. Rather the findings
will serve as a starting point to examine how the self relates to online politics and should
lead to future research which will motivate and inspire future research bridging these two
fields of study.

There are some who question whether this is the best measure of the self-concept
(Cramer & Grace 2003) and the validity of self-construal scales (Levine et al 2003).
However, the Singelis (1994) self-construal scale will not be applied to measuring

individuals fi'om cultures whose values, customs and daily practices differ substantially

12

from the United States, reducing these questions of validity for the project being outlined
here. However, in order to address these criticisms, the index used will be subjected to
factor analysis with full confidence in the validity of this measurement instrument.

Nevertheless, the use of the self-construal scale has become commonplace within
social psychology in recent years. It has surfaced in important current research involving
a wide range of topics, from emotional management (Nezlek et al 2008), interpersonal
relationships (Lydon et a1 2008) Social Networks (Yeung et al 2008), and perceptions of
organizational fairness (Holmvall & Bobocel 2008) just to name a few. While these
measures will be somewhat crude and not intended to fully illustrate the full complexity
of the self, they will suffice for the purpose of categorization of survey respondents. By
scoring respondents along this range, we can see both if different users will use the
intemet to meet their political interests and personal motivations.

What is predicted is that each group will differ significantly in terms of whether
they use the intemet to become more politically informed and active. Instead of simply
predicting that those who use the intemet will be changed by it in terms of political
activity and knowledge, these categories incorporate the “uses and gratifications theory”
(McLeod & Becker 1981) of media audiences, acknowledging that different people will
be motivated to use the intemet in different ways. Specifically, I predict the following
hypotheses to be true:

H1. Those that have strong political identities will be more likely than others to
become politically active and on the Internet. --- People with strong social identities
rooted in political interests will meet those needs to become politically active by using

the Internet more than those who do not.

13

H2. Those that have strong political identities will be more likely than others to seek
political information on the Internet. --- This is also consistent with Uses and
Gratifications theory. People who have strong social identities rooted in political interests
will search the Internet for political information at higher rates than those who lack strong

beliefs about politics.

H3. Those who are interested in politics will be more likely than others to become
politically naive and on the Internet. --- Consistent with Uses and Gratifications theory,
people which are interested in politics will show higher levels of online political activity

that other users will.

H4. Those who are interested in politics will be more likely than others to seek political
information on the Internet. -- Also consistent with Uses and Gratifications theory,
people which are interested in politics will show the highest amount of time seeking out

information on the Internet.

H5. There will be a Positive relationship between Self-Construal and an interest in
politics. --- People who have a propensity to incorporate important others into their own
self-concept will be more likely to take an interest in political issues. Those who see
themselves as unique individuals may have less interest in pursuing collectivist needs via

political movements.

14

H6. There will be a Positive relationship between Self-Construal and holding strong
political identities. --- Since collectivists may be more likely to take an interest in political
issues, they are also more likely than individuals to develop stronger and more intense
political beliefs than others.

Overall, chapter three will focus on how internet usage varies depending on user
characteristics and self-concept. It is not enough to say that the intemet makes people
more or less knowledgeable or politically active. Instead, the user brings an offline self
when using the intemet. These factors then mediate the effects of the intemet on politics
in the US. By using a survey which encapsulates all of these factors and measures of
political activity both online and offline, quantitative measures of association can be
found to indicate whether these personal characteristics have an effect on politics beyond
more exposure to the intemet.

Chapter four contains hypotheses which will attempt to uncover which types of
intemet users are more likely to behave like “Cyber Balkans” (Van Alstyne &
Brynjolfsson 1997) and if these users are similar to “Single-Issue” users (Bimber 1998).
In other words, what leads a person to use the Internet in a way that reinforces their
socio-political identities and what personal needs are these users satisfying with their
patterns of online behavior? Are those needs similar to other types of Internet users?

Chapter four will attempt to understand which variables are related to the desire
of the individual user to act like a “Cyber Balkan” in their online political activities, and
whether there are any similarities in the characteristics of these users to accelerated
pluralists, who invest the vast majority of attention towards one topic that is important to

them. This often has the effect of making a person an expert on one topic at the expense

15

of lacking knowledge about other issues.

Bruce Birnber (1998) points out that there are at least two identifiable traits of
Internet users that lead towards single-issue predominance in usage, or an accelerated
pluralist usage pattern:

1. They seek information only for issues that they find personally
interesting.

2. They are likely to have weaker political affiliations to a specific
party because their political orientation is more issue-based than
ideological.
Likewise, Van Alstyne & Brynjolfsson (1997) identify at least five traits that are

common practices of a cyber Balkan, which includes the following:

1. They tend to define their “community” by those who share their
values rather than those who share their geographical proximity.

2. They have a desire to be surrounded by like-minded individuals
and/or avoid people who disagree with them.

3. They feel uneasy when they are exposed to an oppositional point of
view.

4. They tend to believe that a person should leave a community/ group
if they do not share the views of the other members.

5. They tend to hold more extreme political identities, as they are more
likely to receive constant reinforcement from like-minded others.

Overall, I have two main objectives in chapter four. One of those objectives
involves a hypothesis, and the other seeks to uncover the antecedents of cyber Balkan
patterns of online political computer-mediated communication. My first hypothesis
within this chapter predicts there is a correlation between the core identifying features
of cyber Balkans with the traits Bimber believes lead to accelerated pluralism. It will

be argued that these two groups of people are likely to have similarities in their patterns

16

“an—h al-— A

of online behavior and may not be entirely mutually exclusive categories.

My second objective in this chapter is to identify specific Attitudes, Beliefs, and
Motivations which may explain why someone becomes a cyber Balkan. In doing so, this
dissertation hopes to develop a more integrated and sophisticated view regarding why
people are interested to use the intemet to pursue narrow interest deeply while largely
ignoring other political issues, and to figure out what the common characteristics of users
that engage in these patterns of intemet usage are.

Chapter four will try to predict the levels of cyber Balkanization of intemet users
using linear regression techniques with data collected through an online survey. A
number of survey questions will be developed in order to capture “Cyber Balkan” as an
index measure which contains the five dimensions which are outlined above, then regress
that variable against a variety of independent variables that capture a wide array of user
characteristics. The specifics of the independent variable measures will be discussed in
more detail in the next section.

There are a series of independent variables that are predicted to have a significant
relationship to the dependent variable of cyber Balkanization. As mentioned earlier, it is
believed that these concepts are both a function of three main independent variables: user
attitudes, beliefs, and personal motivations. Each of these three independent variables can
be firrtlrer subdivided into the following (the predicted relationships between these
variables and cyber Balkanization are indicated as + or - in parenthesis):

Attitudes & Beliefs (7): Trust in the News Media (-), Trust in the

Government (-), Trust in Other People (-), Perceptions that the user

has Like-Minded Friends/Family (+), and lives in a Like-Minded

Community (+), the Strength of their Political Identity (+) If the user is
politically independent (-)

l7

W (5): User enjoys engaging in arguments (-), User wants to

avoid criticism when expressing viewpoints (+), User feels anger

towards political opponents (+), User desires bolstering/ support fiom

others online (+), User has an interdependent self-construal (+)

Overall, there are twelve independent variables which will be used to try to
predict the desire to behave online in a cyber Balkan fashion. Additionally, the
demographics of age, race, education, and gender will also be included for two purposes:
to see if there are any intervening/interacting macro-sociological variables that are of
interest, and (2) to see if cyber Balkan online behavior are related to any offline trends in
the US public that are relevant only to certain age groups, racial groups, etc...

Several attitude and belief variables were added to see how offline characteristics
of the user’s life are related to their online behaviors. There are three variables that deal
with trust: in the news media, in the government, and in other people in general. My aim
is to see whether there is a connection to align oneself to self-selected communities and
selective exposure patterns to online information has anything to do with a disregard and
distrust towards relevant institutions and the general public. It is predicted that there will
be a negative relationship here: that a lack of trust in the government, the news media,
and the general public causes one to channel their behavior to a narrow group of self-
selected similar others.

Likewise, I predict that there are correlations between offline realities and online
actions; that there is probably a positive correlation to surrounding oneself to similar
others both online and offline. As James House (1992) points out, the traits of a person’s
immediate environment impact their attitudes and beliefs a great deal, which he refers to

as the “proximity principle”. Others have pointed out (Quan-Haase et al 2002; Hampton

& Welhnan 2002, Haythornthwaite & Welhnan 2002) that online and offline behaviors

18

tend to complement one another instead of being mutually exclusive of each other. That
being said, I have included three measures to capture the homogeneity of a user’s local
environment: asking about the perceptions of how much the user thinks their views are
the same as their fi'iends, family, and their local community. These measures are
predicted to have a positive relationship to cyber Balkanization.

Next, I included variables that seek out relationships between a person’s political
values and their tendencies to act in a self-selecting way. First, I want to know if
Independents are less likely to engage in biased exposure patterns on the intemet than
Republicans or Democrats. It is predicted that self-identified partisans will be more likely
to have a positive relationship to online Balkanization. In addition to this, the strength of
the user’s political affiliations are predicted to have a positive relation to the dependent
variable, as those with strong attitudes/beliefs are less likely to seek out discrepant
information.

Lastly, I want to see if there are specific motivations that might drive intemet
users to act in ways that meet the needs of the self. As theorists have pointed out about

the functions which attitudes serve (Katz 1960; Herek 1986; Shavitt & Nelson 2000), it

has been found that people hold attitudes for any of the following four functional reasons:

to express their values, to make sense of information, to defend their self-esteem, or to
accomplish goals. With this in mind, I chose five motivations which comply with these
functions in examining why an intemet user may choose to act in a self-selecting way to
online information and communities.

In regards to expression of values, it may be the case that (1) the user enjoys

engaging in arguments, and (2) also feels anger/hostile feelings about political

l9

opposition. It is predicted that cyber Balkans will be relatively low on the first trait and
relatively high on the second. The prediction is that a cyber Balkan does not enjoy being
challenged in their political views, nor do they wish to spend time/energy trying to
convert their opponents. Instead, they prefer “preaching to the chorus” while likely
gaining reinforcement and a sense of personal connection to those who feel similar to
themselves.

Additionally, the need to bolster self-esteem can be another key motivation to
becoming a cyber Balkan. It is predicted that self-selecting political activity online is
positively related to those who have a strong desire for acceptance of their beliefs and
feel discomforted by criticism of their views. These people also want to find a feeling of
connection to similar others, as the user seeks a sense of community, belonging, and
reinforcement by contacting like-minded others on the Internet. Moreover, the need to
belong will be strongest for users whose self—construal is more interdependent, as these
people exhibit a stronger desire to connect their political actions/ideals to similar others.

Overall, chapter four is geared to finding two things. The first is to find whether
there are meaningful correlations between the tendencies of cyber Balkanization with
traits that are more commonly associated with accelerated pluralism. In addition to this,
confirmatory factor analysis can be conducted on this measure of Balkanization in order
to discover if this concept consists of multiple dimensions. Second, this chapter focuses
exclusively upon what offline personal aspects of internet users are more or less likely to
drive them towards Balkanizing behavior online. Both of these inquiries will be
addressed though quantitative measures using an online survey. The specifics of how

each variable will be measure will be discussed in detail within the methods and

20

measurements section, along with a discussion of how the online survey will sample and
be distributed to people who are engaged in politics online in the US.

In conclusion, chapter five will summarize the findings of this dissertation.
Within this final chapter, the author will also explore the possible shortcomings that may
exist in either the data or methods. This section will include a disclosure of any issues
that emerged during the data collection process, as well as how the author resolved these
matters. Lastly, discussion will be centered on what these findings imply for future
researchers of computer-mediated political communication, and recommendations will be
made regarding where future research should be directed.

Summary

This dissertation intends to illustrate three important relationships between
computer-mediated political communication and social psychology. First, that the
presence of the Internet greatly transforms political communication in a wide variety of
ways. Not least of which is the blurring of the boundaries between audiences, journalists,
and newsmakers taking place every day in online political news reporting.

Second, I intend to show that the efifzcts of online political communication are not
uniform among its users. Rather, the identities and self-construal of the Internet user
greatly impact whether the person become more or less involved and informed about
politics. Irrespective of the time spent online, some types of users will become more
politically oriented, some other types will exhibit no interest in politics, and there will
even be people who show interest in politics without increased levels of activity.

Third, it will be shown that Balkanizing behavior, self-selectivity and highly

polarized political attitudes are not unrelated to oflline behavior. Instead, there are

21

correlations to online and oflline behaviors and attitudes. Furthermore, some of the
motivations and offline living conditions of Cyber Balkans will be identified, with the
hopes that more traits will be uncovered by future research.

Chapters three and four will focus exclusively upon the Internet user, and how
their personal characteristics are related to the political messages they receive online. But,
before approaching this topic, an analysis on the sources of online politics during the
2008 US Presidential election needs to be addressed in order to appropriately
understanding the nature of online political discourse. This exploration will take place

next in chapter two.

22

W
Political News Sources on the Internet: Similarities and Differences during the US
Presidential Election of 2008.
Abstract

The purpose of this chapter is to explore how online media sources helped to
create imagery and identities around the Barack Obama and John McCain campaigns
during the 2008 US presidential election. A literature review covering the most common
types of online news sources was conducted and discussed. Finally, a content analysis of
537 news postings conducted over the final six weeks of the Presidential campaign will
uncover how the race was covered differently by a variety of online news sources.
Introduction

Every year more Americans go online to meet a variety of personal interests and
accomplish a wide range of tasks. In 2008, the US Presidential election created a great
deal of interest and participation among the public, as the Internet became an important
marketplace of information and locale of political interactivity. In this chapter, I will
focus on those who supplied the Internet with political news. For example, who created
imagery around the candidates, how did different sources cover the major issues of the
election, and what did these writers define as important?

The current time period sees a massive transformation taking place, aided by
modern technological advances, that has the effect of blurring the lines between audience,
source, and reporter which is often referred to as “convergence” (Pavlik 2008; Dessauer
2004). The US is now seeing the rise of “Produsage” (Bruns & Jacobs 2006: 6), where
people are simultaneously becoming the producers and consumers of knowledge. This is

quickly bringing an end to the traditional “top-down” news broadcasting (Gillmor 2006)

23

that defined political communication from most of the 20th century (Pavlik 2008). Today, i
the presence of the Internet has empowered the mass audience and politicians alike. This
allows these individuals to speak directly to millions of users without the filter of a media
organization to edit or limit what they want to say.

The 2008 Presidential election in the US involved the co-existence of professional
journalists, newsmakers, and mass-mediated audience members simultaneously
attempting to provide interpretation and commentary to politics online. Compared to any
other Presidential election, the US public saw a more decentralized and multi-perspective
public sphere. However, this is not to say that the content being presented is always of the
highest quality, nor is the public flee from bias and manipulation. Rather, we see that
different types of online news sources often vary greatly from one another in a variety of
important ways, which will be illustrated here.

Democracy and the Public Sphere

An ideal democracy requires communication and debate among an informed
citizenry that takes place within a public space or fonrm. Jurgen Haberrnas (1989) defines
the public sphere as “a network of communicating information and points of view” that
gradually transforms public opinion. . ...must be accessible and inclusive (Pavlik 2008:
96). The Public Sphere refers to the set of practices that members of a society use to
communicate about matters they understand to be of a public concern. It is “a
sociological term for signifying how people, if at all, speak to each other about what their
condition is and what they ought or ought not to do as a political unit” (Benkler 2006:
178). As Cass Sunstein (2007: 25-27) explains, a well-frmctioning public sphere should

contain at least these three important qualities:

24

1. Speakers have access to a wide array of people

2. Speakers have access to voice their concerns to both institutions
as well as to a heterogeneous public.

3. People will be exposed to a wide variety of people and views,

ensuring both widely shared experiences and exposure to diverse

views.

According to Habermas (1989) the ideal-type of rational debate and discourse in
society has been limited by the existence of corporate agendas and the concentration of
media ownership in the hands of a wealthy few. This situation has limited the pursuit of
truth, has marginalized a wide array of perspectives, and led to a decline in citizen
interest and participation. For much of the twentieth century, the public sphere was a
system dominated by one-way communication tools. Public opinion was said to be
manufactured among passive consumers of news messages artificially by those in power
of the media (Saco 2002). Information sources within the public sphere in all advanced
modern societies were heavily influenced by the owners of those structures. Therefore,
whether it was the government or private enterprise, it was primarily a one-way
technological “hub-and-spoke” model kind of structure that lacked feedback loops and
was prohibitively high in cost for most people (Benkler 2006: 175). This structure led to a
relatively passive news consumer (Benkler 2006; Saco 2002; McChesney 2004; Dessauer
2004) that received finished goods rather than being part of a conversation. The public
sphere was limited in the number of perspectives being presented (Gans 2003), leading to
unreported facts and absence of multiple viewpoints.

However, at the beginning of the twenty-first century, statements coming from the

media can now be seen as invitations for a conversation, particularly on the Internet.

Jurgen Habermas (1989) sees technological expansion as “vital to the development of a

25

meaningfirl public sphere, as innovations and advancements in technology permit new
messengers into the public reahn. This empowers the masses and providing avenues to
both produce and consume information with others” (Saco 2002: 64). Key to these hopes
is that the Internet prevents any specific points of control or concentration of undue
power or influence; it is “flat” (Benkler 2006: 177). The Internet is being seen as a
corrective to the shortcoming of the public sphere, improving democracy by enabling
citizens to be exposed to more perspectives. They become more informed and
knowledgeable, challenging the concentration of power (Bahnisch 2006; Gans 2003;
Benkler 2006; Saco 2002). Even though clustering will most likely take place around
information that the public finds interesting and engaging, additional sources will still
provide a variety of opinions while also serving as a watchdog over society and its
institutions (Saco 2002: McChesney 2004, Benkler 2006; Sunstein 2007; Banisch 2006).
Moreover, the presence of the Internet greatly alters the public sphere simply by enabling
news consumers to access another news source besides print, television, and radio has
transformed the media landscape (Dessauer 2004).

Many aspects of the 20th century model of news reporting and the relations
between source-j ournalist-audience have become outdated today (Pavlik & Ross 2000).
There have long been three major kinds of participants in the world of news: Journalists,
Newsmakers, and the Audience. However, the public sphere in the US has now become a
place where “everybody has the ability to make the news. . . .and newsmakers can no
longer control the tide as easily as they once did” (Gillmor, 2006: 45). This situation is
often referred to as “convergence” (Pavlik & Ross 2000; Pavlik 2008; Gilhnor 2006)

where these three constituents overlap and influence each other in the dissemination of

26

political news in the US. Convergence refers to the blurring of the lines between
consumer and producer.

The public has become empowered and interactive, which has changed the ways
news is both produced and consumed. Many feel that journalists have sat in a privileged
position when it comes to informing the public. They feel that this has skewed the
democratic process (Gans 2003) and has served as a motivation for non-journalists to
become reporters. This serves as a challenge for the journalist to redefine the reporter-
source relationship. Any audience member or source can now contradict or challenge the
information directly to a large audience via their own computer-mediated communication
tool, whether it is through the medium of blogging, emailing, etc... (Pavlik & Ross 2000).
Journalists themselves are using the Internet to research stories, increase their speed of
communication with sources, and are even going online to get story ideas (Pavlik & Ross
2000). All of these examples are seen as evidence of convergence taking place between
news organizations, audiences, and sources.

Additionally, the public and newsmakers themselves have begun to use low-cost
technologies like emails, websites and blogs to speak directly to the public without the
filtering of news media organizations. Research has indicated that 48 million Americans
have posted content to the Internet, much of it in the form of a blog (Pavlik 2008: 75).
Although people are known to blog for a wide variety of reasons (Nardi et al 2004) many
non-journalists have now taken to blogging in order to discuss and report US political
news. They are also providing commentary of both newsmakers and the mainstream
media themselves. Some of these blogs, while still lagging behind more traditional

sources in terms of daily readership, have gained an impressively large following in a

27

short amount of time. Moreover, the most popular political bloggers are not powerless
and anonymous audience members that spout off their personal dissatisfactions into the
emptiness of cyberspace, ultimately having no effect on US politics.

A significant number of these blogs are now quite influential, affecting
politicians, political parties, legislature, public opinion and other coverage of political
news by professional news organizations. Blogs (such as the liberal huffingtonpostcdm
and the conservative drudgereport.com) now boast more than a million unique visitors
per day. Meanwhile, many other blogs on both the political left and right have seen their
readership double over the last year (Perlmutter 2008). These points will be revisited later
in this dissertation.

Convergence can also be found by the newsmakers themselves, who have turned
to the Internet as a way to directly report campaign speeches and political imagery
directly to the public without the filtering of professional joru'nalists. During the 2008
Presidential campaign, Barack Obama and John McCain both spent a considerable
amount of effort each day communicating directly to the public via Internet technology.
Their campaigns continually produced original material to the public in the form of
speeches made and blogs written by candidates themselves. As will be seen later, these
candidates were able to socially construct perceptions about themselves and their
opponents directly via a multitude of online news media websites. These messages were
placed for easily accessible consumption by millions of potential voters each day, without
any filtering fiom professional news organizations.

Before moving into my content analysis of how different news sources portrayed

the candidates and issues during the 2008 US Presidential election, a deeper discussion

28

involving the strengths and weaknesses of each participant will be presented. Although
the boundaries between these constituents continue to blur, it should be noted that the
motivations of audiences and newsmakers to use these communicative tools often stems
from a dissatisfaction with professional journalism (Kaye & Johnson 2006; Singer 2006;
Perlmutter 2008; Pavlik 2008). Politicians aim to use computer-mediated communication
tools in an attempt to set the public-agenda to their own preferences. They purposively
speak about the topics they want the US public to consider in their own language, without
the interference or judgments of professional journalists.

Meanwhile, many citizens (particularly the bloggers themselves) have a large
amount of contempt and distrust of the professional news media industry in the US.
Public opinion polling consistently shows that news media organizations are
“increasingly drawing the ire of the American public” (Klotz 2004: 128). In polling, the
public often strongly expresses the view that news media is biased against their political
views (Manjoo 2008). The public often sees journalists as too entrenched with the
political elite to ask difficult or challenging questions (Gans 2003). In sum, many people
feel highly dissatisfied with the quality and content of news reporting for a variety of
reasons that will be discussed in the upcoming section. Because of this, these people have
taken it upon themselves to act as a corrective mechanism for what they consider to be
sloppy and biased joru'nalism. As Robert McChesney (2004: 57) added:

“Democracy needs journalism to act as a watchdog of the

powerfitl, disseminate truth, and present a wide array of informed

positions on important issues but the US news media currently fails

at all three duties largely due to corporate ownership and the
pressures for profit maximization. ”

29

The current trend has led to expansion of the public sphere, and some proclaim that we
are seeing the end of “gatekeeping” with the “gushing” of information fi'om various
sources and mediums (Gillmor 2006: 46) now occurring (Benkler 2006; Margolis and
Resnick 2000). It is process which cannot be prevented by any specific political or news
organization in the US (Winograd & Hais 2008).

The Journalists

In the US, the mainstream news industry has long consisted of well-known
institutions and large corporations that produce an enormous amount of news that reaches
millions of Americans each day. In previous eras, these organizations generally
distributed their news via a specific medium, such as a newspaper or a television.
However, this is no longer the case. Instead, network television news, prestigious national
newspapers and cable television networks are also present on the worldwide web. Each
day, they are continually supplying online users with news content just as they do within
their more traditional mediums. Indeed, the mainstream news media are still playing a
major role in creating shared experiences and exposure to topics (Sunstein 2007; Davis
2005; Gillmor 2006).

Media organizations have been using the Internet to increase the overall range and
volume of news stories presented. The virtually limitless space offered by the Internet
allows journalist to report an increased the number of news stories, as well as a greater
variety of news stories (Klotz 2004). For example, news stories regarding international
issues and crime have usually been reported online rather than in print or on television. It
has been speculated that this occurs in an effort to customize these stories to niche

audiences rather than to larger heterogeneous traditional media audiences (Klotz 2004).

30

Overall, there has been a rise in demand for hybrid journalism (traditional forms
plus interactivity) that is likely to change content of news, but has been a struggle for
traditional news media organizations to design and implement (Pavlik 2008). These
problems become further exacerbated by the immediacy of the Internet, which leads news
producers to place a premium on speed. Too often, journalists are pressured to value
“getting it out” over “getting it right” (Klotz 2004: 128). The ever-presence of user
accessibility to the Internet sets an ever-increasing premium on speed for news
organizations. Today, “breaking stories” will often be reported perpetually as continuing
work-in-progress for the public, where facts can be placed and retracted as new
information comes in (Klotz 2004). This harms public confidence in the dependability of
the news stories they receive.

In general, the public requires the mainstream news media to construct a shared
focus and public salience to political issues (Sunstein 2007). Because of their extensive
informational and economic resources, these organizations have a significant advantage
in the performance of investigative journalism over all other forms of reporting (Benkler
2006). In general, news organizations and professional journalists see themselves as a
‘mirror’ of reality, which they aim to capture through the professional practices of
neutrality and objectivity (Singer 2006; Gans 2003; Schudson 2003; McChesney 2004;
Gillmor 2006).

However, there are concerns that these professional practices can not be either
objective or neutral because journalists themselves are too far removed fi'onr the citizenry
and the lives of common people (Gans 2003). Robert Park (1940) argued that news

organizations produce what it thinks audiences will be interested in. This process

31

involves the use of subjective judgments by news organizations, where content ultimately
reflects what reporters believe are the audience’s hopes and fears. More importantly,
subjective decision making could never be avoided in journalism even though it runs
counter to the professional values of objective and neutral reporting (Park 1940;
McChesney 2004; Gans 2003).

In addition, journalists lack any full measure of independence and influence.
Journalists are the employees of large multinational corporations, which inevitably
influence the production and attention to tOpics, which journalists largely refuse to
recognize (McChesney 2004). In this view, journalists yield too much power to those in
charge to determine the public agenda. Just as importantly, what they keep quiet about
too much in order to maintain working relationships with their sources over time
(McChesney 2004; Gans 2003; Bagdikian 2004).

Herbert Gans (2003) describes news organizations as being remarkably similar to
factories, where news is processed in a predictable fashion like all other manufactured
goods. News organizations place their emphasis on speed, routine, efficiency and
profitability in news production in order to maximize profit. Probing in-depth research
presents a problem for the organization due to the pressures of keeping budget low and
maintaining high levels of profitability (Gans 2003; McChesney 2004; Bagdikian 2004).
As a result, journalists tend to regard anything done by the government or other ruling
elites as legitimate news (Gans 2003; McChesney 2004). Consequently, they rarely
consider the possibility that their sources are taking advantage of the joumalist-source
relationship in order to influence the public in the desired direction.

Journalists also must maintain relationships with sources that can give them

32

consistently credible stories over time. In the world of politics, that ofien means high-
ranking political officials (Gans 2003). Therefore, journalists often develop relationships
with high-govemment officials, political insiders, and research institutions which can
make facts and figures available to the press efficiently and effectively. This group of
people is sometimes referred to as “Para-journalists” (Schudson 2003). These Para-
joumalists hold the power to create newsworthy events and often have a disproportionate
influence on what gets reported, since journalists rely on these sources for information
(Gans 2003).

US political journalists are not generally willing to challenge or antagonize their
sources for fear of losing them and access to information, which keeps them fiom a
rigorous examination of issues. Instead, journalists are much more comfortable “casting
political debate in terms of strategy and spin rather than locating facts. . .are “obsessed”
with spin and the spinsters” (McChesney 2004: 68-69). Moreover, raising questions of
truthfulness can seem awfully close to be seen by the audience as “taking sides” in a
partisan debate, “making them notoriously reluctant to use the word ‘lie’ when describing
the statements of public officials” (McChesney 2004: 82). In general, journalists are
facing a great struggle for independence from their sources due to convergence. They are
all too often embedded into the organizations that they are covering (such as Iraq war) in
ways that are harmful to the reliability of news stories (Pavlik 2008: 101) Television
news is also heavily engaged in the use of “talking heads” which comment on the news
rather than covering it. In particular, cable news networks often substitute talk, opinion,
and argument for news (Downie & Kaiser 2003).

Overall, there are three main critiques of the US news media: (1) too few

33

information collection points, leaving a multitude of viewpoints underrepresented in the
public arena: (2) Too much control has been concentrated into the owners of the media
over what is said and how it is evaluated: (3) Pressure to maximize audiences and profits
leads to shoddy reporting and infotainment (Benkler 2006: 197). The third concern was
addressed in—depth by Downie & Kaiser (2003) whose work examines in detail the
market pressures on news production. In general, the economic constraints and demands
of television news media cause them to do very little investigative journalism. Their main
intention is to supply “infotainment” to a large audience in order to maximize ratings,
rather than uncovering the causes or relevance of such events to the community. As many
local news directors point out, local television does little original reporting of significant
community issues. This happens because news producers and directors doubt that viewers
will have any interest or patience to watch longer and more complicated stories,
especially when these stories lack vivid photos or video (Downie & Kaiser 2003;
Schudson 2003).

TV news in the United States is characterized by the following elements: the
constraints of time in producing a news broadcast, a shortage of staff due to high capital
investment, the power of video and the temptation to rely on such footage, high levels of
competition to attract viewers, and ownership pressures to maintain high profit margins.
All of these elements combine in the production of news broadcasts which all too often
distort reality. National network and cable news channels have also found itself
increasingly concerned with profit margins in recent years. Before the 1980's, most
national news broadcasts were viewed by their owners as a public service to the

community, but those times have changed. Today, national news broadcasts are very

34

competitive and profit oriented (Downie & Kaiser 2003). At the same time, audience
sizes for each traditional news source “are currently dwindling” as there are so many
producers of news in the US today (Pavlik 2008: 129). This makes the competition for
viewership even stronger and accelerating the pressure to substitute infotainment over
quality information and investigative reporting.

Also, in this rapidly expanding news marketplace there are many suppliers of
political information. Advertisers and producers of news programs fear losing their thin
niche segment of the audience and slicing it too thinly. The end result is that the news is
less controversial or probing and more entertainment-oriented (Benkler 2006).

The Issue of Bias

Concerns about the organization motives of corporate media owners and the
credibility of news producers in the US often lead to accusations of media bias from both
liberals and conservatives (Dye 1990; McChesney 2004; Graber 2005; Alterrnan 2004;
Goldberg 2003; Manjoo 2008; Pavlik & Ross 2000). This has been referred to as “hostile
media phenomena” (V allone, Ross & Lepper 1985), where people believe that the news
media is biased against their own viewpoints and beliefs. Although the current work does
not attempt to resolve whether the US news media is biased, it is important to recognize
that the widespread perception of bias has affected audiences. Skepticism of messages
being received creates a desire to find “the truth” from more “credible” sources, which
has been a central motivation behind the rapid expansion of political blogging. Political
blogging is rapidly on the rise mainly because bloggers and the public generally have a
strong dislike and distrust of mainstream news media. These factors serve as a major

motivation for why many bloggers have started to cover politics (Kaye and Johnson

35

2006; Singer 2006; Perlmutter 2008; Bruns 2006). Perhaps more than any other factor,
these feelings of distrust and cynicism towards the established news media industry are
the driving motivational forces behind the growth of non-traditional news mediums, such
as the rapid growth of the blogosphere.

Some critics on the political left feel that the US news media have become
intimidated by constant accusations of bias that come fi'om the political right. In this
view, the result is a hesitancy to act as a corporate or government watchdog for fear of
further accusations (Alterrnan 2004; McChesney 2004). Instead, mainstream journalists
“have become highly cautious; attempting to do nothing that would suggest it favors
Democrats over Republicans” (McChesney 2004: 109). Additionally, they have been
accused of engaging in “he said, she said” journalism (Singer 2006: 27), where
journalists have become too willing to use anonymous sources that want to manipulate
the journalist into reporting misinformation.

Furthermore, the nature of political reporting is often geared to what will be
popular to large audiences instead of what is important fOr them to know (Schudson
2003; Gans 2003). Reports tend to lack any depth or complexity, especially on television.
While this is not the product of an ideological bias, the competitive pressures of being the
first to report, when combined with the audience demand for entertainment, create an
environment where simplicity and scandal become preferred by the network over depth
and complexity (Margolis & Resnick 2000; Mazzoleni 1987; Altheide 1985). These
pressures seem to be increasing at the present time rather than decreasing (Pavlik 2008;
Gillmor 2006).

Meanwhile, the political right has accused the mainstream news of having a

36

liberal bias, interested in promoting their agenda on a wide array of social and economic
issues (Goldberg 2003; Graber 2005; Dye 1990). These critics point out that most US
political journalists identify themselves as liberal (Dye 1990; Graber 2005). Journalists
are largely autonomous from the organization to select which stories are reported while
enjoying minimal editing or censorship while doing so. They assert that content analysis
of evening television news programs show overwhelming support for the maintenance of
the welfare state in the US (Goldberg 2003; Graber 2005; Dye 1990).

The Audience

Many of the issues discussed in the previous pages have created a demand for
audience members to challenge traditional journalism by actively selecting, reporting,
and analyzing news on their own (Lasica 2003). It even creates the possibility that people
don’t need journalists at all to obtain political news (Klotz 2004). Today, content creation
is now shifling fiom the hands of a few professional journalists to the people who make
up that society. In particular, the Internet “makes everyone free to observe, report,
question, and debate, not only in principle, but in actual capability via blogging and
email” (Benkler 2006: 272). The audience need not be limited to reading the opinions of
opinion makers and judging them in private conversations but can interact with an array
of other audience members to vote intelligently (Benkler 2006).

For the first time, news media audience members have now become the active
participants of news production (Gilhnor 2006). In particular, the Internet has helped to
create a “decentralization” of news production (Bruns 2006) which combines both one-
way and two-way forms of communication (Kaye and Johnson 2006). Production of the

news today is currently transitioning to being “more of a conversation rather than a

37

lecture” according to Dan Gillmor (Bruns 2006: 19).

Audiences are no longer passive spectators in the public sphere but can become
creators and primary subjects where individuals now can publish alongside commercial
media and government entities (Benkler 2006). Research has shown that the Internet has
made it possible for more voices to be heard, with deeper discussion regarding hard news
topics taking place online (Hamilton 2004: 214). Some view the Internet as a great
improvement to the public sphere. They point out that the Internet offers greater depth,
convenience, more involvement with audiences and instant accessibility to users,
improving American journalism as a whole (Dessauer 2004). The presence of the Internet
also allows for greater depth in storytelling, various forms of presentation, and more
news being produced for the public to consume (Pavlik & Ross 2000; Gillmor 2006).

Traditional news mediums such as newspapers and television have not been
particularly adept at allowing the audience to offer direct feedback, criticism, and re-
definition of the messages they receive (Pavlik 2008; Gillmor 2006). However, the
Internet contains virtually no legal or financial barriers to prevent any member of the US
population fiom launching their own “weblogs”, often called “blogs” for short. Blogs
could be defined as an online journal in which the most recent entries are added to the top
of the document (Klotz 2004). The rise of blogs has given voice to those who otherwise
do not hold political power and had also previously lacked the ability to disseminate facts
or opinions towards a large audience. The space where these bloggers co-exist with each
other has often been referred to as a “blogosphere”; which can be thought of as an open
forum in which information is offered, revised, extended, or refuted in an electronically

enabled and decentralized marketplace of ideas (Singer 2006). Blogs have been heralded

38

by media scholar Jay Rosen as an “extremely democratic form of Journalism” (Gillmor
2006: 29).

Research has shown that people blog to document experience, provide
commentary or opinions, self-expression, to communicate ideas, and to connect with
others (Nardi et al 2004; Klotz 2004). The practice of political blogging has placed an
emphasis on transparency and personal authenticity of the blogger in a more post-modem
approach to seeking truth (Singer 2006; Manjoo 2008). Many of these political bloggers
are creators of original news content in addition to being critics of mainstream news
reports. Blogging is often viewed as a form of freelance journalism where the restrictions
of professional objectivity are being replaced by those who use information to reinforce
particular perspectives, reject arguments being made by opponents, and to make
persuasive arguments in favor of a candidate or policy (Gillmor 2006; Lasica 2003). The
Internet also speeds up the news process. This offers the ability to link quickly to similar
stories, and can increase the range of stories being reported in the news, allows for user
interactivity with the news story (Dessauer 2004). A

Additionally, the Internet also facilitates the opportunity to access original
documentation from government and business organizations relatively easily for users.
The possibility of collecting original content and broadcasting findings offers bloggers
and reader alike to form their own news broadcasting. This completely removes large
news organizations from the process of news production. Unlike previous eras, anyone
with access to the Internet has the ability to seek out primary documents, such as
government reports, that they previously relied on journalists to interpret and report for

them. Moreover, any person who accesses these documents can also report their

39

interpretations and opinions to others by launching their own blog report. Here, audience
members are reporting political news to other audience members while completely
bypassing professional journalists altogether (Klotz 2004).
Blogging: the good, the bad, and the ugly

Political bloggers tend to pride themselves on their ability to act as “fact-
checkers” and ‘fivatchdogs” of both the government and the media (Sunstein 2007: 139;
Gillmor 2006; Perlmutter 2008). In doing so, they are able to effectively document
abuses of power and illustrate falsehoods. This process has been referred to as
“Gatewatching” by conservative InstaPundit.com blogger Chris Reynolds (Bruns 2006).
Here, the blogger places a focus on where the information is coming fiom, challenges its
quality, provides personal reaction, and then provides suggestions to audience members
in terms of what information deserves their attention. In doing so, blogs are believed to
reduce the power of the journalistic profession in regards to altering public opinion with
inaccurate or biased information (Bruns 2006), and, as Reynolds adds, “to shame people
into doing their jobs better” (Singer 2006: 28). A

The ascent of political blogging is also a direct expression of the news audiences’
desire for a wider range of available perspectives than had otherwise been previously
available through traditional media (Bruns 2006). While the relationship between
bloggers and journalists may be seen as antagonistic, their co-existence can serve as an
effective “editorial frmction” to mainstream news reporting (Benkler 2006). This serves
as an effective news supplement tool for audiences (Gillmor 2006). Some believe that
this will improve the quality of journalism, as “Journalists can now expect that someone

out there is going to fact-check just about anything they write” (Singer 2006: 28). The

40

ever-presence of blogs in the public eye forces news organizations to improve journalistic
standards and quality.

As with all other people and organizations, the news media can make mistakes,
and blogs can serve as a corrective editorial function which will improve the quality of
mainstream journalism. Instead of being an all-out rejection of mainstream news quality,
blogs can be used as “probers” or could even be viewed as “niche journalism” that acts as
a complement to traditional news sources (Gillmor 2006: 103). David Perlmutter (2008)
shows several illustrated case studies where these positive frmctions all overlap. He notes
how both conservative and liberal blogs had been able to educate audience members
about the Iraq war beyond the scope of what network news coverage has offered.
Simultaneously, these writers indicated how media reports and/or government policies
had made crucial mistakes.

Blogs also have an ability to influence mainstream news production even in the
selection of news topics, indicating that there is really a symbiotic relationship which
exists between the two (Singer 2006; Trammell and Britton 2005). Cases such as the
DailyKos.com reporting of the Blackwater Corporation’s increased involvement in Iraq
and the DrudgeReport.com coverage of Bill Clinton’s sex scandal involving Monica
Lewinsky (Klotz 2004), are examples of where bloggers were the first to break new
content, Their attention to these stories drove the mainstream networks to further
investigation. Through the efforts of bloggers, news stories such as the Dan Rather
“Memogate” scandal in 2004 and Trent Lott’s controversial racial remarks at Strom
Thurrnond’s 2002 birthday party became heavily reported on television and newspapers

(Perlmutter 2008).

41

These examples all illustrate the power of the audience, and indicate that the
consumption of mainstream news messages is no longer passive. Instead, receivers of
news messages can become very active agents that can react to, counter, and sometimes
even refute the facts presented in political news. The current time period sees a hybrid of
both the producer and consumer of political news reports (Bruns 2006). Audience
members have gained “the capability to become a key source of news stories for others,
even if many are heard only by a relatively few people” (Gillmor 2006: 139). These

6

writers hold the ability to provide “random acts of journalism” and ‘ micro-news” to a
small but passionate niche audience (Lasica 2003).

However, there are still serious concerns about blogging. Perhaps first among
these concerns involves the quality of the information reported and the lack of
professional practices from the bloggers themselves. Overall, concerns are arising “about
the credibility, reliability, accuracy and trustworthiness of these content producers exist,
however. . ...along with concerns that blog producers have no training in professional or
ethical journalistic standards” (Pavlik 2008: 116).

Additionally, the negative side of eliminating “gatekeepers” results in an
information forum filled with falsehoods and inaccuracies fi'om people who have little or
no accountability for publishing them (W achbroit 2004). Ultimately, there are no
punishments and few consequences for Spreading untruths and inaccuracies through
blogging. Because of this, news quality suffers when professional reputations are not at
stake (Sunstein 2007; Gillmor 2006). Blogs are also criticized for being an enormous
source of misinformation and personal bias (Keen 2007; Wachbroit 2004; Klotz 2004;

Bahnisch 2006). Perhaps even more troubling is that bad information is seamlessly mixed

42

in with reliable information for consumption by the online public (Wachbroit 2004;
Manjoo 2008).

Because of their open format with other audience members, blogs are especially
prone to the rapid spreading of misinformation about candidates and topics. Blogs can be
a place which fosters blunders, falsehoods and extremism (Sunstein 2007) that all too
ofien reduces political debates into “sniping, snarkiness, and spin” (Bahnisch 2006: 144).
Moreover, blogs allow writers to discuss politics in a very episodic, personalized and
anecdotal way, where political issues can “become personalized reports of self-referential
egotism” (Bahnisch 2006: 145). Political issues become peripheral to the personality of
the writers themselves (Trammell 2006; Keen 2007), falling well-short of any idealized
vision of a healthy and vibrant public sphere (Keen 2007).

Additionally, comment boards from these blogs are often guilty of spreading
bogus claims, even unintentionally, that can even be purposely spread by anonymous
readers who also suffer no consequences for making such claims (Stokes 2006). Other
errors may occur when bloggers link or base their reports to incorrect stories that are
being reported by other Internet users. This usually comes in the form of reports
generated by their other fellow bloggers (Williams 2006; Gillmor 2006). Some believe
that bloggers primarily provide “punditry” rather than true journalism by primarily
offering readers heavy doses of opinion and analysis (Bruns 2006). Others claim that
blogs largely serve as merely a new platform for rhetoric-filled political persuasion
(Bahnisch 2006).

Journalists point out that blogs rarely report original content, but simply

repackage news stories already produced by journalists, providing only personal

43

commentary and quick links to alternative perspectives for their readers. Many journalists
contend that this provides only a minimal added-value from the original stories
themselves (Singer 2006; Gillmor 2006). Because of this, bloggers often face a large
amount of severe criticism fi'om journalists. Blog writers are often ridiculed by journalists
because reports may purposely containing falsehoods, lack impartiality, and engage in
unethical and unprofessional behavior in their reporting of the news (Bruns 2006; Singer
2006; Sunstein 2007; Bahnisch 2006; Keen 2007).

Nevertheless, the Internet enables audience members to utilize a low-cost
technology in order to originate and challenge political claims in a space that is accessible
to millions of other users. This fact has not been overlooked by the major political parties
in the US, who have increasingly turned their attention to these mediums in order to
address the public on their own terms and words. The following section addresses how
politicians and parties use tools like blogs, websites, and email lists to meet their needs
and influence the US public.

The Newsmakers

Perhaps the greatest beneficiaries of new communicative technologies have been
the targets of the news themselves, in this case, the presidential candidates. Before the
creation and growth of the intemet, journalists and media organizations served as a
powerful “gatekeeper” to the candidates. During this period, there were only a limited
number of opportunities for politicians to speak directly to a national audience in their
own terms. The constraints of time, space, topic, and differing interpretations that can be
inserted by the mediator of the message can place powerful constraints and limitations

upon the goals of the candidate in some form. Now, political candidates are increasingly

44

turning to the web to communicate directly to the public without the filter of traditional
news media gatekeepers (Pavlik 2008). Traditionally powerful political groups have
increasingly been able to efficiently use these low-cost methods. They do so in order to
reach out to those who are not party members while also forming coalitions within the
party in order to meet a variety of important goals. This is sometimes referred to as “post-
bureaucratic” political campaign organization (Bimber 2003).

In previous elections, candidates and parties continually developed an
understanding of how the Internet could be exploited to better disseminate information to
audiences and improve their fundraising abilities (Bimber 2003). Research has indicated
that there are four important types of practices on US presidential campaign sites. These
functions include informing users with news, connecting the campaign to other political
actors (much as to the party or linking to important endorsers), involving people to take
actions to support the campaign (such as volunteering time or donations), and mobilizing
the user to connect with others in the campaign with the hopes they will facilitate offline
action for the campaign, such as hosting a rally (Schneider and Foot 2006).

In the 2004 US presidential race, virtually all candidates for office were using
their website to at least inform and involve candidates. Some candidates were using trial-
and-error methods to effectively improve their ability to mobilize and connect supporters.
However, all candidates were using their website to inform potential voters and enable
the user to sign up for more information or donate their personal time/money. These two
elements had already become the core components of online campaigning (Schneider and
Foot 2006; Margolis and Resnick 2000). This chapter will delve deeper into how the

campaigns utilized their informing functions during the 2008 US Presidential race.

45

Previous research by William Benoit (2007) has illustrated some general overall

tendencies in the communication patterns of political campaigns and the media in the US.

Over the past couple of decades, he found the following five propositions to be true:

1.

Discussion of personal character is more common in primary elections, while
attacks and discussion of policy become more common during general elections.

Those who trail in the polls will make more personal attacks.

News coverage tends to over-report character and polling date while under-
reporting policy.

Democrats will discuss policies more, and character less, than Republicans.

Messages from candidates will use more acclaims and fewer attacks than
messages fi'om other sources

While these points serve as an important piece in my analysis of the 2008 US

Presidential election, it is important to note that little attention was given to online

communication being created by political candidates or non-affiliated bloggers. Later in

this chapter, data will be analyzed to see if these tendencies more or less prevailed in this

most current election or if these patterns had changed.

Nevertheless, political parties have found that computer-mediated

communications can put a “human face” on an organization (Gillmor 2006). Political

campaigns used computer-mediated communication as “Parasocial interaction” to give

users “the illusion of a face-to-face relationship” (Stromer-Galley & Baker 2006: 117).

The increased personalization contains the possibility for a political message to fall

“under the radar” from politics and seem more personal, likely resulting in being more

persuasive to the reader (Trammell 2006: 135). Of course, all of this activity overlaps

simultaneously with appearances on television, radio, and links to online news sources

(Margolis & Resnick 2000). The reports stemming fi'om the campaign are often

46

distributed widely to journalists following the campaign each day. Meanwhile, campaign
websites oflen feature favorable reports towards the candidate offered by the press. This
illustrates once again the blurring of the lines between sources, newsmakers, and
audiences (Bimber & Davis 2003).
Presidential Candidates and the Presentation of Self

During the 2008 presidential election, both Barack Obama and John McCain were
able to use the Internet to talk about whatever topic they wanted to focus upon at that
given moment. They both used the Internet to define their own stances on the issues in
their own choice of words and imagery. Via the Internet, candidates were better able to
wield this completely unfiltered communication tool to directly respond to any
allegations made by their opponent. The candidates were also able to create an
impression of their opponent through their carefully crafied choices of language and
topics to flame or define their opponent as they would like their audience to view them.
Moreover, online communication reflects campaign strategy; it is unfiltered and unedited
by anyone outside of the campaign and delivered directly to the public (Bimber & Davis
2003).

In particular, email lists are a highly valued resource for campaigns (Bimber and
Davis 2003) as emails are easily forwarded to a wide audience and carry instant
credibility for campaigns. They are perhaps even more effective than campaign websites
for reaching the public (Cornfield 2004). In the final days of the 2000 US Presidential
race, George W. Bush’s campaign was sending email to over 600,000 recipients per day
(Bimber & Davis 2003: 64). Volunteers and campaign representatives urged people to

“get out the vote” for their candidate. However, campaigns have a need to be efficient in

47

email listserv connections with the public, as too many messages causes “activist fatigue”
(194) from too much “cheap-talk” (Bimber 2003). Campaigns need to utilize alternate
sources of communicating online to supporters in order to avoid “abusing the privilege”
of group email (Bimber & Davis 2003: 52).

Along with email, Presidential candidates use websites to define where they stand
and assign identities to their opponent. It is important for the website to convey both
verbal cues and non-verbal visual cues to the user (Goffman 1959) to let the viewer know
that the candidate fulfills the viewer’s expectations of “the presidential role”. Self-
Presentation on a website becomes particularly critical in a competitive environment
where the opponent is also utilizing online communication technologies to paint a portrait
of the candidate and shape voter’s perceptions (Bimber & Davis 2003). Candidates must
show that they are capable of filling the role of President because they have the personal
characteristics of trustworthiness, strength, competency, empathy, morality and
experience (Williams & Kaid 2006: Bimber & Davis 2003). The candidates use their
home page to illustrate and emphasize these qualities of their self while attempting to
undermine the user’s confidence in whether the opponent has these characteristics.

Research about the websites of George W. Bush and Al Gore during the 2000 US
presidential election showed that authenticity and personalization were seen more
favorably and memorable than other information presented. This election was really the
first where both candidates used the Internet for the purposes of finding and bolstering
support, supplying information, presenting policy stances, and discussing opponents
(Bimber & Davis 2003: 85). Remarkably, neither candidate really utilized the ability to

create identities regarding their own character and assigning definition around their

48

opponent (Margolis & Resnick 2000). Instead, these sites were fairly inflexible and
primarily involved one-way communication from sender to receiver, rarely offering new
information or effectively linking supporters to one another.

During the 2004 US Presidential race, both George W. Bush and John Kerry used
“self-referential framing” for both issues and personal attributes. They also actively
created “attribute flames” to describe the personal qualities of self and opponent
(Williams & Kaid 2006: 85-86). Additionally, almost all mentions of opponents within
these websites were negative (Trammell 2006) a trend which continued in the 2008 US
presidential election. Moreover, this race saw both candidates utilize their website in
more sophisticated ways in regards to raising money and create networking possibilities
for Internet users. Additionally, the election saw improvements in fostering more two-
way communicative tools to make the user feel more personally connected to the
candidate.

Later in this chapter, I plan to show how Barack Obama and John McCain defined
themselves and each other. In addition, comparisons of how the candidates used self-
referential and attributive flames will be made versus how other online news sources also
created imagery and interpretation of the candidates during this election. However, before
moving into my methodology and findings, some discussion should be made to define
what media flames are and why they are important to study.

Media Frames and Political Discourse
As an operational definition, one may view a frame as a cognitive map which helps a
person understand how information is to be recognized, understood, and processed

(Lakoff 2004). Media flames can be understood as “a central organizing idea for news

49

 

sir

ne‘

content that supplies a context and suggests what the issue is using selection, emphasis,
exclusion, and elaboration” (Entrnan 1993: 52). Frames can also seen as a tactic of
persuasion, used in politics in order to get the voting public to think of any given social
issue within their own language and terms. They are important because they organize
information, help us make value judgments, and set priorities within any given debate.
Moreover, people have a tendency to accept only facts which fit within the flame, as
flames are pervasive yet critical to political discussion and thought, as they structure our
ideas and our reasoning (Lakoff 2006). In a very real sense, frames are at the heart of
political debate, as those who are able to activate flames built around their own
assumptions and worldviews will control political discourse and thought (Lakoff 2004).

Still, people are not necessarily slaves to flames, even though these can be both
pervasive and powerful. Lakoff' (2006) also conceives of political discourse consisting of
differing interest groups and competitors trying to place their flames into the minds of
others. So instead of one dominant flame dominating political discourse, several
competing discourses exist simultaneously in a marketplace of ideas. The audience may
be thought of as the consumers of flames as they experience political content in the news
media, shifiing through content and experiencing a variety of views and assumptions.
“On most policy issues, there are competing packages available. . . .one can view policy
issues as, in part, a symbolic contest over which interpretation will prevail.” (Gamson &
Modigliani 1989: 2)

It is imperative that we view the media as multivalent, where flames have been
simultaneously created by a wide variety of opinion leaders, professional politicians,

news journalists, and the general American public over time. As Gamson and Modigliani

50

(1989) add, every policy issue has a culture, and there is an ongoing discourse that
evolves and changes over time. Political discourse, as well as political flames, will
change as social circumstances and political values also change. Media flames, or the
social construction of meaning, should not be thought of as inflexible. Instead, they offer
readers a way of thinking about an issue and suggest interpretation and meaning to the
reader. Both language and non-verbal communication such as photos can be used to
construct meaning around the candidates and issues. As an example of such research on
political discourse and media flames, Semetko and Valkenburg (2000) uncovered five
major flameworks that defined European political coverage in the media: Conflict,
Human-Interest, Economic Consequences, Morality, and Responsibility flames.
Methodology: the 2008 Presidential race and online news coverage

Instead of focusing upon one type of message or news source in uncovering
frames, the purpose of this study was to see if these types of sources varied significantly
flom one another in their news coverage, and if so, how they differed. For example, were
blogs more or less likely to discuss issues than the candidates or journalists? Which
topics were the most prevalent during the election, and which news sources were more or
less likely to cover particular issues? Did these stories portray the supported candidate in
a positive light, or focus mainly on defining an opponent in negative terms? Also, did the
writers of these stories allow the candidates to speak in their own words, or did they try to
put words in the candidate’s mouths and try to interpret their thoughts via the filtering of
the writer?

With these kinds of questions in mind, I set out to observe online political news

stories and postings for roughly the last forty days before the Presidential election. These

51

news stories were purposely selected flom different but very popular news sources, and I
read fourteen presidential race news stories per day: two popular leftist blogs, two
popular conservative blogs, John McCain’s website, The RNC website, Barack Obama’s
website, the DNC website, two randomly selected network TV websites (CBS, NBC,
ABC), CNN, Fox News, and a prominent national newspaper (either the New York
Times or the Washington Post). When visiting each website, I selected the first story that
I saw on the website about the 2008 US Presidential race. Although the site may have
supplied many stories at any given time, I assumed that the reader would prefer to read
the article which was the most recent or seemed to be the “headline” story. This decision
was based upon factors such as the size of the headline or whether the story placement
was made in the central area on the website, for example.

The four daily blogs represented the “audience” portrayal of that day’s political
news. The four “candidate” websites cover the perspectives of the news “source”, and all
of the others represent traditional news media organization online reports. The selection
of which specific blogs to be used depended on their popularity and stated (or assumed)
partisanship. Many of these had been previously identified by Adamic and Glance (2005)
as having a stated or otherwise easy to identify general partisan stance. In many cases,
these blogs are already well-known among the US public as having an openly stated
allegiance to a political party (such as redstate.com or the liberal DailyKos.com)
ideology, or agenda. Overall, I was able to read and encode 537 online articles during this
time period.

The purpose of this was to learn whether different sources really were

substantially different flom each other in covering the election. For example, were any

52

sources really “biased” in the sense that they are disproportionately negative to one
candidate and positive to the opponent? Were some sources more likely to talk about the
issues, while others focused primarily upon punditry and personal flaws? Were blogs
more likely to use inflammatory language and anger than the more mainstream news?
Were the candidates using their ability to talk directly to the audience to talk about
themselves, about the issues, or about their opponent? By using a closed coding system to
measure these differences, I answer those questions statistically while also qualitatively
finding the most important words/phrases being used during the campaign that were so
critical in constructing meaning during the election.

As the researcher, it was necessary for me to make a few key decisions in order to
collect the data First, I made every effort to access information flom the Internet flom a
variety of different sources at random times during the day in order to avoid coverage
errors. Second, the online user was assumed to be selecting entirely from the most
popular online news websites for information. Therefore, I selectively chose to look at
only the most popular political blogs, established mainstream news websites and official
candidate websites. There is no reasonable way to conduct a content analysis project of
all online presidential news coverage, so it became necessary to assume that a “typical”
person would seek information from the most popular and generally credible sources. A
listing of all the online news sources that were among the sampling flame can be seen in
appendix A. Third, I randomly sampled the intemet at different times of the day and night
in order to avoid coverage errors in online content. In doing so, I hoped to capture an

adequate cross-section of differing types of online users and exposure patterns.

53

Analysis and Findings
Question #1: For each article, I asked if the article was mainly about Barack

Obama/Joe Biden or John McCain/Sarah Palin? Are these candidates generally
portrayed by the media in a positive or negative framework?

Table 1 seeks to find if different online news sources were disproportionately
favorable or unfavorable to the Presidential Candidates. By categorizing each article in
terms of who is the main focus of the article and how were they portrayed, analysis of
source differences could be made. The Chi-Square flom this analysis was extremely large
(321.83, p < .000) and indicates that different sources cover the campaigns in different
lights. We can see flom table 1, that liberal blogs portrayed the McCain/Palin campaign
negatively 50 out of 79 times, while portraying the Obama/Biden camp negatively only 3
times during that same time period. Conversely, conservative bloggers illustrated the
Obama/Biden campaign negatively 51 out of 80 times, and only portraying them in a
positive light a mere four times during the race. This is not a surprise, but does indicate
that blog reports have a strongly partisan bias, and lack objectivity in their news
reporting-

Additionally, the websites of the campaigns were mostly used to attack their
opponent. Table 1 indicates that out of 78 GOP website news stories, 56 (71 percent)
articles portrayed the Obama campaign in negative terms while the remaining articles
were in praise of the McCain/Palin campaign. On the other hand, while the Democratic
Party websites portrayed McCain/Palin negatively 50.8 percent of the time, they also
spent a considerable amount of effort (44.3 percent of the articles) attempting to portray

their own candidate favorably. This indicates that while both candidates were willing to

attack their opponent via the Internet, the Obama camp utilized their website

54

considerably more often to talk about the positive aspects of the Obama/Biden ticket,

while the GOP websites seemed designed mainly for attacks.

Table 1: Candidate Framing by Different Sources .
What is overall Frame of Article? I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Veggies
or Obama/Biden Obama/Biden McCain/PalinlMcCain/Palin
Neutral Positive Negative Positive Negative Total
Media. . Cou 52 34 44 14 95 23
does orgamat'm % 21.8% 14.2% 18.4% 5.9% 39.7%[10009']
article Liberal Blog Count[ 101 14 3 2 sol 0:3
% 12.7% 17.7% 3.8%L 2.5% 63.3% 100.

? Conservative Countl 5 4 51 1. 11 9 80]
B'°9 % 6.3% 5.0% 63.8% 13.8% 11.3% 100.0%
Democratic Cou 3 27 0 Ol 31 61
Pam 5‘“ % "1 4.9% 44.3% .0% .0% 50.8% 100.0%
GOP Site Cowl 01 0 56 22 0 7

% .0% .0% 71 .8%l 28.2% .0%|100.04:|
Total Coun 70 79 154 49 185 53
% 1 13.0%! 14.7%l 28.7% 01%| 34.5% 100.0%]

 

 

Overall, the most “balanced” reporting came from the mainstream news websites,
although a disproportionate number of these articles seemed to be critical of the
McCain/Palin campaign. Although the coverage flom mainstream news media sites was
neutral 21.8 percent of the time, they were also critical of the McCain/Palin ticket 39.7
percent of the time, while being critical of the Obama/Biden ticket 18.4 percent of the
time. When these percentages are added together, we can see that the news media
organizations covered the election in negative flames 58.1 percent of the time.

Queg’on #2: What is the relationship between the media sources and whether the
candidates speak for themselves? Are the writers or opposition candidate within these
stories attempting to “speak for them” and define in their own words what the

candidate stands for? Likewise, is the candidate able to speak about their opponent
and redefine their stances in different language?

55

Next, differing sources were analyzed to see which ones were more likely than
others to allow Barack Obama to define himself and the political issues directly in his
own quotations. Table 2 indicates that Barack Obama was most likely to assert where he
stands on the topic of issue that day (whether it be personal character, policy stance,
etc...) via his own website, something he took advantage of in 27 out of 61 (44.3 percent)
of the postings. Interestingly, GOP blogs and party websites also allowed Barack Obama
to be quoted directly many times, but virtually all of those were made in order to refute
the claim he was making, or as being evidence of deficiencies (such as poor character,
relationships to ACORN, etc..). Overall, the Chi-Square statistic was significant (13.535,
p = .009) indicating that some sources were more likely to allow Barack Obama to define

important moments and issues in his own words, while others were less likely to.

Table 2: Does Source Allow Barack Obama to assert his stance in his own words?

 

 

 

 

 

 

 

 

 

 

 

Is Barack Obama directly
quoted?
[ No Yes Total 1
does article Media Organization Count 179 60 2
m“? % 74.9% 25.1% 1000::
Liberal Blog Count 65 14 71
% 82.3% 17.7% 100.0°/
Conservative Blog Count 57 23I 80!
% 71.3% 28.8% 100.0%
Democratic Party Count 34 27 61
S“ % 55.7% 44.3% 100.0%
GOP Site Count 54 24 7
% 69.2% 30.8% 100.0%
Total Count 389 148 53
°/o 72.4% 27.6% 10001:]

 

 

 

Interestingly, GOP websites and Conservative Blogs oflen took direct quotes flom
Barack Obama regarding personal character questions and policy stances, but did so

mainly to illustrate why these ideas were troublesome. So although these sites were

56

willing to quote Obama and give him voice, it was largely to serve as the beginning of an
attack, as we can see from the large amounts of negative flaming that came from these
sources in table 1 earlier.

However, John McCain did not utilize these sources the same way as Obama. As
can be seen in table 3, McCain was able to assert his own stance in only 20 out of 80 (25
percent) of conservative blog articles, and 24 out of 78 times 930.8 percent) of GOP
website stories. This is reflective of a strategy where most of the efforts flom these
writers were spent trying to assign identities and construct perceptions of Barack Obama,
and not spending as much effort attempting to illustrate or define McCain. Not
surprisingly, McCain was rarely directly quoted in the Democratic Party websites, as they

allowed him to voice his stances in his own words only 6 times in 61 articles that were

 

 

 

 

 

 

 

 

 

read, a mere 9.8 percent of the time.
Table 3: Does Source allow McCain to assert himself in his own words?
Is John McCain directly
quoted?
[ No Yes Total I
does article Media Organization Count 137 102 23
"a“? % 57.3% 42.7% 100.
Liberal Blog Count 45 34 07:]
% 57.0% 43.0% 100.
Conservative Blog Count 601 20 eel
% 75.0% 25.0% 100.0%
Democratic Party Site Count 55 6 61
% 90.2% 9.8% 100.
GOP Site Count 54 24 76'
% 69.2% 30.8% 100.0%
Total Count 351 186 53
% 65.4% 34.6%] 100.0%]

 

 

 

 

Interestingly, mainstream news organizations were very permissive to the McCain

camp, allowing him to speak in his own terms in 42.7 percent of their news stories, where

57

table 2 showed that Obama was able to do so only 25.1 percent of the time, bringing into
question whether the traditional news media really had a favorable bias to Obama and
was dismissive of McCain. Overall, these sources differed significantly in terms of how
often they allowed McCain to voice himself (Chi-Square = 29.658, p < .000)

The next part of the analysis looks at how different sources tried to describe the
presidential candidate and assign meaning and definition to the candidate that were not in
the terms of the candidate’s choosing. Table 4 shows that Obama and Biden were
particularly successful in being able to define the McCain campaign in terms of their own
words and frameworks within traditional media organization news stories, doing so in
37.7 percent of all mainstream news articles. At the same time, writers and pundits used
their own words to construct imagery or definition around McCain, something that was
done 40.2 percent of the time. Taken together, it can be seen that people other than
McCain were describing his platform and actions in their own words 77.9 percent of the
time in news media organization, which may have given some conservatives the
impression that the news reporters were working against them and showing bias.

We can also see in table 4 that Obama was active in assigning identity and flames
onto McCain within Democratic Party websites, doing so 25 times out of 61 postings, or
41 percent of the time. At the same time, liberal blog writers and pundits were defining
the McCain campaign in their own words and flameworks as well, as 46.8 percent of
these articles saw the writer or pundit within the story define McCain’s stances or
character in their own words. Again, these differences were found to be statistically
significant (Chi-Square = 131.06, p < .000), showing that writers and candidates were

able to voice critiques and redefine McCain’s stances and image in their own words

58

largely through traditional news sources, democratic party websites, and the liberal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

blogosphere.
Table 4: Which Sources are allowing others to define John McCain?
Who is assigning perception and
meaning to McCain?
Writer or Obama or
Nobody Subject Biden Total
does article Media Organization Countl ml 96 90 23
me from? o o o
/o 22.2/.. 40.2 n 37.7% 100.0%
Liberal Blog Countl 33 37 9 79'
% 41.8% 46.8% 11.4% 100.0%
Conservative Blog Countl 49 23F 81 801
% 61.3% 28.8% 10.0% 100.0%
Democrat Party Countl 19l 17 25 61
Site % 31.1% 27.9% 41.0% 100.0%
GOP Site Countl 64 14 0 7
% 82.1% 17.9% .0%l100.m:l
Total Cou 218 187 132 537
% "1 40.6% 34.8% 24.6% 100.0%I

 

 

 

 

 

Additionally, we can see in table 5 below, political writers and pundits were
disproportionately creating meaning in their own wording around Obama within
conservative Blogs (in 53.8 percent of those articles) and GOP websites (in 55.1 percent
of those stories). Meanwhile, the GOP candidates themselves were defining where
Obama stands in their own words 43.5 percent of the time in mainstream news articles,
while the writers of these articles attempted to define Obama 29.3 percent of the time.
Taken together, writers and opponents were able to try to define where Obama stands
72.8 percent of the time in traditional news organization articles online, which is similar
to the 77.9 percent of the time that writers and opponents try to shape where McCain
stands. It is likely that both Democrats and Republicans feel that the traditional news
media reporting “work against them” given these findings, but fail to recognize that these

patterns of putting words into the candidate’s mouth is something that these sources do to
59

both candidates, not only their own. Findings like this may give some justification to the
“hostile media syndrome” theory discussed earlier.

Table 5: Does the Article allow others to define where Barack Obama stands?

 

 

 

 

 

 

 

 

 

 

 

 

 

Who is assigning identity to Obama? I
Writer or I McCain or
Nobody Subject Palin Total
does article Media Organization Count 65 701 104 23
"a“? % 27.2% 29.3% 43.5% 100.le
Liberal Blog Count 51 17 11 71
% 64.6% 21.5% 13.9% 100.0%
GOP Blog Count 20 43 17 063
% 25.0% 53.8% 21.3% 100.
Democrat Party Site Count 491 11 1 61
% 80.3% 18.0% 1.6% 100.0%
GOP Site Count 15 43 20 76'
% 19.2% 55.1% 25.6%I 100.0%
Total Count 200 184 153 53
% 37.2% 34.3% 28.5% 100.07]

 

 

 

The pattern seems to show that McCain and Palin used traditional news outlets
themselves to refute the claims made by the Democratic Party nominee, and attack
Barack Obama in terms of both policy and personal character. The GOP campaign oflen
relied on a mixture of pundits and conservative bloggers to construct negative beliefs and
perceptions about Obama. These differences were statistically significant (Chi-Square =
133.605, p < .000) showing that some sources were willing to allow others to develop
meaning around Obama in their own words, while some did not often do so.

Overall, these findings suggest a few things. First, the traditional news media
seem to allow both candidates the opportunity to describe themselves in their own terms.
Both McCain and Obama were able to speak directly to the public within the stories of
CNN, Fox, MSNBC, the network news sites, and the New York Times in their own

words. Nevertheless, both of the candidates found ample opportunity to voice their

60

opposition towards the candidate of the other party within these sources, an opportunity
both parties engaged in heavily. Second, the news media seems to promote a fairly open
dialogue between the candidates, where the writers themselves are disproportionately
unlikely to voice their own opinions about the candidates or define them within the
language and flameworks of the writes themselves. In terms of perspective, the news
media organizations relied heavily on the candidates themselves to discuss themselves
and their opponents, while other perspective and definitions were underreported,
consistent with the top-down approach to political reporting that Herbert Gans discussed
earlier. Third, these findings suggest that the traditional news media journalists are
concerned about the perception of being biased by their readers, and are more willing
than other writers (the candidates and bloggers) to take a hands-off approach to the public
debate, allowing the candidates to be quoted directly and being more likely to keep their
own perspectives and interpretations to themselves.

Question3 #3 :Is anyone criticizing the Coverage of the Election? Where are these
criticisms coming from?

So where are the critics of the news media and their coverage? Not surprisingly,
these criticisms come mostly flom the blogosphere. Table 6 shows that bloggers,
particularly the conservative ones, were most likely to voice criticism of the political
coverage (their articles criticized the media coverage 33.8 percent of the time).
Interestingly, the candidate and party websites seemed particularly cautious against
portraying the media as biased or that they were doing a poor job. The Democratic Party
sites criticize the media only once in 62 stories, while the GOP websites criticize them
only 3 times out of 78 stories. In sum, more than half of criticism about news media

coverage of the candidates and the election came flom bloggers, who often voiced

61

fl'ustrations about the coverage and claim that the media worked against their candidate,
as both sides demonstrated the “hostile media phenomenon” discussed earlier, where both
sided felt that the news media favored the opposition’s candidate. Once again, these

differences were found to be statistically significant (Chi-Square = 50.031, p < .000)

Table 6: Which Sources are criticizing the media coverage?

 

 

 

 

 

 

 

 

 

 

 

l ls the media Criticized?| I
No Yes I Total
does article Media Organization Count 216 231 23
W“? % 90.4% 9.6% 100.0171
Liberal 8109 Count 71 8 71
% 89.9% 10.1% 100.0%
GOP Blog Count 53 27 80}
% 66.3%] 33.8% 100.0%
Democrat Party Site Count sol 1 61
% 98.4% 1.6% 100.0%
GOP Site Count 75 3 7
% 96.2% 3.8%l100.0°:l
Total Count 475 62 53
% 88.5% 11.5%I100.0°/:|

 

 

 

 

Question #4: Are the articles in some sources more likely to cover the election as a
“horserace” that is full of polling numbers and pundit advice?

Perhaps the reason that conservative bloggers felt that the news media was biased
was related to the “horse race” flamework that dominated their coverage. Table 7 shows
that online news stories referred to an Obama advantage 107 times out of the 537 news
stories that were covered here, while McCain was said to lead or be gaining momentum
only 8 times during this same time period. The polling reports consistently mention that
Obama is “in the lead” or is “winning” key states and national polls. By the last few
weeks of coverage, most news stories that discussed John McCain nearly always
followed up with phrases such as “trailing in the polls”, or “running out of time”, and

would allow Obama or Biden the opportunity to suggest their campaign was “desperate”.
62

Table 7: "Horserace" Coverage by News Source

 

 

 

 

 

 

 

 

 

 

 

Is anyone seen as being in the lead? I
Not Mentioned Yes. ObamaT Yes. McCain
or Tied leading/gaining leading/gaining Total
does Media Count 151 84 4 23
mg; °°'“° WWW” % 63.2% 35.1% 1.7% 100.ij
Liberal Blog Count 63l 15 1 7
% 79.7% 19.0% 1.3% 100.09:
Conservative Count 74 3 3 801
B'°9 % 92.5% 3.8% 3.8% 100.0%
Democrat Count 57 4 0 61
Pa“ Site % 93.4% 6.6% .0% 100.0%
GOP Site Count 77 1 0 7
% 98.7% 1.3% .0%l 100.092l
Total Count 422 107 8 537
% 78.6% 19.9% 15%| 100.0%

 

 

 

 

 

Overall, thirty-five percent of all mainstream news stories referred to polling
numbers which indicated Barack Obama was leading or gaining momentum in the polls.
Interestingly, both bloggers and party websites seemed very unlikely to talk about poll
results or horserace coverage. This “horserace” flamework was predominantly a product
of the news media organization news stories, findings which are consistent with many
critics of mainstream political news coverage. These findings were significant (Chi-
Square = 77.24, p < .000), and show that “horserace coverage” was almost exclusively
the domain of mainstream news coverage online.

Question #5 : How do sources difl'er in terms of what topics are covered? Are ditferent
sources more likely to talk about political issues, the personal character of the
candidate, or to cover the election as a “horserace” that is full of polling numbers and
pundit advice?

Horserace coverage was not the only way in which different sources had a
propensity to cover the presidential election in varying ways. Table 8 shows that there are
strong and significant differences (Chi-Square = 106.523; p < .000) in the topics that
were under discussion in political news coverage during this time period. In particular,

63

the blogs (both liberal and conservative) were more likely than other sources to talk about
the personal character of the candidates and significantly less about specific issues, such
foreign policy or health care.

At the same time, both of the candidate websites were disproportionately likely to
talk about the issues and policy while downplaying poll numbers. The Democratic
websites discussed political issues 47.5 percent of the time, while GOP party sites talked
about the issues 64.1 percent of the time. However, the GOP websites were also likely to
speak about personal character (doing so 24.4 percent of the time) while the Democratic
websites did so only 8.3 percent of the time. Additionally, as already mentioned, the news
media organizations were overly focused upon polling numbers, which were often the
centerpiece of the news story. They were much less likely to talk about personal character
of the candidates when compared to the blogs, but also less likely to talk about the
important political issues than were the candidate websites.

Table 8 is interesting because it shows that different sources were likely to
emphasize different topics, trying to set the agenda and focus the public’s attention to
different places. While mainstream news organizations are often preoccupied with giving
snapshot and scoreboard like reporting of the race, the candidate websites were largely
the places that discussed the issues themselves. As we have seen, the Obama sites often
defined Barack Obama’s stances on many key issues, while the McCain sites often tried
to define where their opponent stands and to question his character. Simultaneously, the
blogs were places in which personal character was often discussed, along with polling
nmnbers and other items, such as writer or pundit speculation and predictions of what

should be happening or will happen next.

64

Table 8: Topics Covered During the Presidential Race by Source

 

 

 

 

 

 

 

 

 

 

 

iWhat is article discussing? I J
Other Political Personal Political
Topics Issues Character Horserace Total
does article Media Count 46| 76 24 93 2
°°"‘° "a“? 0'9“"iz°“°" % 19.2% 31.8% 10.0% 38.9% moot/j
Liberal Blog Count 21 16 21 21 79]
% 26.6% 20.3% 26.6% 26.6% 100.0%
Conservative Count 20 231 26 11 so]
B'°9 % 25.0% 28.8% 32.5% 13.8% 100.0%
Democratic Party Count 20 29l 5 7 61
S“ % 32.8% 47.5% 8.2% 11.5% 100.0%
GOP Site Count 5 50 19 4 7
% 6.4% 64.1% 24.4%l 5.1% 100.0%
Total Count 112 194 95 136 53
% 20.9% 36.1% 17.7%l 25.3% 100.0%

 

 

 

 

 

Question #6: Did the articles coming from different sources use difl'erent amounts of
anger or hostility towards a candidate? Does the writer suggest that the campaigns are
purposely being deceitful or dishonest to the public?

Finally, I tried to understand what levels of hostility and skepticism were being
drawn out by different sources. Table 9 shows that, perhaps not surprisingly, blogs
showed the most animosity and anger in their news stories. In the 79 liberal blog stories I
read, 32 of these stories (40.5 percent) were personally hostile towards John McCain
and/or Sarah Palin, while only 1 story attacked Barack Obama or Joe Biden. Similarly,
the GOP blogs attacked Barack Obama and/or Joe Biden 29 times out of 80 times (or
36.3 percent of the articles).

However, hostilities were also seen in the websites of the candidates as well.
Although we saw earlier that the candidate websites were hesitant to show contempt for
the news media coverage, they did show hostility towards the opposing candidate. Table

9 shows that the Democrat Websites were hostile towards McCain/Palin 17 times out of

65

61 stories, while the GOP sites attacked Obama/Biden 33 times out of 80 stories. In both
cases, these figures exceeded expectations based on the null hypothesis that different

media sources showed the same patterns of behavior.

Table 9: Does the Writer Express Hostility Towards the Candidate?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hostile towards Candidate?
Yes. toward Yes, toward
McCain or Obama or
Palin No Biden Total
Wheredoes article Medla. . Count 191 204 16 2
mm"? organ'zatm % 7.9% 85.4% 6.7% 1000::l
Liberal Blog Count 32 46 1 79'
% 40.5% 58.2% 1.3% 100.0%
Conservative Count 5 46 29F 60F
B'°9 % 6.3% 57.5% 36.3% 100.0%
Democrat Party Count 17 431 1 61
Site % 27.9% 70.5% 1.6% 100.0%
GOP Site Count 0 45 33 78
% .0% 57.7% 42.3% 100.0%
Total Count 73 384 80 53
% 13.6% 71.5% 14.9%[ 100.093

 

The news media organization websites were the least likely to engage in personal
attacks on the candidates, doing so below expected levels for both Barack Obama and
John McCain. Again, these findings are consistent with the critiques of some media

scholars that these organizations are very concerned about being accused of bias and
losing critical market share, and are therefore less willing to openly criticize or attack the
candidates during the election than others were. Likewise, it also supports the critiques
that journalists have made towards the blogs: that they are sites that focus on personal
character and hostility, and are detrimental to serious political discussion and insight into

the important issues of the day. These findings were significant (Chi-Square = 174.923, p

66

 

 

---=
_——

< .000) and show that some sources of news tend to be more venomous towards

candidates than others are, and too often in the direction of party bias.

Table 10: Does the Article Accuse the Candidate of Dishonesty?

 

 

 

 

 

 

 

 

 

 

 

 

Dishonest? ,
Yes. I Yes, Yes. Both
No GOP Democrats candidates Total
‘ does Media. _ Count 193 191 19l BI 23
gm 0mm % 80.8% 7.9% 7.9% 3.3% 1000;]
Liberal Blog Count 63 14 0 2 71
% 79.7% 17.7% .0% 2.5% 100.0%
Conservative Count 62 4 14 0 sol
B'°9 % 77.5% 5.0% 17.5% .0% 100.0%
Democratic Count 53 8 0- 0 61
Pam 5“ % 86.9% 13.1%l .0% .0% 100.0%
GOP Site Count 461 0 32 0 7
% 59.0% .0% 41.0% .0%l 10009:]
Total Count 417 45 65 10 53
% 77.7% 84%| 12.1% 1.9% 100.07]

 

 

 

 

 

 

In regards to skepticism, I wanted to know if different sources varied in
questioning the Claims being made by the candidates. Table 10 shows that the blogs were
more likely than the news media to question the claims being made by the candidates, but
they predominantly only challenge the claims made by their opponents. We see that 14
out of 79 (17.7 percent) of liberal blog stories accused McCain and Palin of being

dishonest, while only 2 accused both Obama and McCain of being dishonest, and none of

those stories singled out Obama for dishonesty. Likewise, Obama was accused of
dishonesty 14 times out of 80 (17.5 percent) of conservative blog stories, but only
accused McCain of doing so on 4 occasions

Interestingly, we see that the mainstream news media organizations were entirely

balanced in their claims of candidate dishonesty. Table 10 shows that 19 traditional news
media organization articles accused the Obama camp of dishonesty and 19 accused the

67

McCain camp of dishonesty, while 8 articles accused both candidates of doing so.
Meanwhile, we can see that the GOP websites had a disproportionate amount of accusing
Obama of dishonesty, doing so 32 times in 78 (41 percent) of their news stories. This is
almost entirely a function of the “BarackBook” theme that was run by the rnc.org
website, a continuously updated chronicle of questionable Barack Obama relationships
and potentially scandalous cover-ups or conspiracies he was believed to be engaging in,
which served as the leading story of this website on a daily basis during the US
Presidential election.

These differences were statistically significant (Chi—Square = 105.33], p < .000)
and suggest once again that source differences in election coverage are real across a wide
array of dimensions: fl'om what topics are discussed, in regards to whom is able to make
their claims in their own words, in levels of hostility, where different sites were more
likely to question the facts, yet while also securing to exhibit very high levels of political
bias within blogs and candidate websites. Meanwhile, traditional news coverage often
engaged in a lot of “top-down” reporting; relying on the candidates themselves to fill the
spaces of their articles while placing a strong emphasis on polling numbers.

Discussion and Conclusion

These figures indicate that different online news sources were highly likely to
portray the candidates in varying ways from each other. In particular, the bulk of political
blogs contained a heavy dose of negativity towards the opposition’s candidate, while
spending significantly less time and effort praising their own candidates. While bloggers
often complain about bias in news media organization coverage, they ironically seem to

show the highest amount of bias in their own reporting. Reading the blogs illustrates the

68

“hostile media” hypothesis; as writers flom both the right and left criticized the coverage
of their candidate, and sometimes even suggested this bias was intentional. In contrast,
media organizations were the only source category that reported neutrally at least 20
percent of the time, exhibiting the least amount of favoritism towards any candidate. All
other sources engaged in negativity towards their opponent at disproportionately high
rates.

Meanwhile, mainstream news had a propensity to cover the election as a contest.
Similar to what Williams (2006) found in earlier elections, the prevalence of horse-race
coverage predominated over detailed attribute and issue content of the issues and
candidates. The news media were also more objective in their criticisms of the
candidates, and did so in more civil terms. However, blogs and mainstream news both
had a tendency to see the candidates in a negative light, which could have an effect on
voter turnout and apathy during a general election.

If one were interested in hearing candidates discuss the issues, their best option
was to go directly to the candidate websites. On these sites, the candidates talked about
the issues more often than did other sources, although they also had a tendency to speak
in terms of personal character as well. Both candidates spent a great deal of time trying to
use the Internet as a medium to define the election. Much effort was spent trying to
describe themselves in characteristics that were desirable of performing a presidential
role, and questioning their opponent’s ability to do the same. Candidate websites are
informative, yet are clearly constructed with the motivations of persuading the reader to

think in ways which are desirable to the candidate’s platform. Perhaps the best strategy

69

for a web surfer interested in gaining unbiased political knowledge is to seek out
information flom multiple sources in order to avoid a reliance on one perspective.

Taken together, the criticisms that journalists and bloggers make towards each
other seem to have some justifications. The reports coming flom blogs often seem biased
and hostile, which firels criticism made by journalists about blog reporting. In contrast,
reports from media organizations largely rely upon candidate quotes flamed between poll
numbers, often speculating about what may happen next or how it voters may be swayed.
Bloggers portrayed journalists as being too comfortable with their careers and too
embedded with political elites to risk challenging facts or in giving voice to critics, and
used these feeling to resonate with an Internet public that may have felt similarly.

I think future research should focus on the feelings that journalists and bloggers
have towards one another, and how those feeling effect their selection of news stories and
their methods of reporting. For example, research may find that bloggers wear their poor
reputations with journalists as a badge of honor, while journalists place a high premium
on objectivity so that articles will not be seen by peers as low—quality rhetorical pieces.
Likewise, experiments may also indicate whether blog postings are more or less
persuasive than news articles, and under which conditions and topics? In this chapter, I
have assumed that all articles are equally likely to be persuasive to the reader. However,
it may be the case that blog viewers enjoy reading a favorite blog but do not consider the
stories to be accurate. Just because one prefers to read blogs and does so consistently, this
does not necessarily indicate that they believe what they read. It is possible that readers
are entertained by blog reports and share the writer’s feelings on the topic, yet are

skeptical of the facts or perceived bias, using mainstream news as a supplement.

70

Nevertheless, there are plenty of areas for future research regarding political
information on the Internet. Although beyond the scope of this chapter, it would be
interesting to see how consistently supplying negative messages about candidates to
subjects over multiple days in an experiment affect the subject’s interest in politics.
While it may be expected that negativity towards the candidates would dampen the

public’s interest, the Presidential race of 2008 appeared to be as lively and intense among

the public as any in recent memory. Does the presence of the Internet, by giving direct
connections to political groups and candidates, make Internet users feel an extra sense of
importance or vested interest in the outcome of the election? Alternatively, it is possible
that the enormous amount of political stories being reported via so many news sources
could lead to information overload and social withdrawal. While social theorists have
long suggested that such a situation could arise with an expansion of communicative

technology (Sunstein, 2007; Manjoo 2008; Cooley 1998), research has yet to show this.

As blogs are still relatively new, it will be interesting to see how their business
models may change over time. How will these online news sources generate revenues?
Will they be uniquely different from other major news sources, or will it lead eventually

to the same type of conflicts suffered by Mainstream news media? (Hamilton 2004).
Given their levels of partisanship and rhetoric, it is not irnprobably that the most
influential of bloggers will receive financial support flom political parties in the future.
Will this make their reports undesirable for readers, given their economic incentive to
generate biased reports? How will the quality of the reports be altered, and how might the

influences of money and power make these articles feel disingenuous to the reader?

Overall, the 2008 election, as seen on the Internet, showed a high amount of

71

 

convergence. Hundreds of reports were generated by politicians, journalist, and bloggers
each day and placed alongside one another for the Internet user to consume. In the next
two chapters, the emphasis will shift flom the senders of online political news to the
receivers. The remainder of this dissertation will examine whether Internet users engage
in high levels of selective exposure in online news consumption, and if these patterns are
related to the personal characteristics and motives of the users. Additionally, specific
identities thought to drive differential usage patterns will be explored, as will the

implications that these findings suggest for future political discourse in the US.

72

 

CHAPTER 3
The Self Goes Online: How Personal Attributes Mediate Political Activity on the
Internet.

Abstract

The purpose of this chapter is to incorporate social psychological literature
regarding the self to the study of online politics. It is proposed that the presence of the
Internet, by itself, does not alter political activity or attention in the US. Instead, the
political identities and self-construal of lntemet users drives whether people use
computers to meet their political interests. An online survey was distributed to 640
respondents during the winter of 2009 to test four specific hypotheses made in order to
clarify the relationship between personal characteristics and online politics.
Introduction

Since the earliest days of the Internet, people have wondered how this new
technology would alter everyday life. Some feel that the Internet can help facilitate and
expand the principles of democracy, strength the public sphere, and create a public which
is more active and informed about politics than ever before. Alternatively, there are
others who see a dark side to the presence of the Internet, seeing it as a device which can
enable people to ignore important societal issues altogether in the pursuit of personal
pleasures and entertainments.

But while this debate has generally centered around aggregated statements of how
the Internet is universally “good” or “bad” for society as a whole, there has been
surprisingly little consideration of how identities (both personal and social) can lead to

widely variant patterns of usage fiom one person to the next. Individuals have an offline

self that they bring to the computer when they sit down to use it; and these identities are

73

likely to shape what we seek out online. There is reason to believe that some Internet
users will seek out spaces where they can become more informed and/or politically
active, while others will look to avoid these issues. In a sense, the arguments being made
are not mutually exclusive; the Internet leads to both high levels of political activity for
some types of users yet will simultaneously have no positive impact on political activity
levels for other types of users.

The next section will explore the debate regarding how the Internet has a positive
or negative effect on communities and politics. I will then introduce some important
findings about the self-concept and identities that must be considered in serious
discussion of how and why people use the Internet. At that point, four hypotheses will be
developed and tested showing how the Internet usage of people with varying identities
and self-construal types will engage in significantly different patterns of exposure and
activism regarding online politics.

Community and the Internet

As mentioned earlier, there are some scholars who feel that the Internet will
atomize people and make them disconnected from their community needs similar to the
effects of television (Putnam 2000), and are sometimes referred to as “Dystopians” (Katz
& Rice 20028). For dystopians, the Internet can be seen as primarily an individual-
centered entertainment and shopping venue, where serious collective political action is
unlikely to surface (Putnam 2000; Galston 2002; Galston 2004). The Internet is viewed as
an electronic technology which allows us to consume hand-tailored entertainment in
private, even utterly alone. People may become reasonably well-informed spectators of

public affairs, but many fewer will actually partake in politics (Putnam 2000). In other

74

words, Putnam sees the Internet as rendering our leisure time as more private and passive
“spectatorship” (Jensen 2006: 40); where more and more of our time and money are spent
on goods and services consumed individually and a declining interest in making personal
sacrifices towards the cdmmon good (Putnam 2000). Yet, when people do undertake
political action, it is too often something that requires no more effort than a few seconds
of pointing-and-clicking, often in the form of donating money instead of personal time
(Putnam 2000), which is heralded as being far from ideal or adequate for a healthy
democracy. People will typically spend little/no personal time with meaningful offline
political activities, but will instead sign up for mass mailings or other passive
consumption of information that requires virtually no forms of personal commitment or
sacrifice (Putnam 2000: 63).

Others have emphasized that time spent on the intemet acts as a substitute for
other offline activities such as spending time with friends and family, leading users to
feel a sense of isolation from their immediate community, even leading to higher rates of
depression for those who use the intemet for several hours a day (Klotz 2004). Contrary
to utopian claims, the Internet is likely to resemble a contentious and chaotic marketplace
of ideas containing many different groups, leading to little consensus or collective unified
action (V acker 2000). Some have even suggested that the quality of material being
reported on the Internet is increasingly problematical, with concerns that this may lead to
an even further reduction in political participation (Katz & Rice 2002B).

Additionally, since the Internet is a vehicle that enables the user towards endless
choice and personal selection, some scholars feel it is unlikely that the Internet can

reduce already declining social capital or sense of community among users (Galston

75

2002, 2004; Wachbroit 2004; Putnam 2000). They discount any notion that online groups
are solutions for our current civic ills, let alone be comprehensive models of a better
future (Galston 2002). These scholars generally see what they refer to as “Choice”,
“Voluntary”, or “Associational” communities being distinguished by: low barriers to exit,
low barriers to entry, and interpersonal interactions shaped by mutual adjustment
(Galston 2004; Putnam 2000; Wachbroit 2004; Sunstein 2007 ; Bishop 2008). These
characteristics lead to online communities that are highly homogenous, where there are
low levels of personal commitment, and activities are largely centered on personal
interest and entertainment; while noting that the ability to easily ignore pe0ple does not
create any incentives for accommodation (Galston 2005). Moreover, these kinds of
groups threaten to undermine democracy and traditional community (Wachbroit 2004;
Sunstein 2007) while fostering intolerance.

By their very nature, Internet groups are almost entirely selective communities
with very low barriers to entry and no barriers to exit. These “weak” communities are not
seen as sufficient for meaningful social change, as people often leave these communities
whenever asked to sacrifice personal time or effort for the cause. They act as a pseudo-
community (Putnam 2000) and do not create meaningful ties which are bonded strongly,
the kind of connections that can emotionally connect people to make personal sacrifices
towards a cause (Galston 2004). These communities have no tangible real-world value, as
memberships in them are by-and-large characterized by low levels of commitment and do
not often lead people to participate in “real” off-line activity.

According to Putnam (2000) the Internet only exacerbates the decreasing amounts

of social trust and social capital that have been suffered over the last 40 years, a trend

76

which has cut across all population and geographic sections of the US (Putnam 2000).
Moreover, Putnam (2000:171) sees “virtual social capital” as “a contradiction in terms”
and purports that at a macro level, the Internet will lead to decreasing amount political
involvement and possibly less knowledge of political issues across society, especially
among younger people. In this view, the enthusiasts of “virtual community” who see
these networks as the basis for utopian communitarianism are overly optimistic about the
emergence of the Internet, Rather, this new communicative technology is more likely to
fiagment the public and make them too passive and individualistic to engage in political
endeavors (Putnam 2000).

Critics of this perspective point out that before the advent of the Internet there had
already been a move from all-encompassing socially controlling communities towards
more individualized and fragmented personalized “network ” communities
(I-Iaythornthwaite & Wellman 2002; Wellman 1999). For decades in the US, people have
lived their lives through a series of overlapping personal social networks or partial
communities, giving way to the opportunity of “networked individualism”
(Haythornthwaite & Wellman 2002). Overall, people have remained connected to their
communities, but do so in a more socially networked and selective way. Individuals
manage their social networks as a more-or-less selective community, and utilize this
network to obtain information, personal support, and develop a sense of belonging
(Haythornthwaite & Wellman 2002; Wellman 1999). Moreover, these scholars point out
that the Internet has been unfairly blamed for a loss of civic-mindedness and political

involvement. To the contrary, the presence of the Internet actually encourages and creates

77

new forms of connectedness and involvement for its users, which often leads to higher
levels of civic engagement (Hampton & Wellman 2002; Haythomthwaite 2007).

These scholars, sometimes referred to as “utopians” (Katz & Rice 2002B), point
out that there is a positive connection between online and offline activities that leads to
more civic engagement. The Internet can expand community by making people
“Glocalized” (Hampton & Wellman 2002), which allows people to re-define community
to include people living beyond their immediate locale rather than replacing them with
like-minded others who live far away. The Internet is viewed as a tool which empowers
people to become more active both within their local communities while also
transcending the limitations of time and space (Giddens 1991) making Internet users both
more locally and globally active than are non-users. Most importantly, however, is a
rejection that online and offline behaviors are mutually exclusive. Instead, online
activities have the ability to generate increased knowledge and interest in a person’s
community and make them more active in their local communities and national political
interests. ‘

Findings have shown that those who use the Internet are found to be more
informed about what is happening within their community than are those who do not
(Kavanaugh & Patterson 2002; Katz & Rice 2002A; Oates & Gibson 2006). As a
refutation of dystopian claims, researchers have shown that online community members
have report the kinds of strong emotional and social bonds commonly associated with
local community (Haythomthwaite 2007). Additionally, others have noted that Internet
users were also more likely to engage in political activities such as voting, and also that

the Internet provided a platform for a significant amount of additional forms of political

78

activity that were not available offline (Katz & Rice 2002A). The Internet is not a
substitution for political life, but a vehicle which empowers and creates activity: The
more time people spend on the Internet, the more they are involved with organizations
and politics, both offline and online (Haythomthwaite & Wellman 2002: 31).

Instead of seeing the Internet as a distinctly separate entity from the communities
which we physically live within, these scholars support the notion that these communities
often overlap (Davis, Elin, & Reeher 2005; Kolko 2003; McKenna 2007); that being a
member of an online community does not make one less connected to an offline
community (Rice et a1 2007; Rheingold 1995). Instead, people interact with many of the
same people in both spheres (Quan-Haase et al 2002) a point which others have claimed
extends to political communication in the US (Jensen 2006). Moreover, others have
argued that the Internet has been effective at reaching people who might not have been
interested in the political process, engaging and encouraging them to become active
participants in the political sphere, and emphasizing a positive relationship between time
spent online and political activity (Davis, Elin, & Reeher 2005; Katz & Rice 2002A).
Online political activity is seen as a supplement to offline political behavior; where both
mutually reinforce each other rather than act as substitutes. If a person uses the Internet to
discuss politics, there is reason to believe they will become more interested and involved
in politics offline as well, which may genuinely create new forms of community-oriented
activism.

These scholars see the Internet as a vehicle which leads a resurgence of
community, goodwill and democracy. In their view, people flee to the Internet to find the

purpose, affect and community that are otherwise absent in their lives during this

79

postmodern era (Wise 2003). They envision the Internet as a space where all voices could
be heard, greatly reducing the influences of media elites and opinion leaders in the
democratic process. This is partly due to the fact that these people themselves are
enlarging their own social networks and increasing their number of daily interactions
(Weinberger 2005).

The Internet serves numerous positive functions for users, such as becoming more
knowledgeable and integrated in local community affairs and expanding a person’s sense
of community belonging (Rheingold 1995). Ultimately, the Internet is a space where new
community ties can be developed, new coalitions can be nurtured, and individuals can
even mobilize with people that they may have otherwise never even met (Anderson
2003). Having a space where an endless amount of possible new connections can be
made while simultaneously enabling people to form stronger ties to the people they know
ofiline can only be a good thing for communities and democracy.

What is striking in this debate is how rarely (if ever) utopian or dystopian scholars
have considered any social psychological components of the people who are actually
using the Internet. While these theorists have sought correlations between time spent
online and increased levels of political activity, for example, nobody in this debate has
really considered that some people will be motivated to seek out politics while others will
not. The effects of the Internet are unlikely to be uniform across all types of people across
society, leading to a decisive seismic shift towards one direction or another. It seems
more likely that people who are oriented to collectives and are interested in politics will
become even more connected to these goals by using the Internet, while those who are

excessively individualistic may use the Internet similarly as Putnam (2000) sees it: as a

80

means of private news consumption that does not translate into political activity.
Additionally, it seems plausible that those Internet users which are uninterested in politics
and/or do not define themselves in collectivist terms will become interested in politics
regardless of how much time they spend online.

Although there has been much debate over how the intemet affects political
knowledge, activity, and beliefs of its users, it has almost entirely developed in the
absence of social psychological literature about the self-concept. This has left the current
discourse of how politics on the intemet affects its users both fragmented and incomplete.
The integration of recent developments within social psychology regarding the self-
concept along with the varied arguments made by computer-mediated communication
scholars should prove to be beneficial in a number of ways. It is likely that mere exposure
to the Internet does not lead to any uniform impact, nor do all users share the same
motivations or experience the same patterns of exposure. Instead, the self-concept of
users impacts how people use the intemet. Differences in the overall self-concept are
likely to have strong impacts on what any specific person will see on the intemet, who
they talk to, whether they are persuaded by the information they see, whether they
become involved in political activities, and whether they use the intemet as a source of
information.

As Eric Uslaner (2004A: 108) claims, “There is little evidence that the Internet
creates meaningful new communities, and even less that the Internet breaks down
traditional social ties or trust in others.” The Internet does not uniformly make people
become one way or another. Instead, people bring their offline identities and self-

concepts with them when they sit down at the computer; that the Internet “allows us to

81

become ever more ourselves” (Katz & Rice 20023: 13). Put another way, the Internet
does not remake people’s personalities. We bring our prior selves online; the Internet
neither creates nor destroys trust and communities (U slaner 2004B). As an example of
how these users may differ, earlier research has show that online political activists are
more likely to be well-educated, affluent, younger, and male than other Internet users
(Norris 2002; Owen 2006). These attributes were believed to affect the type of political
discourse that existed on the Internet while raising concerns that there was an under-
representation of those who were less educated, female, poor, and/or older which left a
gap in the representativeness of online political actions and discussions.

Moreover, the Internet is not likely to create the kinds of communities that
generate social trust, but excels in bringing together people who already share something
in common (U slaner 2004B). The Internet does not create a new world; activities
undertaken online generally reflect people’s interests and experience (Klotz 2004). As
Eric Uslaner (2004A: 108) explains “While the Internet makes the world better in some
ways, and worse in others, it is not transforming: if you want to make a revolution, you
have to go offline.” Where previous research has made attempts to show large-scale
changes on political discourse across US society due to the existence of the intemet, this
research project attempts to specify which people will be the most affected by the
presence of intemet exposure, and in what direction will the intemet impact the direction
and intensity of their beliefs, as well as what specific kinds of behaviors are likely to

result fiom the combination of offline identities to online activities.

82

The Self and Identifies

During the twentieth century, Sociology developed an enormous body of
literature regarding the study and measurement of the self, particularly from Symbolic
Interactionist theorists. Overall, the self was seen to be very fluid and in a constant state
of change fueled by reflexivity and constant interactions with others, and could be
thought of in generalities as a ‘self as process’ (Mead 1934; Blumer 1968; Cooley 1998).
These theorists often placed an emphasis on how the self becomes a function of his/her
situational and environmental settings, and called for qualitative observational methods to
understand how people act and why (Blumer 1968, Mead 1934, Cooley 1998, Goffman
1959). Because these scholars discounted notions of a self that transcends situational
demands and the expectations of others, methodologies designed to quantitatively
measure the self-concept were not pursued.

However, the later part of the century saw the emergence of social psychologists
asserting that people had tendencies to act in patterned ways based upon identities which
they held a deep commitment towards (Stryker 1980, Serpe 1987, Burke & Reitzes 1981,
McCall & Simmons 1978, Rosenberg 1978) and “identity” refers to internalized sets of
meanings attached to roles played in a network of social relationships (Stryker 1980).

These theoretical premises assumed that the self was a stable object that could be
quantitatively measured with the hopes of predicting human behavior. Much of the
research within this ‘self as object’ paradigm set out to see which identities were
important to people, how these identities are related to social roles, what motivated
people to enact these identities, and how these identities became salient by situational

factors (Stryker 1980, McCall & Simmons 1978, Burke & Reitzes 1981; Rosenberg

83

1978; Higgins 1987). In sum, the self should be seen as multifaceted, comprised of parts
that are sometimes highly interdependent and sometimes not (Stryker 2000: 27).

Nevertheless, research within this paradigm was difficult due to the lack of
quality measurement instruments of the self and the identities they hold. Perhaps the most
popular measurement technique involved the “Twenty Statements Test” (Kuhn and
McPartland 1954), where identity salience (McCall and Simmons 1978; Stryker 1980)
and Psychological centrality (Rosenberg 1978) could be captured by seeing which
identities were the first to be listed in response to the question “I am _” and other self-
reporting techniques (Burke & Reitzes 1981). These efforts were supplemented by other
social psychologists who were eager to understand how the self incorporates empathy
with close others into their own self-concept (Wright, Aron, & Tropp 2002; Crandall,
O’Brien, & Eshleman 2002). Other social psychologists pointed out the reasons and
processes by which the self often minimized differences with others who share important
identities in order to meet collective needs (Abrams & Hogg 2001; Turner 1999; Cerulo
1997; Howard 2000; Simon & Kampmeier 2001). Likewise, other researchers have
discovered that the self-concept is multi-dimensional, and have attempted to illustrate just
how many unique factors could be identified within the self-concept of an individual
(Deaux et al 1995; Brewer & Gardner 1996).

Others have studied the tendencies of the self to be influenced by external social
and cultural factors which the self is located within (Kitayama et al 1997; Kashima et a1
1995; Roberts & Helson 1997; House 1992). For example, researchers had shown that
people fiom different cultures (Kitayama et a1 1997; Kashima et a1 1995; Gecas & Burke

1995), age groups (Demo 1992), and time periods (Zurcher 1977) have a tendency to

84

define themselves within a certain dimension over time, such as how people think of
themselves either in individualistic or collectivist terms. For example, Kitayama and
colleagues (1997) showed that people socialized in Japanese culture tend to view
themselves primarily in collectivist terms, seeing themselves with regard to how they are
related to important others. In contrast, Americans were more likely to view themselves
in individualist terms, where self—definition is based mainly on how unique and different
they see themselves compared to others. Similarly, Louis Zurcher (1977) also found that
different generations within the US varied in terms of defining themselves in individualist
or collectivist terms, again showing that self-construal is driven by proximal social forces
and the social contexts of the time periods they live within. Newer research within a
similar vein has indicated that modern western societies have transferred the defining of
identity from the group to the individual, making the external pressures of community
less consequential than one’s inner resource; where identity has become a problem, a
goal, or a project (Dunn 2008).

In more recent times, social theorists have expounded upon how current social
conditions and technologies are likely to uniquely affect the experiences and reflexivity
of people living in modern times (Giddens 1991; Thompson 1995; Baumeister 1986;
Cerulo 1997; Callero 2003; Wood & Smith 2005). Communication devices such as the
Internet are thought to help the self transcend the limitations of personal experience and
understanding that are imposed on it from immediate physical environments (House
1992; Goffman 1959) and empower people into pursuing “identity projects” (Giddens
1991: 107) and to have expanded their understanding of the world through vicarious

experience and expanded interaction patterns (Giddens 1991; Thompson 1995; Agger

85

2004). In modern times within capitalist democracies, people find themselves in a culture
that permits them to pursue desired identities and pursue a “life trajectory” (Giddens
1991: 109) where people can attempt to define themselves within an extraordinarily wide
array of possible identities. As stated by Dunn (2008: 160) “The modern self is a complex
collection of identities that nevertheless organizes itself around a constantly developing
narrative that exhibits varying degrees of integration and coherency.”

Further works that explore relationships between the self and modernity have
incorporated materialism and identities constructed via consumption patterns and an
ethos of personal selectivity as an unalienable right of individuals (Schwartz 2005; Slater
1997; Dunn 2008; Walker 2008). These authors have discussed how these conditions
force the self to find balance between independence and interdependence from others in a
struggle which is simultaneously liberating but creates anxiety, a basic condition of living
in modern life (Giddens 1991; Hewitt 1989; Schwartz 2005).

However, by the end of the 20th century, some scholars were pointing out that
modem (or perhaps post-modem) living conditions had made previous conceptions of the
self-concept antiquated. The basic premise of these theorists is that the self lacks a core or
an organization; that there are simply too many identities and complex social situations
available today for the self to navigate its way through life in such a straightforward or
predictable fashion. Some have suggested a more post-modem approach to the study of
the self is now necessary, sometimes referred to as ‘self as narrative’, where people are
more or less a jumbled collection of stories and ephemeral roles that may or may not
overlap with each other and is Lmlikely to be organized into a cleanly defined hierarchy of

importance or commitment (White 1992; Gergen 1991; Holstein & Gubrium 2000).

86

Similar to the Symbolic Interactionists, they believe research of the self should focus
upon “discursive practices” (Holstein & Gubrium 2000: 90) in social situations and how
these identities are constructed and integrated with one another. Ultimately, it is
purported that this is a superior way to understand people rather than assessing an overall
structure of the self, and does not attempt to uncover how constellations of identities
within it are organized.

Instead of driving social psychologists away fi'om further studying the self,
however, many have been attempting to incorporate these opposing and even
incompatible viewpoints regarding the self-concept towards a more coherent whole; a
perspective which conceives of the self as being both flexible and yet relatively stable
and measurable guides much of the approaches towards synthesis (Stryker & Serpe 1994;
Sedikedes & Brewer 2001; Stryker & Burke 2000; Stets & Burke 2000; Aron &
McLaughlin-Volpe 2001; Wright, Aron, & Tropp 2002). The fruits of this research have
indicated that the self-concept consists of several distinct core components which exist
simultaneously with each other and vary in salience at any given moment (Rosenberg
1978; Stryker 1980; Deaux et al 1995; Brewer & Gardner 1996), leading to the
emergence of the “tripartite self” in the current literature (Sedikedes & Brewer 2001).

The tripartite perspective of the self-concept contains three overlapping
dimensions which are distinctive and yet often overlap with each other. These dimensions
are defined as the individual self, the relational self, and the collective self The
“individual self” pertains to the aspects of the self-concept that are unique and
idiosyncratic to the individual. The “relational self” incorporates the important social

roles (such as being a “father” or a “student”) and the identifiable relationships that the

87

self has with specific other people (for example, the friendships someone may have with
pe0ple named “Bob” or “Mary”), and looks at how these relationships impact the
person’s own self-definitions of who they are. The “collective self” focuses upon the
social groups which a person belongs to, and how that person incorporates these group
memberships and perspectives into their own (Sedikedes & Brewer 2001). These
membership groups are often more abstract in nature, and might include, for example,
such traits as your political identity, nationality, ethnicity, or other organizational
affiliations.

Development of the tripartite self-concept synthesizes and incorporates much of
the literature of the twentieth century within social psychology, giving researchers a more
complex but realistic view of the self-concept. While other leading researchers had
previously focused almost exclusively upon one specific dimension of the self-concept in
their research (Rosenberg 1978; Stryker 1980), the tripartite self seeks to gain a fuller
understanding of the self by looking at how all of these identity components
simultaneously co-exist with each other within the self over a person’s life. Within the
self, each identity can become more or less salient as situations and contexts drive
different identities to the forefront (Stryker, Owens & White, 2000). The tripartite self-
concept purports that, at any given moment, an individual’s identity is altered by
situational factors which make some of their identities relevant to the immediate situation
more salient than others. Supporters of the tripartite self-concept also view the self as
something which is a both stable and fluid, persistent over time and amenable to a variety
of measurement techniques (Deaux et al 1995; Brewer & Gardner, 1996; Oyserman &

Packer 1996; Sedikedes & Brewer 2001).

88

However, all of these complexities and conditions need not discourage one from
conducting research on how the self-concept and identities may be related to attitudes and
behavior. While the self may have a great many number of identities contained within a
multidimensional self-concept, there is reason to believe there are very specific identities
and self-construal difference that relate to how people use the Internet to become
informed or involved. Moreover, the tripartite self conceptualization is a synthesis of
previously divergent findings and assumptions about the self. This rubric of
understanding sees the self as both process and object, recognizing that the self’s
behavior can at times be situation driven while still maintaining some tendencies to act
according to patterns of personal characteristics and shared social identities.

The tripartite self-concept sees all people containing within them all of their co-
existing identities at all times in all places. These identities may take their roots within
personal values, important interpersonal relationships, previous experiences, and
important group memberships. Moreover, this framework lends itself to studying the self
within social situations, as there is little emphasis placed on analyzing all of the inner
workings of the self, but rather focusing on those identities which are the most important
and salient within the situation the researcher is interested in understanding.

In other words, one does not need to uncover all of the identities held within the
self-concept, nor reveal how each are related to the other. Likewise, this conception of
self concedes that people have an immeasurable number of identities in a social world
that is complex and ever-changing. Nevertheless, people can be understood within these
situations as actors which have a semblance of long-term permanency and some

predictability where there are personal tendencies that relate their self to their social

89

environment. Although situational factors, identity conflicts, and external pressures that
influence identity salience can alter how a person behaves in a social situation, measuring
the most relevant identities of the towards a specific field of interest is likely to yield
results that state how a person is likely to act in a given situation, with all other things
being equal.

For our purposes, the situation of interest is how the political identities of
individuals are related to how they use the Internet. What is known from previous
research is that the Internet is really not introducing new people into political
participation; it is strengthening the resources of those who were already involved (Oates
& Gibson 2006) and that political partisans seem to strengthen their political interest and
involvement via the Internet (Jensen 2006; Bimber 2003). There also seem to be
important personal demographic variables that influence how someone uses the Internet
for political information, as these users tend to be younger, male, and better educated
than the rest of the public (Bimber & Davis 2003; Davis 2005; Owen 2006). So even
though the Internet is a unique environment, users behave in accordance with the findings
of the main personality theories, where personality-specific characteristics can explain
differential behavior of people on the Internet (Amichai-Hamburger 2007; Davis 2005).

However, there is still more to be uncovered. Bruce Bimber (2003) calls for more
research on this topic, suggesting that many personal variables have yet to be discovered
regarding political info seeking and engagement on the Internet. However, it seems
plausible that the Internet is unlikely to lead to great political transformations, but will

extend “politics as us ” to a new turf (Katz & Rice 2002; Oates & Gibson 2006)

90

appealing to some users and not appealing at all to those who lack strong or clearly
defined political identities.

With respect to seeking political information online, Bruce Bimber (2003)
speculates that the Internet is a place where information costs fall and sources multiply,
creating a situation in which the information-rich get richer and the information-poor stay
poor. In short, people will generally use the Internet as a supplement to political
information rather than as a replacement; that those who follow political news in
traditional sources also follow it online (Bimber & Davis 2003). For the most part, it is
the same audience. Research from Klotz (2004) indicates that only 50 percent of
Americans say they enjoy keeping up with the news, and increased availability is unlikely
to alter interest in it. While none of these authors made an attempt to utilize social
psychological research on measuring identities, their finding do suggest that one should
expect a relationship between political identity and the amount of political information
the person seek online. Those who lack an interest in politics or have weak political
identities are unlikely to use the Internet for political purposes. However, those who are
interested and knowledgeable about politics seem to use the Internet to meet these
interests.

In regards to political identities, it has been found previously that people rarely
change their political affiliations once they decide they are Democrat or Republican.
People have the notion that a political party is filled with "your kind of people" (Bishop
2008: 23). Additionally, other research has shown that attitude strength and the
importance a person places upon politics serve as the most important determinants of the

amount of political behavior (Skitka & Mullen 2002). Yet, there is still a question

91

regarding whether there is a strong linkage between seeking out political information and
being actively engaged in politics. Some have found that being informed does not
necessarily seem to indicate political activity, as the new informational environment has
not previously been seen to change levels of political engagement in any substantial way
(Bimber & Davis 2003). However, other researchers have indicated that online political
involvement supplements high levels of offline political activity, indicating that the
Internet is a place where the usual “gladiators” come to participate (Jensen 2006: 49).
This seems to be supported by the research of other scholars, who show there is very little
evidence for the notion that the existence of new communicative technologies makes
political activists out of ordinary people (Davis 2005). It is far more likely that people
who are politically active offline will use the Internet to localize and disseminate
information and engage in political actions more often than others will. The Internet is a
medium for political influencing public opinion and policy, not just a space for discussion
(Margolis & Resnick 2000).

Nevertheless, it appears that political activity is not clearly related to higher levels
of time spent online nor political browsing and discussion, as these have previously been
found to be unassociated with levels of political activity (Katz & Rice 2002B). This
implies that social scientists need a deeper understanding user motivations and identities
to understand these linkages between online attention and offline activity. Although some
of these works have related political identity to political intemet usage, they have not
considered the self-construal of individuals as well in considering how these tendencies
to see oneself predominantly within individualistic or collectivist terms may motivate

some people to become active, while leading others to be more passive spectators of

92

online politics, bypassing offline political activity despite defining themselves as interest
in politics.

These differences in personal self-construal may indicate why some scholars such
as Putnam (2000) find civic engagement to be harmed by the Internet, and others such as
Wellman (1999; 2002) find the Internet to lead towards increased levels of civic
engagement. Perhaps the self-construal combines with important political social
identities that motivate people to become knowledgeable and connected or occasionally
to become knowledgeable but disconnected or for some others to even completely avoid
politics altogether, regardless of how much time they spend online overall. The next
section will discuss how the self strives to meet important personal desires and needs
online, and explains how identities can serve as important motivations in their online
patterns of behavior.

Uses and Gratifications: The Self goes online

While one can think of identities as the understandings, social roles and narratives
that make up the self-concept, identities can also be thought of as containing a
motivational element, as they move people to actions in a predictable manner (Stryker
2000; Stryker 1980; McCall & Simmons 1978). The combination of motivations and
personality determine online interactions, where different types of people will use the
Internet in varying ways to meet personal needs they seek to satisfy online (Amichai-
Hamburger 2007).

This chapter adopts the assumption that people use the intemet to meet personal
needs and goals that they develop primarily offline. Overall, The Internet helps people

meet these needs by finding others who share mutual interests while providing the

93

freedom to choose the groups and social roles that suit them (Amichai-Hamburger 2007:
197). Additionally, the Internet permits the fieedom to leave online groups with no
negative social consequences when these needs are not being met.

Moreover, Internet users have certain predispositions and needs that motivate
them to seek different programs and sources, often referred to as “Uses and Gratifications
Theory” (Norris 2002: 60). This is a prominent theory is Mass Communications Research
which purports that “individuals have needs and wants that can be satisfied through
media consumption” (Traudt 2005: 200). Uses and gratifications theory sees media
audiences as active recipients of messages, where “social and psychological elements
mediate the persuasiveness of messages; stresses individual usage and choice” (Rubin
2002: 525-526). In short, uses and gratifications theory is rooted around five basic
assumptions (Rubin 2002: 527-528) illustrated in the following:

1. People choose media content; their selection and use of media is goal—driven,
purposive and motivated

2. People actively select and use communication devices to meet their needs or
desires

3. Our predispositions, living environments, and interpersonal interactions shape
what we expect in media communications; responses are filtered through our
personalities, social categories and relationships

4. The media compete with other forms of communications and interactions for our
attention; the ability of the media to gratify personal needs varies between

different people
5. Our own initiative mediates the patterns and consequences of media use

While these are the basic premises of online communication, the theory recognizes that
there will be occasions where uses and gratifications assumptions will be violated. For
example, there are likely to be meaningful differences in personalities that moderate
differential uses and exposure to differing stimuli, such as attention to news sources and

the need for cognition, or personality traits such as masculinity or aggression leads to

94

increased desires to play violent video games (Rubin 2002). In some instances, the
existence of certain individual characteristics may heighten or intensify the influence of
messages, or may even provide a necessary condition for attention and influence to take
place at all (Rubin 2002). As a consequence, the media rarely have direct effects on all
users; instead, there are important differences in the selection and attention to messages
which is problematic since people can differ in a virtually infinite number of ways (Rubin
2002: 507).

Uses and gratifications theory recognizes that personality, social context,
motivation and availability all impact the possible influences of the media (Rubin 2002)
but that personal motivations have a strong impact on what messages that we seek and
attend to online. Perhaps most importantly, this theory indicates that Internet users are
active participants in their online experiences that can either challenge or ignore what
they see on their computer screens. The Internet does not make the user become one way
as so many of the scholars discussing online community have insisted; rather, the user
utilizes the Internet to meet their own interests and needs, virtually all of which were
formed prior to their online usage. Furthermore, by measuring the most relevant identities
and motivations regarding politics and combining these with how the individual sees
his/her own general self-construal (or understanding), research should be able to indicate
how various users approach the Internet in differing ways.

Hypotheses

Given the theoretical discussion regarding the self and the basic tenets of uses and

gratifications theory, some predictions of how the online self is expected to pursue online

politics begin to emerge. In this section, 1 specify six specific hypotheses.

95

Gm

\m'h

If

the,
\\V

HI. Those that have strong political identities will be more likely than others to
become politically active and on the Internet. -- This is consistent with Uses and
Gratifications theory. People who have strong social identities rooted in political interests
will meet those needs to become politically active by using the Internet more than those
who do not.

H2. Those that have sa'ong political identities will be more likely than others to seek
political information on the Internet. --- This is also consistent with Uses and
Gratifications theory. People who have strong social identities rooted in political interests
will search the Internet for political information at higher rates than those who lack strong
beliefs about politics.

H3. Those who are interested in politics will be more likely than others to become
politically active and on the Internet. -- Consistent with Uses and Gratifications theory,
people which are interested in politics will show higher levels of online political activity
that other users will.

H4. Those who are interested in politics will be more likely than others to seek political
information on the Internet. -— Also consistent with Uses and Gratifications theory,
people which are interested in politics will show the highest amount of time seeking out
information on the Internet.

H5. There will be a Positive relationship between Self-Construal and an interest in
politics. -- People who have a propensity to incorporate important others into their own
self-concept will be more likely to take an interest in political issues. Those who see
themselves as unique individuals may have less interest in pursuing collectivist needs via

political movements.

96

Ofer

faceb
this
5 Ur

FE‘

SR

index]

H6. There will be a Positive relationship between Self-Construal and holding strong
political identities. -- Since collectivists may be more likely to take an interest in
political issues, they are also more likely than individuals to develop stronger and more
intense political beliefs than others.
Methods

Data used to test these hypotheses were collected using an online survey during
February, 2009. This survey purposively sampled those individuals who were members
of either a political or non-political facebook.com group. Approximately 30 separate
facebook.com groups were contacted and each member was personally invited to take
this survey by clicking on a link which took them directly to the website that hosted the
survey. During this time period, 624 different respondents completed the survey. The
results were recorded by a website provider named zapsurveycom and were coded by the
author upon the respondent’s completion of the survey.
Variable Construction and Measurement

There will be six variables used in this chapter: online political activity levels,
online political information seeking, having an interest in politics, the strength of political
identity, independent self-construal, and collectivist selfoconstrual. The variables “online
political information seeking” and “online political activity online” will be used as
dependent variable, and regressed against all of the other variables mentioned above in
three separate models. The first model will contain only two key Specific identities
(political interest and strength of political identity), 3 second model that includes
independent and collective self—construal, and a third model which includes some

interaction variables. The results from the second models will indicate whether a person’s

97

self‘“
Scckin‘

mania

\

self-construal has any direct effect upon political activity online and political information
seeking online, while the third will indicate if these independent variables have
meaningful interactive effects on the dependent variables.

Additionally, regressions will be performed to predict whether an interest in
politics or the strength of a person’s political identities can be predicted by their self-
construal. Analysis in this chapter seeks to find out if self-construal has more of a direct
or indirect effect on political participation levels on the Internet, and whether collectivists
or individualists are more likely to be political participants online. Before testing the six
hypotheses in this chapter, a brief discussion of how each of these variables was
measured will be conducted.

Measgrm’ g Self-Comm

It should be noted that my factor analysis of the data suggests that the Singelis
(1994) self-construal scale was two-dimensional. I used eight questions fi'om this scale,
and found that two significant dimensions emerge by the analysis of the initial
Eigenvalues using direct oblimin rotation in table 11. The rotated component matrix (see
table 12) indicated that five of these eight questions loaded onto a single factor
(component 1), while others loaded onto the other component. Looking closer at these
dimensions indicated that component one reflects the respondent’s degree of collectivist
sentiments, while the other dimension points to how much the user sees him/herself in
unique and idiosyncratic ways. After standardizing all of these eight observed indicators,
I created two separate indices for self-construal: a five-item “collectivist self-construal”

index, and a three-item “individualist self-construed” index.

98

fab

 

$0.9

"w
N

Table 11: Factor Analysis of Self-Construal Scale

 

Total Variance Explained
Rotation
Sums of
Squared
7 Initial Eigenvalues Extraction Sums of Squared Loadings Loadings

 

% of
Variance Cumulative 96 Total

 

iComponenti Total lee of VariancJCumuianve % Total

 

 

 

 

 

 

 

 

1 1.996 24.923 24.923 1.996 24.923l 24.93022] 1.
2 1.719 21.493 46.306 1.719 21.493 46. 1.71
9 .995 12.314 58.619
4 .979 10.992 69.602

.734 9.171 79.773

.679 9.490 97.263

.600 7.499 94.759

.419 5.241 100.000

 

 

 

 

 

Extraction Method: Principal Component Analysis.

Table 12: Component Loadings

 

 

 

 

 

Matrix for Self Construal
Component
1 2

-.041 .594
.075

-.016 .7
.596 - 07
.695
.563 -.022;
.559
.715 .01

 

 

 

 

The individual items were simply added together and were also unweighted in the
calculation of these indexes. Both will be used measures will be used to regress against

99

the relevant dependent variables in this chapter: political activity online and political
information seeking online.
Measuring Strength of Political Identity

This variable was constructed in two parts. First, six separate five-point Likert-
scale questions were asked to indicate where the respondent stands on a number of well-
discussed and prominent political issues (Tax Levels on the Wealthy, Size of
Government, Government Regulation, Social Welfare Spending, Gay Marriage, and
Separation of Church and State) to determine the direction and strength of political
identity by adding all of these scores together (where the maximum “conservative” score
was +12 and maximum “liberal” score would be -12). However, since strength of
political identity was Our only concern here, this variable used the absolute values of each
respondent: thus, all negative integers were transformed to positive numbers, so that -12
would now be +12, -11 to +11, etc...giving this variable a range ofO (no political
affiliation) to +12 (strong political affiliation to a political party).

Table 13: Factor Analysis Results for Strength of Political Identity

 

 

 

 

 

 

 

Total Variance Explained
Initial Eigenvalues Extraction Sum of Squared Loadings
Cumulative
nent Total % of Variance Cumulative % Total % of Variance %
1 3.402 56.706 56.706 3.402 56.706] 56.7
2 .913 15.216 71.922 91
3 .562- 9.374 91.296
4 .407 6.779 99.073
.405 6.752 94.925
t .31 oi 5.175 100.000

 

 

 

 

 

Extraction Method: Principal Component Analysis.

100

tend:

vane

 

 

b
’00;

,2 _
‘ --__‘

In doing so, the respondents are not identified by their conservative or liberal
tendencies, but instead by the degree of intensity they hold these values. In a sense, this
variable separates the politically minded {tom the apolitical. Factor analysis indicated that

this measure is one-dimensional, as can be seen from the data in tables 13 and 14.

Table 14: Component
Loadings Matrix for
Strength of Political
Identity

 

 

 

 

.741

 

 

 

 

Measuring Political Interest Levels

This variable was measured using two separate five-point Likert scale questions
to indicate the amount of interest and commitment the respondent had to their political
beliefs. The questions asks the respondent if they feel certain that their political views are
correct, and asks him/her how interested they are in following political issues. Responses
ranged from “strongly agree” to “strongly disagree”, with an option for having no opinion

01' 11012 sure.

10]

vane

\m

 

These questions had a correlation of .581, which was found to be significant at a
99 percent confidence interval. This can be seen in table 15. Overall, the scoring of this
variable ranged between -4 (not being interested in politics) to +4 (being highly interested
in politics).

Table 15: Correlated Measures for the Variable Political Interest

 

 

 

 

Correlations
Political Certainly
in my Beliefs Politically Interested

Political Pearson Correlation 1 .581
Certainty in . .

N 624 62

Pearson Conelation .581" 1
WW Sig. (2.1111166) .000

N 624 62

 

 

1" Correlation is significant at the 0.01 level (2-tailed).
Measuring Pglitigl Activig 0min;

This index asked the respondent five Yes/No questions to indicate the political
actions they have taken online. These actions included the following items: signing an
online petition, using the intemet to meet for offline political action, donating money,
volunteering time, or trying to persuade others to vote for their candidate in online
discussion. Data that indicates the loading and Eigenvalues for these analyses can be seen
in tables 16 and 17, respectively.

Each “Yes” to the relevant survey question (listed in appendix b at the end of this
dissertation) was scored as 1 point, each “No” scored as 0, giving the index a range of
scores from 0 to 5. Factor analysis (see table 16) indicated that this measure is one-

dimensional.

102

Tab!

 

 

 

Table 16: Factor Analysis Results for Political Activity Online

 

 

 

 

Total Variance Explained ,

Initial Eigenvalues [ Extraction Sums oquuared Loadings j

Total | % ofVariance 1 Cumulative % | Total I % ofVariance Cumulative 91. |

1 2.220 44.403 44.403 2.2201 44.403 44.4031
2 .955 17.099 61.502
3 .730 14.600 79.101
4 .650 12.995 99.096
15 .545 10.904 100.000

 

 

 

 

 

Extraction Method: Principal Component Analysis.

Table 17: Component
Loadings Matrix for
“Political Activity Online f

Component]

 

 

 

1 T

.731

.71

 

Measuring Online Political Information Seeking

This variable asked the respondent to indicate if he/ she had sought political

information from any of four possible sources online (message boards, emails, blogs, or a

pOpular online news source, such as cnn.com, for example). Each “Yes” response was

scored as 1 point, and all “No” responses were scored as 0. Those who found news online

from each source were scored as a 4 (maximum) and those who sought political news

from none of those sources were scored as a 0. Factor analysis indicated that this measure

is one-dimensional, which can be seen in tables 18 and 19.

103

Table 18: Factor Analysis Results for Political Information Seeking
Total Variance Explained

 

 

 

 

 

 

 

 

initial Eigenvalues I Extraction Sums of Squared Loadings J
nent Total % of Variance I Cumulative % I Total % of Variance I Cumulative %J
1 1.921 49.029 49.029 1.921 49.029I 49.029I
2 .872 21 .811 69.841
3 .758 18.951 88.792
ii .449 11.209 100.000

 

 

Extraction Method: Principal Component Analysis.

Table 19: Component
Loadings Matrix for
Political Information

Seeking

 

L Component]

 

 

1 |

.77

 

Data Analysis and Results

my Those which have strong political identities will be significantly more
likely than others to become politically active and on the Internet. -- And HMLEBS—iil
-- Those who are interested in politics will be significantly more likely than others to
become politically active and on the Internet.

Table 20 is used to predict the dependent variable “Political Action Online” using

three models. The first model uses only two independent variables: interest in politics,

and the strength of their political identity. While these two variables have a moderately

strong correlation with each other (Standardized Beta = .356, P = .000), this correlation

indicates that they are quite distinct. The second model seeks to find if there are any

direct relationships between collectivist self-construal and individualist self-construal to

104

the dependent variable, Political Action Online. Meanwhile, the third model tests the

interactions of all the predictor variables to see if any of these are significant and

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
   

 

 

 

 

meaningful.
Table 20: Regression Results for Predicting Political Activity Online
Coefficients _
Unstandardized Standardized l
Coefficients Coefficients

3 Std. Error Beta t 319.
(Constant) .625 .21 1 2.970 .
Strength of Political identity .132 .01 .262. 6.919 .3
interested in Politics? .954 21(91I .175 4.546 .
(Constant) .687 .2 3.286 .001
Strength of Political Identity .135 .01 .268 7.038
Interested in Politics? .868I .209 .159 4.155 .
Collectivist Self-Construal -.o17 .01 -.033 -.888 .37
Score
lndlvidualist Self-Construal .103I .027 .144 3.827
Score
(Constant) seal .29 3.915]
Strength of Political Identity .094 .025 .188 3.822
interested in Politics? .737 .1 3.341
Collecflvist Self-Construal .004 .008 .1 14
Score
Individualist Self-Construal .058 .081 1.046
Score
Interact Pol Int 9. Pol Strong .016 .132 2.591
Interact Coll so 9 P01 Strong .000 -.007 -.097
Interact lnd sc 9 Pol Strong .005. .051 .629
interact ind SC 81 Pol Interest .007 .020 .385
Interact Coll SC & P01 -.011 -.052 -.986
Interest

 

Dependent Variable: Political Action Levels Online

What was found in the first model is that both Strength of Political Identity
(Standardized Beta = .262, P <.000), and Interest in Politics (.175, P < .000) have positive
and Significant relationships in predicting Online Political Action Levels. This data

indicates that we can reject the null of hypotheses 1 and 3, and conclude that those users

105

who are highly interested in politics and have strong political identities engage in higher
levels on online political activity than do other users.

In model 2, we can see that the coefficient of Collectivist Self-Construal was
found to be both negative and not significant, indicating that collectivist self-construal
does not have a direct effect upon online political action levels. Interestingly,
Individualist Self-Construal was found to have a positive and direct effect with online
political activity levels (Standardized Beta = .144, P < .000). On the surface, it would
seem that collectivists may be more likely to be interested in politics and social
movements. However, the data here suggests that individualists more likely to turn to
their computers to feel connected to politics. Perhaps this is reflective of Putnam’s (2000)
claim that the Internet is largely a vehicle of individual entertainment or usage: while
collectivists who are interested in politics may want to participate in physical proximity
of others, individualists wish to connect in more distant terms, being able to plug in and
out of politics on their own personal schedules and terms. Although this is not what was
predicted and is beyond the scope of this chapter, this finding may serve as an interesting
topic for future research.

Model three of this regression shows us that self-construal, whether it is strongly
individualist or collectivist in nature, has no significant interactions with Political Interest
or Strength of Political Identity. There does not seem to be any special combination
between self-construal with Political Interest of Strength of Political Identity that led to
an additional amount of online political activity. The only interaction term which was
significant and positive (P = .01) was between political interest levels and strength of

political activity. It appears that the combination of having strong political attitudes and

106

being interested in politics lead to significantly higher levels of online political activity
levels than other Internet users. As can be seen in table 21 below, including self-construal
proved to be useful in predicting online political activity (Incremental F = 7.028, P =
.001), while the interactions in model three were not useful (Incremental F = .896, P =
.483). Overall, it appears that model two was the best predictor of online political
activity, suggesting that self-concept has some direct effict upon this dependent variable.

Table 21: Which Model Best Predicts Political Activity Online?

 

 

 

 

 

Moggl Adjusted R-Sguare Incremental F-Test
One .113 40.817“
Two .132 7.028“

Three .137 .896

 

 

 

 

 

*" F-Value is significant at the 0.01 level.

Hmthesis 2: Those which have sa'ong political idena’ties will be significantly more
likely than others to seek political information on the Internet. --And Hymthesis 4 ---
Those who are interested in politics will be significantly more likely than others to seek
political information on the Internet.

Table 22 turns our attention to the dependent variable “political information
seeking on the Internet” while using the same four independent variables to predict it:
Independent Self-Construal, Collectivist Self-Construal, Interest in Politics, and Strength
of Political Identity. Similar to our previous analysis, three separate models are used. The
first model predicts political information seeking using only two independent variables:
interest in politics, and the strength of their political identity. The second model seeks to
find if there are any direct relationships between collectivist self-construal and
individualist self-construal with the dependent variable. Meanwhile, the third model tests
the interactions of all the predictor variables to see if any of these have significant and

meaningful impact on political information seeking.

107

Using the data from Table 22, we can see from the standardized beta coefficients

of model one that the independent variables “Strength of Political Identity” (Beta = .131,

P = .001) and “Interest in Politics” (Beta = .262, P < .000) have positive and significant

effects with regards to how much political information one seeks online, as predicted in

our hypotheses. The data here indicates that we can reject the null of hypotheses 2 and 4,

and conclude that those who are interested in politics and have strong political identities

will seek out more political information online than do other people

Table 22: Regression Results for Predicting Online Political Information Seeking

 

 

 

 

 

 

 

 

 

Coefficients
Unstandardized Standardized
Coefficients Coefficients

Model 8 Std. Error Beta 1 Sig.

1 (Constant) 1 .674 .095 17.597 .
Strength of Political Identity .049 .015 .131 3.232 .03
Political Interest .186 .029 .262 6.470| .

2 (Constant) 1 .673 .095 17.588
Strength of Political Identity .050 .015 .132 3.258 001
Political Interest .185 .029 2601 6.374
Colbctive Self Construal .016 .015 .042 1.097 27
Individualist Self Construal .018 .020 .034 .892 .37

3 (Constant) 1.558 .113 13.784 .
Strength of Political Identity .078 .021 .206 3.744 .
Political Interest .291 .056 .410 5.243 .
Collective Self Construal .029 .029 .076 1.021 .
lndividualist Self Construal -.017 .042 -.032 -.410 .68
Inter SC Coll 8 Pol Strong -.002 .005 -.026 -.334 .7
Inter SC Ind & Pol Strong .010 .007 .121 1.467 .1
Inter SC Ind & Pol Interest -.015 .013 -.065 -1.203 .
Inter SC Coll 8 Pol Interest .000 .009 -.002 -.029 .9
Inter Pol Int 8. Pol Strong -.019 .009 «205. -2.168 .031

 

 

 

 

 

 

Dependent Variable: Seeking out political information on the intemet

Model 2 seeks to discover if there are any direct relationships between Self-

Construal and Political Information seeking online. However, the coefficients for

108

Individualist and Collectivist Self-Construal were not found to be significant. The data
shows that Individualist Self-Construal Internet users did not seek out more political
information online than did others. We can see from the data in this chapter that the
driving direct influences on differential usage regarding online politics is an interest in
politics and a strong affiliation to a political party rather than the overall self-construal of
the Internet user.

Model three shows that Self-Construal does not have any meaningful interactions
with Political Interest and Strength of political identity. The only significant interaction
term in model three was between Political Interest and Strength of Political Identity,
which strangely enough was a negative (Beta = -.205, P = .031). This may suggest that
only one of these qualities is necessary to seek out information on the Internet, where
there are no additive effects between these two personal aspects that cause a person to
seek out additional political information online.

Overall, table 23 indicates that Self-Construal is not an important predictor of
online political information seeking, nor is any of the interactions meaningful. Model one
is the best predictor of online political action (Incremental F = 38.481, P < .000). Table
13 shows that while the political identities and motivations of the user are important
factors in information seeking, the overall self-construal is not. Model two provided no
improvement to the prediction of the dependent variable (Incremental F = 1.002, P =

.368), nor did model three (Incremental F = 1.458, P = .202).

 

Table 23: Which Model Best Predicts Online Political Information Seekigg?

 

 

 

 

Model Adjusted R-Sguare Incremental F-Test
__ 0118 .107 38.481"
L_ Two .107 1.002
Three .111 1.458

 

 

 

 

 

** F ~Value is significant at the 0.01 level.

109

Hypothesis 5: There will be a Significantly Positive relationship between Self-
Construal and an interest in politics. -- And Hypgthesis 6 — There will be a
Significantly Positive relationship between Self-Construal and holding strong political
identities.

The next analysis in this chapter wants to understand if Self-Construal is a
predictor of an Interest in Politics and Strength of Political Identity. If so, it might have
direct effects on Political Activity Online and Online Political Information Seeking.
Table 24 shows that Political Interest levels has no significant relationship to Collective
Self-Construal (Beta = -.05, P = .213) but does have a very weak yet positive and
significant relationship to Individualist Self-Construal (Beta = .091, P = .023). These
findings suggest that we can, at least partially, reject the null of hypothesis 5, and
conclude that in this population those who have an individualist self-construal are slightly

more likely than others to become interested in politics.

Table 24: Regression Results: Self-Construal and Political Interest

 

 

 

 

Coefficients ,
Unstandardized Standardized
Coefficients Coefficients
Model 3 Std. Error Beta t 819
1 (Constant) .904 .012 76.9161
Collectivist Self-Construal -.005 .004 -.050 -1.246 .21
Score
Individualist Self-Construal .012 .005 .091 2.273 .02
Score

 

 

 

 

 

 

Dependent Variable: Interested in Politics?
R-Square = .011; R-Square Adjusted = .008; F = 3369*

However, the results in table 25 show that there is no relationship between Self-
Construal and the Strength of Political Identity. Both the Individualist (Beta = -.019, P =
.63 8) and Collectivist (Beta = .001, P = .976) self-construal have near zero and non-

Significant relationship to the Strength of Political Identity. These findings show that we

110

cannot reject the null of hypothesis 6, indicating there are no direct relationships between
Self-Construal and Strength of Political Identity.

Table 25: Regression Results: Self-Construal and Strength of Political Identity

 

 

 

 

 

Coefficients #
Unstandardized Standardized
Coefficients Coefficients
9 Std. Error | Beta t Sig.
1 (Constant) 5.922 .129r 45.279
Collectivist Self Construal .001 .041 001 .030 97
Score
Individualist Self Construal -.027 .057 -.019 -.471
Score I

 

 

 

 

 

Dependent Variable: Strength of Political Identity
R- Square = .000, Adjusted R-Square = .000, F = .112

Other Relevant Findings Coming from the Data Analysis

The correlation matrix in table 26 yields some interesting findings. We can see
that the correlation between the two dependent variables (seeking out political
information and online political activity) is very large and positive (r =.576, p < .000)
indicating that the politically active are also seeking out political news online. It seems
that those who are active in online politics and seeking news are many of the same
people, or the ‘political junkies’ that Margolis and Resnick (2000) pointed out dominate
online politics. Interestingly, correlation analysis shows that seeking out political
information have a positive correlation with time spent online (r = .169, p < .000) and
also correlates with the direction of political identity (where positive scores indicate
conservative scores, negative scores indicate liberal scores), showing that Liberals spent
more time seeking out political information in this sample (r = -.O79, p = .048), although

this effect is apparently very weak.

111

Table 26: Correlation Matrix of Important Chapter 3 Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Correlations
Collectivist Indlvlduallst Seek out
Self Self political Political Strength
Construal Construal info on Activity Political
Score Score intemet Online Identity
Collectivist Pearson 1 -.004 .030 -.041 .001
Self Correlation
WW’ Sig. (2- .922 .453 .302 .975
tailed)

N 624 624 624 624 624
Individualist Pearson -.004 1 .057 .153" -.019
Self Correlation
Cmm' Sig. (2- .922 .155 .0001 .637

tailed)

N 624 624 624 624 624
Seek out Pearson .030 .057 1 .576" .224“
political info Correlation
" mm Sig. (2- .453 .155 .000 cool

tailed)

N 624 624 624 624 624

Pearson -.041 .153" .576" 1 .295“

Correlation

Sig. (2- .302 .0001 .0001 .0001

tailed)

N 624 624 624 624 624

Pearson .001 -.019 .224" .295" 1

Correlation

Sig. (2- .975 .637 .000 .000 .000 .000 .51

tailed)

N! 624 624 624 624 624 624 624 62
Political Pearson -044 .099 .309“ .309" .356" 1
Interest Correlation

Sig. (2- .270H .014 000+ .000 .000

tailed)

N 624 624 624 624 624 624
Directionof Pearson .094' -.041 wail -.226" -.290" .042
Political Conelation
"WW Sig. (2- .019 .301 .049 .000 .000 .291

tailed)

N 624 624 624 624 624 624
Tlme Online Pearson -.015 .055 .169“ .215" .026 .026

Conelation

Sig. (2- .707 .167 .000 .000 .517 .509

tailed)

N 624 624 624 624 624 624

 

 

 

*. Correlation is significant at the 0.05 level (2-tailed).
*"'. Correlation is significant at the 0.01 level (2-tailed).

1

12

 

There were similar findings regarding political action online, where the politically
active were found to be online more than others (r =.215, p < .000) and more liberal (r = -
.226, p < .000). These correlations might suggest that political liberals in the US are
turning to the Internet to find connection and take action more so than are conservatives,
seeking out in cyberspace rather than (or perhaps as a complement to) face-to-face
interactions.

Discussion and Conclusions

These results strongly suggest that the main variables among those that we have
examined that drives political activity online is an interest in politics and a strong identity
with a political party, which supports hypotheses 1 and 3. People interested in politics
and that have a strong sense of connection to a specific party ideology used the Internet to
pursue and promote their political values online. They used their computers to make
connections and become involved, while those who were not political did so at
significantly lower levels. The significance of independent self-construal in model one
may also suggest that those who are strongly individualistmay spend more time online
than others. Perhaps these people look to the Internet to connect politically from a
distance rather than to be in close physical proximity to other people who share their
views and wish to act.

Overall, there seems to be a linkage between offline identities and online
behaviors. The data shows that the peOpIe who conducted political actions online were
those who had long been interested in politics and had a strong sense of connection to a
political party, consistent what Margolis and Resnick (2000) have found. The Internet is a

new battleground for the same political gladiators that also do battle against each other

113

offline. This is certainly consistent with “Uses and Gratifications Theory” discussed
earlier. People bring important identities with them to the Intemet, and these identities
channel where the user spends their time and attention. Given the many possible items to
focus attention on the Internet, people who are interested in politics seek to satisfy their
interests and motivations such as political activism and news seeking. Internet usage is
not making a person more or less political by itself. Instead, the identities of the computer
user which determines whether a person seeks out information online or not.
Although this is beyond the scope of this chapter, future analysis may show that
online political people spend disproportionately more time online. The Internet is a
medium which allows the politically minded to seek collaboration and form movements
among like-minded individuals, activities which take a great deal of time and effort.
Future research may better indicate whether or not ‘political junkies’ that spend more
time on the Internet than other people do would reverse some of the declines in social
capital that Putnam (2000) has decried previously. In other words, is the time being spent
online forming tangible and valuable social capital and political movements, or is much
of this investment in personal time and energy proving to be unfruitfirl? Research
questions that shift from the motivations of the user to the efficacy of these online efforts
could prove to be very useful.
However, it is still unclear to how network ties encourage or discourage activism

(Snow & McAdam 2000: 63). Others have pointed out that structural and political
inequalities in combinations with the constraints within the political system are likely to
dampen the effects of extensive political change stemming from Internet communications

(Margolis & Resnick 2000). There is a danger to assume that increased grassroots

114

political activism and community building can fully be able to transcend offline realities.

Longitudinal and qualitative research conducted among politically active lntemet users

could prove to be very informative as well.

Finally, other research could better identify which other important identities are
related to differential rates of online political activity and information seeking. Surely,
there are many more identities that serve as political motivations and make collectivist
social identities salient when a person sits down at their computer. Likewise, the
everyday offline living situations and relations to other institutions (such as the
potentially competing demands of work and family, and the influence of friends) are
likely to impact online behaviors, and these factors have not been addressed here. Using a

variety of methodological approaches, future research could clarify and illustrate these

important relationships to online politics.

115

CHAPTER 4
What Makes Someone a Cyber Balkan? Finding the Linkages Between Social
Psychology and Self-Selectivity in US Politics Online.

Abstract

This chapter uses survey data collected in the winter of 2009 to uncover the
possible linkages between community selectivity, online selective exposure patterns, and
the personal characteristics of the computer user. The purpose is to uncover which
personal attitudes, values, living environments and motivations can likely lead a person to
become a cyber Balkan. Several potential independent variables will be discussed and
explored while also indicating what these results mean regarding the quality of political
discourse in US politics today.
Introduction

Concerns have grown about the possibility of the lntemet creating a more
politically divisive and polarized nation, a claim that has been made by political
researchers such as Stanley Greenberg (2004). From such a vantage point, the lntemet is
seen as having the potential to divide users by areas of interest, creating polarization via
selective attention (Van Alstyne & Brynjolfsson 1997; Sunstein 2007; Bishop 2008).

Other critics have pointed out that the lntemet by its very nature is antithetical to
political action because time spent online is primarily a solitary activity devoid of
meaningful contact or action with others (Putnam 2000). In this view, the lntemet acts as
a substitute for time that could have been spent productively offline and with other

people: Hence according to this view, the more time someone spends online, the less time

they spend offline in activity with others. Consequently, increased online activity will

116

result in a more politically ignorant and uninvolved user that will be less likely to take
offline political action (N ie & Erbring 2005; Putnam 2000).

However, earlier chapters in this dissertation have pointed out that the effect of
the lntemet on individuals and communities may not be uniform. Some will challenge the
viewpoints of others through rigorous debate, others will drop out of political discussion
entirely, and there will be those who look to surround themselves with like-minded
others. I have predicted that the self-concepts and identities of the user determine whether
one expands or narrows their own horizons. In particular, those who narrow their
horizons will receive focus here in chapter four.

This chapter aims to uncover why someone acts as a cyber Balkan online and
attempt to understand what motives of the self are being met in behaving this way?
Additionally, this chapter seeks to uncover the relevant connections between ofiline living
environments and online balkanization. Does being surrounded by like-minded others
offline translate into a desire for the same online? Also, which beliefs and values are
contributing to the desire to self-select their communities and patterns of online
interaction?

Overall, this chapter attempts to understand the personal and social factors which
cause one to act as a “Cyber Balkan” (Van Alstyne & Brynjolfsson 1997). A cyber
Balkan is defined as a person that (1) tends to define their “community” as containing
those who share who regardless of physical proximity share one’s own values, (2) Prefer
to be surrounded by similar others online, and (3) selectively expose themselves to
websites where they expect others to confirm their own pre-existing viewpoints. In this

sense, a cyber Balkan exhibits similar offline and online behaviors, and their

117

values/attitudes drive these behaviors. Additionally, cyber Balkans have motives, such as
a desire to maintain a sense of certainty in their beliefs, a tendency towards dogmatism,
and to express anger towards political opposition. All of these points will be explored
later in this chapter. In short, Cyber Balkanization can be thought of as patterns of actions
and self-selection. I assume that these actions are motivated by a person’s attitudes
regarding different topics and how they deal with disagreements.

Ultimately, these divisive qualities spread both online and offline, mutually
reinforcing one another. Online relationships are likely to be brought into one’s everyday
offline life and vice-versa, indicating overlap between the online and offline self
(McKenna 2007; Kolko 2003). The lntemet mirrors everyday life. What people do online
is pretty much what they do offline (U Slaner 2004; Katz & Rice 2002; Margolis &
Resnick 2000). As a result, it is feared that public discourse in the US will be reduced to a
place where “We don't know, can't understand and can barely conceive of ‘those people’
who live just a few miles away” (Bishop 2008) if their views are substantially different
than our own. In addition, these homogenous groups tend to march towards extremes
(Sunstein 2009; Manjoo 2008; Bishop 2008; Sunstein 2007), which is a central point of
concern that will be explored throughout this chapter.

Like-Minded Selected Communities and Attitude Polarization

As noted earlier, some scholars feel that online communities are little more than
self-selected pseudo-communities where people share a common identity (Galston 2004;
Putnam 2000; Norris 2004). The lntemet is believed to likely develop more common
beliefs-values groups based on non-affective ties; where people can join and quit at the

click of a button with no feelings of obligation (Levine 2004). The lntemet is seen as a

118

place of consumer choice meets lifestyle: where individuals can transform their abstract
world-views into virtual worlds. Specialized communities can be constructed to reflect
any particular philosophical, identities or ideological perspective to reinforce their own
pre-existing worldviews within a narrow range of dialogue and challenge (V acker 2000).
Increasingly, Americans are seen to be segregating themselves into their own political
worlds, blocking out discordant voices and surrounding themselves with reassuring news
and views (Bishop 2008; Sunstein 2007)

In addition, people lack any accountability or responsibility within them, and can
leave whenever they see fit (Margolis & Resnick 2000: 118) which detracts fiom
“meaningful” and “real” communities offline. These Choice (self-selected) communities
are distinguished by the following characteristics: low barriers to exit, low barriers to
entry, and interpersonal interactions shaped by mutual adjustment (Galston 2004).
Moreover, web groups are often driven purely on interest-based self-selection, where the
basic norm for online group membership criterion centers on having a similarity in
interests and viewpoints (Weinberger 2005). In other words, online political groups are
typically places for support and agreement, which leads to patterns of reinforcement and
extreme attitude formation (Sunstein 2009). Irnportantly, like-minded online groups are
quicker than other groups to stop talking to members with contrary opinions and were
more willing to exclude deviance and foster intolerance (Bishop 2008; Sunstein 2009).

Also, the ready availability of exit makes online communities more homogenous,
and the ability to easily ignore does not create incentives for accommodation (Galston
2004; Putnam 2000). The lntemet is seen to generally excel at coordinating self-selected

organizing and mobility. As Robert Putnam (2000: 172) states, “Social networks based

119

on computer-mediated communication can be organized by shared interests rather than
by shared space.” Overall, self-selected communities (both online and offline) intensify
fragmentation and breeds attitude polarization among its members (Galston 2005;
Sunstein 2007; Wachbroit 2004; Galston 2004; Sunstein 2009). People grow more
extreme within homogenous groups as a way to conform; people hope to gain

reward/ status from the group for being even more extreme than the group average,
propelling each other to move the whole group further away from neutrality (Bishop
2008; Sunstein 2009; Galston 2005).

What might be contributing to this process? Some have suggested that selective
community and clustering into like-minded communities is an unconscious process
(Bishop 2008). It has been suggested that Balkanizing may be a symptom of anomie; the
sense of emptiness and disorientation brought about by industrial modern life. This sense
of disconnection leads people to pursue their own community, and ultimately leads to a
nation that is segmented by regions (Bishop 2008) both online and offline. It is suggested
that selective community “clustering” is “a general reaction to a society, a country, and a
world that were largely beyond an individual's understanding or control” (Bishop 2008:
255), suggesting that the rise of these communities are driven by important needs of the
self living in complex modern-day societies.

While Bishop (2008) largely deals with the “unplanned polarization” stemming
from the growth of like-minded communities, Cass Sunstein (2009: 34) claims that much
of the polarization is “planned”, as people actively attempt to create their own like-
minded communities and living spaces. This type of polarization is highly relevant to the

study of online politics, as the lntemet in a place which “allows us to become ever more

120

ourselves” (Katz & Rice 2002: 13). Sitting in front of a computer monitor where there are
virtually endless possibilities to invest one’s time and attention enables the user to exert
fiee choice in all that he/she attends to. The lntemet enables us to confine our
communication to people who share precisely our interests and values (Putnam 2000).

However, Americans have been heterogeneously sorting themselves into political
parties the same way they had sorted themselves into like-minded communities,
churches, and organizations since the mid-1970's (Bartels 2000; Putnam 2000). This is an
important point because it indicates that personal motivations drive selectivity and the
polarization stemming fiom it, and not necessarily the presence of the lntemet itself. One
should not suggest that rising polarization or selective communities are caused by pe0ple
spending too much time online, or that the lntemet prevents a reversing of this pattern.

Nevertheless, problems stemming from the growth and expansion of self-selecting
can be seen in increased levels of political polarization in the US (Bishop 2008). In the
realm of current day US politics, "Americans are increasingly living in communities and
neighborhoods whose residents share their values and they are increasingly voting for
candidates who reflect those values" (Abramowitz , Alexander & Gunning 2006: 88). It is
a bottom-up process; driven by self-selection into like-minded communities both
physically and socially (Oppenheimer 2005; Bishop 2008).

But to better understand Cyber Balkanization, focus needs to be shifted towards
the personal identities and motivations of the user. The motives and self-concept that
each user brings with him/her to the keyboard are different. Because of this, it is expected
that cyber Balkanizing is prominent only among some specific types of Internet users.

The rest of this chapter explores which characteristics are behind these patterns of

121

lntemet usage. It will also examine and outline possible theoretical explanations which
underpin selectivity and polarization online.
SIDE Theory and Online Interaction

Perhaps the most prominent theory explaining attitude extremism online is what is
referred to as “Social Identity and De-Individuation (SIDE) Theory (Spears & Lea 1992).
The roots of SIDE theory are based in previous social psychological works on social
identity and self-categorization theories (Tajfel 1981). Additionally, a “social identity” is
defined as “that part of the individual’s self-concept which derives from his knowledge of
his membership in a social group (or groups) together with the value and emotional
significance attached to that membership” (Taj fel 1981: 225). In general, SIDE theory
purports that the processes of depersonalization and de-individuation which occur offline
when a person is immersed within a groups that shares the same social identity happen
online as well (Spears, Lea, & Postmes 2007).

However, SIDE theory also posits that these effects can be particularly strong in
online communication. Social identities can become even more powerful and
consequential in online communication (Spears, Lea, & Postmes 2007) where a person is
not in the presence of others, particularly those who are different from themselves. The
personal physical isolation combined with the co-presence of similar others increase the
salience of their group membership online while also deepening the effects of self-
categorization and depersonalization of out-group members (Spears, Lea, & Postmes
2007). Moreover, online political groups are very likely to make group identities highly
salient as in-group members are continually present while out-group members are absent,

influencing the actions and thought processes of all of those involved within these

122

interactions (Sassenberg & Jonas 2007).

These elements of social isolation are crucial in understanding how groups
produce extreme attitudes, as polarization is most likely to occur where group
membership is made salient (Sunstein 2009). Like-minded people who engage in
discussion with one another are more likely to adopt more extreme rather than more
moderate variants of the group's shared beliefs (Galston 2005). These effects are
particularly strong in online conversations where there is a lacking presence of out-group
members, but reduced where we can see other people visually via a media-device
(Walther, Gay, & Hancock 2005). In other words, being alone and surrounded by like-
minded others where we are fairly anonymous and unaccountable for what they say or do
produces attitude extremity.

There is a concern that on the lntemet, many people are mostly hearing more and
louder echoes of their own voices; which is of questionable value from a democratic
standpoint (Sunstein 2007). Many lntemet groups, especially those organized around
support for political candidates or policies, are unified bythe sense of personal
camaraderie or shared values among its members. This drives people to a strong tendency
towards group polarization (Sunstein 2009). If people on the lntemet are deliberating
mostly with like-minded others, they will be shifted to more extreme points. The most
important reason for group polarization online involves the exchange of new information
both through conversation and the selectivity of the news sources which are consumed by
the user (Sunstein 2009). As Cass Sunstein (2007: 71) points out “When people
deliberate together, they give disproportionate weight to ‘common knowledge’ and look

to other in-group members as a reference point for ‘social comparison.”’

123

Likewise, the nature of the lntemet makes it easy to find like-minded individuals
to interact with and encourage people to strengthen their own pre-existing judgments
(Sunstein 2007; Manjoo 2008; Sunstein 2009; Bishop 2008). Additionally, the ability of
moderates to leave an online group at any time can fuel extremism (Sunstein 2009),
where the ever-increasing levels of like-minded interaction can breed extreme attitudes
and overconfidence while fueling political division (Sunstein 2007). Although new
communicative technologies such as the Internet hypothetically permits the development
and exploration of newly constructed ‘identity-proj ects’ (Giddens 1991), it is more often
a space where people exploit their own current social identities and personality traits that
were primarily formed offline in their everyday lives (Spears, Lea, & Postmes 2007;
Margolis & Resnick 2000; Galston 2005; Putnam 2000; Norris 2004; Katz & Rice 2002).
Since the lntemet is a medium which enables the user to engage in a virtually endless
array of choices of where to look and who to speak with, collectivist-oriented social
identities thrive on the lntemet. In regards to online politics, group belonging and
interaction lead to depersonalization and de-individuation processes similar to but
perhaps even stronger than in offline interactions (Spears, Lea, & Postmes 2007).

These perspectives are consistent with the major tenets of structural symbolic
interaction and social psychological theories of the self. People typically hold numerous
identities which compete for expression (Stryker 2000), and are motivated to seek out
opportunities to express the identities which are the most important to their self-concept
(Stryker 1980; McCall & Simmons 1978). Persons will seek out opportunities to act out
the roles underlying salient identities, and those with strong identities will move people to

select and engage in situations that evoke these identities (Stryker 2000). Furthermore,

124

Viktor Gecas (2000) has shown that value-based identities (such as political values) give
meaning, purpose, and direction to individuals. They serve as a powerful motivational
force for the self. Values serve as standards by which people live their lives; they “also
tend to be both relatively enduring and trans-situational, tend to be less situational
bounded” (Gecas 2000: 95) making them highly influential of a person’s behaviors.

As mentioned already, personal choice is a critical component to cyber
balkanization. The ethos of the lntemet is consumer choice, but these personal choices
may have taken place without reflection or an awareness of alternatives (Levine 2004).
lntemet users have unparalleled discretion in choosing when and what communication to
receive (Klotz 2004), but it is unlikely that individuals would search for information on
topics in which they have no interest or find uncomfortable (W achbroit 2004). People
who hold different political identities make varying selections on what they wish to see
and what they avoid via “filtering” (Sunstein 2007: 3).

To some extent, with all of the potential political messages one could read or see
online, this selectivity is unavoidable. Without selectivity, people run the omnipresent
risk of information overload: too many options, too many opinions, too many topics.
However, there is a natural human tendency to make choices that do not disturb our
preexisting view of the world (Sunstein 2007) unplanned encounters that disconfirm
viewpoints often prompts one to reassess one’s own view and reduces attitude extremism
(Sunstein 2007; Sunstein 2009). Unfortunately, many lntemet users actively seek to find
only what they want to see. They are conscientiously attempting to design a personalized
communications universe of their own preferences online (Sunstein 2007; Bishop 2008;

Sunstein 2009).

125

The lntemet allows you to easily see exactly what you want to see (Sunstein
2007). People are actively developing personalized environments to seek connection and
interaction with meaningful (like-minded and similar) others in a safe setting (Bishop
2008). Here, the self seeks out those who share their personal tastes, beliefs, styles,
opinions and values. These processes result in “Cyber Balkanization” where the lntemet
produces “A Daily Me” in terms of exposure to political information. People will exclude
troublesome issues or disfavored voices on a continual basis (Sunstein 2009: 79-80).
These consumer preferences can have a devastating effect on the survival of democratic
communities and lead to polarization if all one bears are echoes of one’s own concerns
and perspectives (Wachbroit 2004).
The Daily Me: Consumer Ideology and Selective Exposure

Online political discussion in the US takes place in a cultural context that
emphasizes the rights of free choice and personal independence. These rights are
considered inalienable by most Americans, and are applicable to both consumer goods
and personal ideologies or values. In general, the right and ability to be a consumer is the
ideological birthright of the modern western individual, where all social relations, objects
and activities can be transformed into commodities (Slater 1997). We see ourselves as
people who choose, who are inescapably “fi'ee” and self-managing, who are allowed to
make decisions of who we are and what we want to be (Slater 1997: 5).

Moreover, “individual choice dominates our sense of the social”, where personal
interests center on choice, liberty, desires are the central and overriding social value in
modern society (Slater 1997 : 40). People in US culture have the right and ability to

formulate their own plans and actions, and extend relationships to others nestled with a

126

consumption-based ethos of free choice (Slater 1997; Dunn 2008). To a large degree, a
person becomes self-defining, achieving identity through choice, using his/her own
motivations to determine what they desire and find interesting (Slater 1997). We seek
connections to similar others through our lifestyles, whether these similarities take the
form of materialistic status markers or shared ideological values (Dunn 2008). Moreover,
these lifestyle groups are elective communities; memberships which we choose rather
than have ascribed or allocated to us (Slater 1997).

These situations involving choices are seen more often in the US compared to
other countries, which some have described as a “choice culture” (Iyengar & Lepper
2002: 77). The spread of consumption values spreads to the general society because
consumption itself becomes a central focus of social life; we see ourselves as a society of
choice (Slater 1997). These notions regarding consumer sovereignty “threatens to
become an iron logic of all social interactions” in the US (Sunstein 2007 : 80), and can be
particularly fruitful in understanding how people behave online.

It appears logical that notions of consumer sovereignty act as catalysts to the rise
of selective and largely homogenous commrmities, whether they appear offline or online.
These kinds of communities, and the individuals who find themselves within them,
produce decreased intergroup community and increased political fragmentation (Galston
2004). This can have harmful social consequences as people remain in the safety in the
company of those with similar views and interests (Levine 2004; Sunstein 2007).
Moreover, because the lntemet places virtually no restrictions on what the user attends
to, the problems associated with free choice and selective communities may be even

greater than they are offline. Choice, as a core value, “has run amok on the lntemet” and

127

wherever choice increasing becomes a core value, social bonds will tend to weaken
(Galston 2005: 48).

Consumer sovereignty, the idea that individual consumers are permitted to choose
exactly as they wish, underlies the motivations for users to produce “the daily me” each
day. This occurs when people have the unlimited power to filter the information they
receive (Sunstein 2007), and choose to engage in selective exposure patters, a point
which will be discussed in the next section. These commitments to consumer sovereignty
will undermine political discourse if these kinds of lntemet users have an insufficient
understanding of social problems and Shirk rights and duties to the community while
continually stifling dissenting views (Sunstein 2007). Additionally, the exercise of choice
can have powerful psychological consequences on the self, such as increased levels of
confidence in their beliefs which in turn, reduces their motivation to challenge them
(Iyengar & Lepper 2002).

These values, combined with the complexities of modern life, manifest
themselves online in the form of selective exposure to particular types of users. People
today have access to more information that at any other time in human history, and are
required to decide what information they regard as true or false largely on their own
devices (V acker 2000; Manjoo 2008). Because of this, they are likely to expose
themselves primarily to sources they already accept as trustworthy, and engage in
patterns of selective exposure online.

Selective Exposure
While the lntemet may be relatively new, social psychology has been aware of

selective exposure for many decades. Selective exposure occurs when people focus their

128

attention upon sources with which they agree (Michener, Delamater & Myers 2004: 204-
205). In general, people look to interpret and understand new information in a way that
bolsters their previously held views. In particular, selective exposure is fueled by the
personal motivations of the self such as interests, need, and goals which channel attention
(Michener, Delamater & Myers 2004: 134). Additionally, important personal identities
provide a basis for deciding where people focus their energies and cognitive resources
(Michener, Delamater & Myers 2004: 92).

Research conducted by Muzafir Sherif (1934) showed that when people are
presented information, they are likely to find evidence that confirms their own
preconceptions. Brock and Balloun (1964) also showed that people focus more attention
on messages that confirm their beliefs, and purposely tune out the messages that do not.
These findings are also consistent with the theory of “Cognitive Dissonance” (F estinger
1957), which states that we steer clear of information that disconfirms what we think we
already know.

Furthermore, people are also more willing to accept information they previously
believed to be true at face-value, and tend to pick apart information that disconfirms
those beliefs; 6 process referred to as “biased assimilation” (Lord, Ross & Lepper 1979).
This suggests that selective exposure is related to how someone thinks; as people with
strong identities or feel very certain about their beliefs are likely to engage in heuristic
processing. Research has shown that our self-schemas influence how we select and
process information, selectively focusing upon information towards the confirmation our
most important and salient hierarchies (Michener, Delamater & Myers 2004: 94).

The important points of these frndings have strong implications regarding attitude

129

extremism and lntemet usage regarding political communication.

We should expect to see that those individuals who have already made up their
mind on where they stand ideologically will largely seek out only those sources that
confirm these views. This process seems mostly intentional, as pe0ple have made
decisions about what sources are credible and which are not. They have motivations to
avoid discrepancy, which alters their behavior and selection on political news. The more
control people have over their exposure to political information; the less likely they were
to be exposed to ideas dissimilar to their own; with selectivity being highest among the
strongly partisan (Mutz & Martin 2001; Bimber & Davis 2003; Iyengar & Morin 2006).
The Balkanized would rather be comforted rather than challenged, making mutual
understanding and meaningful discourse very difficult (Sunstein 2007). Research by
Iyengar and Morin (2006) found that when Republicans were giving a choice of which
news stories to read, they strongly preferred ones which displayed to the Fox News Logo,
while liberals had somewhat weaker but significant preferences for CNN and NPR
stories. I

Today, with the presence of the Internet, we live in a new order based on
individual choice with the result of consistently living within a giant feedback loop
(Bishop 2008). The paradox is that greater selection often serves as a motivation to
reduce the number of places where we choose to receive our information from. The more
heterogeneous the marketplace of ideas is, the more likely people will engage in selective
exposure in order “to bolster feelings of certainty” (Bimber 2003: 209). Other research
has indicated that when citizens are given a greater capacity to choose among multiple

media sources, they are more likely to make selections that expose themselves to

130

 

narrower and more personally compatible viewpoints (Mutz & Martin 2001; Bimber
2003).

With media fragmentation, there are now Republican news organizations and
Democratic news organizations, and each side feels a deep antipathy towards the other’s
sources (Manjoo 2008). People are drawn to the messages that are important to them, that
reflect their interests, and do not upset their psychological equilibrium (Whillock 2000).
The liberal blogosphere has been said to be “basically a giant echo chamber” of like-
minded reinforcement (Manjoo 2008: 47) while network analysis has shown that
conservatives, in particular, get their news from densely connected online news sources
(Adamic & Glance 2005). The concern here is that the acquisition of new information
does not make someone more politically knowledgeable if these users tend to avoid any
information which disconfirms their pre-existing beliefs (Bimber 2003). Some have
suggested that Americans have lost the ability to speak in a common civic tongue, as
virtually all social and political issues seem to fall under some type of discord (Bishop
2009; Manjoo 2009) ‘

However, it seems unlikely that all people using the lntemet to gain political
information are engaging in extreme forms of self exposure and constant reinforcement.
One might suspect that some kinds of users are predisposed to these patterns based on
their specific identities, values, and offline social relationships that overlap with what
they do online. In the next section, these specific characteristics will be outlined, and
predictions regarding how they impact online political behavior will then be illustrated

and tested.

131

User Characteristics Potentially Tied to Cyber Balkanization

Although there seems to be important links between social identities and attitudes
to selective exposure, many questions remain. For example, it is still unclear what roles
individual diflerences serve in determining different levels of selective exposure patterns
to media messages? (Bryant & Zillman 2002) In regards to online politics in the US,
there seems to be evidence that selective exposure exists, but who is acting this way?
Also, what specific identities, attitudes and personal characteristics are likely to lead a
person to engage in high levels of self-selective online?

I predict that there are five major factors that impact whether someone acts as a
cyber Balkan online: ofi‘line living environments, trust levels, the Dimensions of a
person’s Political Identity, Personality and Demographic Factors, and Personal
Emotions or Motivations. Each of these factors will be discussed in firrther detail below,
and will include predictions for whether they are positively or negatively associated with
cyber Balkanization.
Offline Environments

As discussed earlier, there is likely to be overlapping between what people do
offline and online. It seems reasonable to believe that those users who live in fairly
homogenous political social networks offline want to find like-minded others online.
James House (1992) illustrated the “proximity principle’, which states that people often
feel the effects of society through interpersonal interaction and communications with
people around us. Others have found that people living within like-minded communities
don't often encounter political disagreement (Bishop 2008), and that online communities

tend to complement offline relations (Katz & Rice 2002).

132

These immediate social and physical environments are also very likely to
influence the attitudes and values of an individual. People within these realms are very
likely to be members of a person’s “reference groups” which are people we look to as a
source of standards and identity, and “primary groups”, the people that we are close to
and interact with regularly (Mead 1934; Cooley 1998) having a profound impact on their
socialization and development of self. The relationships formed with these people serve
as important reference points for the “social comparison” of their political values to
significant immrtant others (Rohall, Millie and Lucas 41; Sunstein 2009: 26). The self is
motivated to make social comparisons between in-group and out-group attitudes to
achieve a positive social identity (Skitka & Mullen 2002). These primary and reference
group members serve as an important benchmark for how we view our self, and finding
consistency between the values we share with loved ones offline with our online
encounters is likely to serve as a strong personal motivation for the lntemet user.

In regards to this chapter, homogeneity with these groups is the key: people with
higher levels of cross-pressures within their primary groups are likely to have weaker
political convictions and are more likely to be uninvolved with politics (Orum 2001).
Through the construction of an online survey, my survey questions will seek to find out if
respondents feel their fiiends/family share their political values. This survey also asks
whether the respondent feels their community largely shares the same values that they do.
It is predicted that there is a positive relationship here: the more people perceive
themselves as living within a like-minded oflline social environment, the more likely they

will engage in cyber Balkanizing behavior online.

133

Lug

Overall, this dissertation considers three forms of trust: generalized trust in others,
trust in the Government, and trust in the media. Putnam (2000) has indicated that most
Americans believe that we live in a less trustworthy society than our parents did (Putnam
2000), which is likely to affect the motivations of the self. In regards to the first category,
research has previously analyzed whether Internet users have higher or lower levels of
trust in others than non-users have. Although Eric Uslaner (2004) found that lntemet use
neither increases or decreases general trust in others, surveys have shown that the lowest
amount of trust in others is found in heavy lntemet users (Robinson et al 2004). It has
also been shown in other research that online political discussants are slightly less
trusting of others compared to those who do not discuss politics on the lntemet (Davis
2005). In short, I expect to find a negative relationship between trust in others and cyber
balkanization, which appears to be consistent with many of the findings above.

In regards to trust in the government, Daniel Bell (1999) believed that people
would find their sense of community in the corporate institutional structure; where
corporations and government become seen as “community”. However, recent times have
shown a loss of trust in traditional institutions, and that these institutions have been
insufiicient in satisfying people's needs for belonging (Bishop 2008). Furthermore, a lack
of trust in government officials discourages people fi'om participating in politics either
offline or online, as the lntemet provides enormous other possibilities which they can
choose to invest their attentions (Davis 2003). Nevertheless, Richard Davis (2005) has
found that online political participants are slightly more trusting of the government than

the general public. I predict that people who hold higher levels of trust in the US

134

government will be more likely to discuss politics online with like-minded others, feeling
some sense of connection and efficacy towards the institution. Those who trust the
government will talk politics with similar others, while those who distrust the
government are unlikely to discuss politics at all, due to a lack of efficacy which causes a
sense of connection to politics (U slaner 2004).

Lastly, research has indicated that people using the lntemet for political
information are more likely to distrust the traditional news media (Bimber 2003).
Additionally, distrust of the news media is widespread and bipartisan among the US
public (Sunstein 2007), as people often cite the belief that news organizations are biased
against their own viewpoints. This has been referred to in the literature as “Hostile Media
Phenomenon” (V allone, Ross and Lepper 1985). Regardless of political affiliation, those
who feel the media is stacked against them results in greater levels of distrust in
mainstream news sources, making the media less persuasive to those who hold this belief
(Manjoo 2008).

Moreover, it should be expected that those who seek out like-minded online
communities and alternative news sources (such as a politically biased blogger) do so
because of disdain for the mainstream news media. In turn, these users attitudes become
Balkanized, or divided into small and hostile units of like-minded groups that pit
themselves against dissimilar others. The more someone distrusts the mainstream media,
the more likely they are attending to like-minded others for information, indicating a
negative relationship between trust in the media and cyber Balkanization.

Dirnflions of Polit_ical Identity; Direction, Sggngm, Emofion and Certainty

There are two dimensions of political identity that deserve attention here: the

135

direction (fi'om conservative to liberal), and the strength of their identity (extreme versus
moderate). In regards to the first point, conservatives were shown by Iyengar and Morin
(2006) to be more willing to consume news which restricts them to perspectives that are
likely to toe the ideological line, avoiding challenges to their strongly held pre-existing
viewpoints. While strongly partisan users fiom all sides of US politics are likely to
engage in cyber Balkanization, this chapter will test whether conservatives are more
likely to engage in this kind of online selectivity as Iyengar and Morin (2006) suggest. In
terms of the direction of political identity, it is predicted here that being conservative will
positively relate to cyber balkanization.

Nevertheless, it is likely that anyone that identifies strongly with a particular
political group will desire to surround themselves with people who share that social
identity. People identify with a party, then adopt the ideology of the group to minimize
in-group conflict or personal inconsistency. Those who have weak or moderate
attachments towards a specific political group will be less likely to act as a cyber Balkan
than those who feel a strong allegiance to their group. I predict to find that those who
identrfil themselves as “strongly” conservative or liberal will be more likely to engage in
more cyber Balkanization than will moderates.

Personglitv. Persongl Motivations, and Demoggtphic Factors

Other research has not found meaningfirl differences in online balkanization and
most demographic factors (Bimber 2003). Nevertheless, I will see if there are any
important relationships between cyber Balkanization with regards to four important
demographic categories: age, race, gender, education. However, I predict that there will

be no relationship between these four demographic factors with Cyber Balkanization.

136

Nevertheless, in regards to age, there may be at least some reason to believe that younger
online users will be more prone to cyber Balkanization than others. Young people have
been found by some researchers to be more likely to identify with like-minded
communities and reject classic definitions of community (Galston 2002; Putnam 2000).
Additionally, younger people are said to be more pessimistic about the government
(Owen 2006) as well as the traditional news media (Bimber 2003). These factors have the
potential to drive younger users towards like-minded others.

Personality factors are also expected to be related to online self-selection, namely
the person’s self-construal and the degree of closed-mindedness, or Dogmatism. In
regards to the former, interdependent (collectivist) self-construal type values choice less
than others and will be more opposed to the disharmony of group consensus (Iyengar &
Lepper 2002), making them more likely to practice online Balkanization. Similarly,
research on personality factors has shown that people differ in terms of Dogmatism, or
how much open or closed-mindedness they exhibit in their thinking (Goldberg 1992,
1993). It is predicted here that those who score high in dogmatism may be predisposed to
selectivity, and are self-motivated to avoid a complex marketplace of ideas online or
offline. This makes them more likely to be cyber Balkans than more open-minded people
are, as they are more uncomfortable than others with ambiguity and uncertainty in their
political views.

Research has shown that argurnentativeness is an important dimension of
personality (Goldberg 1992, 1993). People vary greatly in terms of their willingness to
engage in an argument. Thus, lntemet users that like to argue are less likely to engage in

selective exposure patterns than are those who avoid arguments. My prediction is that

137

those who do not enjoy arguments are more likely to cyber Balkanize.

In regards to feelings of certainty, it is already established that certainty is an
important factor is selective exposure. Those who feel confident in their beliefs are more
likely to engage in cyber Balkanizing behavior (Bimber 2003). Although certainty in
political beliefs may overlap significantly to the strength of their political identity
measures, this dimension will be tested as if the two are not. In general, I predict that
people who feel very certain about the correctness of their political stances will be
positively related to cyber Balkanizing behaviors online.

Likewise, people may only take interest in politics online in a very narrow-casted
way, only following one key personal issue that is considered personally important. If a
person attends only to stories that are of interest to them, they are likely to be surrounding
themselves to similar others, whether intentional or not, which is bound to effect the
strength of those attitudes by consistently engaging with people who share similar values
and understanding on single topics. Thus, those who are selective to only one topic of
personal interest (show high levels of single-issue interest) are likely to show high levels
of C yber Balkanization.

Emotions

I expect that there is overlap between the strongly partisan and feelings of anger
towards the party of political opposition. Some have shown how current US political
identities are often rooted in affective attachments to a social identity rather than specific
policies. Political party affiliation in the US today often has more to do with social
identity and social comparison than ideology (Bishop 2008; Dowd et al 2006) where

people "pick a team" to compare the self to. People identifying themselves as politically

138

liberal or conservative often “adopt an appropriate ideological script” on political issues,
which minimizes in-group conflict but increases out-group conflict (Skitka & Mullen
2002: 122). Additionally, the more dislike for opposition party, the more consistent
ideologies they will hold, as ideologies may be a product of the desire to maintain
affective consistency (Skitka & Mullen 2002).

In short, those who identify strongly with a particular group are likely to develop
hostile feelings towards other political parties, and the self is then more motivated to be
surrounded by like-minded others. Therefore, it is predicted that those who feel anger
towards people who don ’t share their political values will be more likely to segregate
themselves, having a positive relationship to cyber Balkanization.

It is also likely that the desire to Balkanize is be related to the stresses of modem-
day life. Some people may self-select which issues they attend to because they wish to
avoid more aggravations in their Iifiz, such as having a desire to avoiding arguments and
criticism, while desiring bolstering and support from others who share their social
identities (Bishop 2008).

As a final point, it is believed that time spent online is not related to being a cyber
Balkan. It is proposed here that personal factors drive self-selectivity and Balkanization,
not merely the amount of time spent on the computer. To show this, I have measured this
as a variable, but predict that time spent online is not related to Cyber Balkanization.
Methodology

Data used to test these hypotheses were collected using an online survey during
February, 2009. This survey purposively sampled those individuals who were members

of either a political or non-political facebook.com group. Approximately 30 separate

139

facebook.com groups were contacted and each member was personally invited to take
this survey by clicking on a link which took them directly to the website that hosted the
survey. During this time period, 624 different respondents completed the survey. The
results were recorded by a website provider named zapsurveycom and were coded by the
author upon the respondent’s completion of the survey.
Dependent Variable Construction and Measurement

Cyber Balkanization, the dependent variable being used, was found by me to have
more than one dimension when using factor analysis with direct oblimin rotation. The
following table shows that the five questions I used to measure Cyber Balkanization in
my survey yielded two separate dimensions (see table 27 below): one that more or less
describes emotions and motivations of the user, and the other describing the values and
actions of the respondent.

Table 27: Factor Analysis Results for the Variable "Cyber Balkanization"

 

 

 

 

 

 

 

 

 

 

Total Variance Explained
Extraction Sums of Squared Rotation Sums of Squared
Initial Eigenvalues Loadings Loadings

nil % of Cumulative % of Cumulative % of Cumulative
Compone Total Varlance % Total Variance % Total Variance %
1 1.7461 34.914 34.914 1.7461 34.914l 34.914 1.651[ 33.029 33.023
2 1.065 21.300 56.214 1.065 21.300 56.214 1.159 23.186 56.21
6 .843 16.867 73.081
4 .808 16.160 89.241
'5 .538 10.759 1oo.oool

 

 

 

 

 

 

 

 

 

At this point, I broke the dependent variable into two separate factors: the Motives
of Cyber Balkanization (“CB Actions ”) which is comprised of the three items which
loaded onto component one below (Comfort with similar others + Feeling Uneasy around
difference + Change Topic to avoid arguing), and attitudes about what constitutes a

community (“CB Community ”) which is comprised of the items “A Community

140

Comprised of Shares Values” and “People Should Leave Group if they Differ from the

Others”. These factor loadings can be seen in table 28, and lead to my justification in

determining which items aligned with one another.

Table 28: Rotated Component Matrix

 

 

 

of Cyber Balkanization
Rotated Component Matrix
Component
1 2
Community -.098 .831
comprised of
Shared Values
With .668 223+

Slmllar Others
Feel uneasy .821 .071
differences
Change Topic .682 . -

Avoid Argue
Share values .235

leave group

 

 

 

 

 

In each variable, scores were added together to form a 3-item scale measure of

“CB Actions” (where values range from -6 to +6) and a 2-item scale of “CB

Community” (where values range fiom -4 to +4). Throughout the remainder of the

analysis conducted in this chapter, the independent variables discussed earlier will be

regressed each dependent variable separately. Table 29 describes the mean, standard

deviation, and range for each of these measures. These two measures were also found to

have a low and non-significant correlation to each other (r = .042, P = .292).

Table 29: Descriptive Statistics for both measures of Cyber Balkanization

 

 

7 Descriptive Statistics , -
E Range I Minimum Maximum Mean l Std. Deviationl
Cybelbalkan by Motives 624 16 -8 8 -1.35 2.82
Cyberbalkan by Actions 624 8 -3 5 1.10 1.
alid N (listwise) 624

 

 

 

 

 

 

141

Predicted Relationships to Cyber Balkanization

Table 30 is a summary of the predicted relationships of each independent variable

discussed earlier and how each was measured.

Table 30: Predicted Independent Variable Relationships to C ber Balkanization

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Measurement of Variable Predicted Relationship to
Obtained by using... Cyber Balkanization
Perceived like—Minded Friends Question “My friends share Positive
mmlitical Views”
Perceived like-minded Family Question “My family shares Positive
my political values”
Perceived living in a like- 2 Survey Questions (r=.301") Positive
minded city
Trust in Others 2 Survey Questions (r=.465‘f) Negative
Trust in Government Question “I trust U.S. Positive
government to usually do the
right thing”
Trust in the News Media 2 Survey Questions (r=.43 l **) Negative
Conservative Political Identity Dummy Variable (Eysenck Positive
Scale) where l = Conservative
Strength of Political Identity Eysenck Political Identity Positive
Scale (6 Questions; One—
Dimensional)
Certainty in correctness of 2 Survey Questions (r=.581") Positive
Political Beliefs
Feeling Anger towards 2 Survey Questions (r=.472**) Positive
litical opponents
Age 1 Open-Ended Question Negative
@ontinuous Measure)
Race (White or Non-White) Dummy Variable(White= 1) None
Gender Dummy Variable(Male = 1) None
Education Level of User Dummy Variable (is None
respondent a college
graduate?)
User has lnterdependent Singelis Self-Construal Scale Positive
(Collectivist) Self-Construal (3 Questions; One-
Dimesnional)
User avoids arguments Infante and Rancer Scale (5 Positive
Questions; One-Dimensional)
User seeks bolstering and 2 Survey Questions (r=.343") Positive
support in views
User desires to avoid Criticism 2 Survey Questions (r=.345") Positive
Dogmatism Rokeach Scale (5 Questions; Positive
One-Dimensional)
Time Spent Online Each Day 2 Survey Questions (r=.762”) None
Selectivity in issues followed 3 Survey Questions (One- Positive

 

dimensional)

 

 

142

 

Wherever possible, multiple questions were used to measure each independent
variable. However, three of these variables (Perception of having Like—Minded Friends,
Perception of having Like-Minded Family, and Trust in the Government) would not
correlate strongly or significantly to the other questions that were used in my survey to
form a scale measure for these variables. In these instances, I am forced to rely on single-
item measures. The construction of all other independent variables are comprised on
multi-item scale measures, where correlations among 2-item scales are listed in the table
below and factor analysis conducted on larger scales all yielded one-dimensionality of the
underlying variable. A more detailed description which illustrates precisely what
questions and scoring methods were used to calculate each variable is shown in Appendix
A at the end of this dissertation.

Regression Results: “CB Actions” as the Dependent Variable

Table 31 shows the regression results when predicting Cyber Balkanization as
defined by actions and motives. We can see from the data that variables connected to the
personal emotions and motivations of the user, such as holding feelings of anger towards
political opponents showed a positive and significant relationship (Beta = .248, P < .000)
to the dependent variable. We found the same with respect to maintaining a desire to
avoid criticism fiom others (Beta = .141, P < .000). Although the relationships are of
moderate strength. these findings indicate that users who feel anger towards political
opposition are more likely than other people to engage in Cyber Balkanizing behavior
online, as are those users who wish to avoid criticism from others when expressing their
viewpoints. Both of these points are consistent with the predicted relationships discussed

earlier. These results suggest that cyber Balkanization is highest among users that don ’t

143

want to argue over whether their feelings and viewpoints are correct, and is particularly
strong for people who feel disgust towards people whose political viewpoints difierfi'om
their own.

Cyber Balkanization was also found to have a modest positive and significant
relationship with single-issue selective news followers (Beta = .109, P < .000), indicating
that those who only have an interest in topics that are personally relevant or interesting to
them are likely to avoid arguments or contrary viewpoints. There seems to be a
relationship between selectively choosing which topics to engage in and selectively
deciding whom to engage in regarding online political discussion. This is also consistent
with predictions made earlier. Additionally, Cyber Balkanization was linked to the living
conditions and social networks that people are within offline. Those who felt that their
friends (Beta = .173, P < .000) and their families (Beta = .074, P = .041) share their
values were more likely to engage in cyber Balkanization. Although these effects are
somewhat weak, they are significant an in the direction that was predicted.

The results in table 31 also show that there are personality characteristics which
drive online Balkanization. Two personality factors were shown to reduce Cyber
Balkanization: The somewhat weak effect of having a Collectivist Self-Construal (Beta =
-.088, P = .009) and the strong effect of Argumentativeness (Beta = -.346, P < .000). The
third personality variable that also showed a positive but weak relationship to Cyber
Balkanization is Dogmatism (Beta =.085, P = .024), indicating that those who dislike
complexity and are uncomfortable with uncertainty are more likely to engage in online

Balkanization and self-selectivity.

144

Table 3]: Regression Results: Predicting Cyber Balkanization as defined by Actions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Coefficients
Unstandardlzed Standardized Colllneal'lty
Coefficients Coefficients Statistics
Std.
Model B Error Beta 1 Sig. Tolerance VlF
1 (Constant) -.961 .458 -2.099 .036
Friends Share Views .424 .088 .173 4.790 .
Family Shares Views .154 .075 .074 2.047 .041
Trusts Government .136 .071 .068 1.922 .055
Trusts Public -.021 .055 -.01 3 -.380 .7
Tnlsts Media .022. .058 .014 .379 .705
Is respondent a -.468 .252 . -.076 -1 .854 .064
conservative?
Strength of Political .006 .031 .006 .181 .857
ldenfity
Feels Certainty in Political -.006 .135 -.002 -.048 .962.
Beliefs
Feels Anger towards 1.028 .146 .248 7.029 .000
0 1’
Age -.008 .010 -.030 -.855 .393
Is respondent White? .122 .253 .016 .482 .630
Gender .247 .192 .044 1.288 .198
Is respondent college .082 .195 .014 .419 .675
gmduate?
Collectivist Self Construal -.080 .030 -933 -2310 .009
Argumentative -.244 .026 -.346 -9.376 .000
Dogmatism .103 .046 .085 2.263 .024
Tlme Spent Online -.056 .052 -.036 -1.078 .281 .905 1.1
Single Issue Selectivity .116 .037 .109 3.157 .002 .857
Lives in Like Minded City -.075 .054 -.049 -1.372 .170 .801
Avoids Criticism .604 .149 .141 4.054 .000 .845
Bolsters Viewpoints .242 .145 .061 1.673 .095 .780

 

Dependent Variable: Cyber Balkanization defined by Online Actions
R-Square = .381; Adjusted R-Square = .360

 

Durbin-Watson Statistic = 1.933
F-Statistic = 17.669"

This finding is certainly consistent with my predictions. Also of interest is that the

variable “Time Spent Online” has no significant correlation to Cyber Balkanization (the

145

online actions of the lntemet user) (Beta = -.036, P = .281). This indicates that time spent
online, in and of itself does not make one more self-selective. In other words, spending
time on the lntemet leads one to avoid online arguments or steer themselves towards
interactions that reinforce their own viewpoints.

While table 31 indicates there are numerous personal motives, personality
dimensions and offline living environments that are related to Cyber Balkanization, there
are a number of interesting characteristics that are apparently not related. First, the results
show that none of the political identity or strength variables are significantly related to
Cyber Balkanization. It seems that online selectivity is not a function of being either
conservative or liberal, nor is it a product of having strong political identities, or even
feeling highly certain that your political views are correct. These variables in and of
themselves are not enough to constitute online self-selectivior. An online political
partisan who feels strongly about his/her political views is not necessarily going to avoid
counter-views online and surround themselves by like-minded people.

In the final part of analysis within this chapter, the dependent variable is redefined
based upon the findings of factor analysis conducted earlier. All relevant findings will be
discussed and the implications of these findings will be illustrated.

Regression Results: “CB Community” as the Dependent Variable

This section considers and measures Cyber Balkanization as a tendency or a
person to define a community in terms of being a collection or social network of like-
minded rather than other common and traditional definitions. Table 32 shows that a very
strong and positive relationship with Cyber Balkanization exists with those who feel like

they live in a like-minded community (Beta = .541, P < .000). The size of the

146

standardized beta here is surprisingly large, and indicates that those who live in cities
where they feel most people share their viewpoints believe that people who do not share
the same viewpoints as other community members should leave it. People that live in
cities where other people mainly share their views seem to believe that a community is
comprised of people who share your views, whether they live close to you or not.

Again, this indicates that there is a strong amount of overlap between online and
offline Balkanization. When people Cluster together and share like-minded views, they
tend to become more intolerant of dissention and lose empathy for people who do not
share their viewpoints, even to the point of not considering those people as part of their
communities. Instead, people fiom like-minded areas seem to define community as
something that is personally selective and reinforcing, likely creating more extreme
attitudes and values over time.

The results from table 32 also indicate that there are four other personal
characteristics and motivations that are related to how people define their community in
terms of selectivity: A desire to avoid criticism from others (Beta = .292, P < .000),
Collectivist Self-Construal (Beta = -.O68, P = .031), Argumentativeness (Beta = .082, P =
.018), and Age (Beta = .065, P = .050). Other than living in a like-minded city, the desire
to avoid criticism from others was the strongest factor in the tendency to define
community by those who share their views, which is consistent with my predictions.

The final two variables related to the community definition of Cyber
Balkanization, Argumentativeness and Age, are hard to interpret. In regards to the
former, the data indicates that those who like to argue are more likely to be self-selective

in regards to who they consider to be part of their community, which seems counter-

147

intuitive. It may be the case that while people may enjoy arguing with dissimilar others,

these people still do not consider those holding differing views as part of their community

 

 

 

 

at the end of the day.
Table 32: Regression Results: Predicting Cyber Balkanization as idea of Community
Coefficients
Unstandardlzed Standardized Colllnearlty
Coefficients Coelficierlts Statistics
Toleranc
B Std. Error Beta t Sig. e VIP
1 (Constant) .222 J .251 .376
Friends Share Values .004 .049 .003. .078 .938 .791 .
Family Shares Values -.074 .041 -.060 -1 .792 .074 .786 .
Trust Government .067 .039 .057 1.725 .085 .88! .
Trusts Public .018I .030 .019 .582 .581 .814 .
Trusts Media .007 .032 .008 .234 .815 .7481 1.33
ls respondent a -.121 .138 -.034 -.877 .381 .605. .
conservative?
Strength ofPolitical Identity .023 .017 .045 1.3481 .178
Political Certainty in .131 .074 .063 1.771 .077
Beliefs/Values
Feels Anger towards -.156 .080 -.064 .051 82 1.21
Opposition
Age .011 .005 .065. .050 .814 1
Is respondent White? .009 .139 .002 .949 928 1 07
Gender -.173 .105 -..052 .101
Is respondentcollege -.166 .107 -.050 .120 877 1 141
graduate?
Collectivist Self Construal -.0361 .017 -.068 .031 901 1 11
Argumentative .034 .014 .082 .018
Dogmatism .028. .025. .040 .257 .727 1 37
Time Spent Online .046 .028 .051 .105 .905 1 1
Single Issue Selectivity .003 .020 .005 .888 857 1 16
Live in Like Minded City .483 .030 .541 .000 .801 1 2
Avoids Critidsm .733 .082 292 .000 .845 1 1
Bolsters Viewpoints .130 .079 .055 .102 .780 1 28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dependent Variable: Cyber Balkanization as Definition of Community
R-Square = .462; Adjusted R-Square = .443
Durbin-Watson Statistic = 2.063
F-Statistic = 24.638"

148

 

More research could be fruitful in determining whether this relationship is valid or
not. Likewise, the results also show that older people are more likely to define their
community in terms of including people who share their own values. Yet, there is no
theoretical justification that I can find to indicate why that would be true. Overall, I am
skeptical of this finding, and would suggest that more research be conducted before
accepting this relationship as valid.

Nevertheless, table 32 also indicates that several dimensions discussed earlier are
not significantly related to defining community in self-selective ways. First, there were
no significant relationships between a user’s political identities and Cyber Balkanizing.
Similar to findings fi'om earlier, these variables in and of themselves are not enough to
cause online self-selectivity. An online political partisan who feels strongly about his/her
political views is not necessarily going to avoid dissimilar people online or ofiline, and
avoid considering difierent others as part of their community. Additionally, time spent
online had no significant correlation to Cyber Balkanization as an idea of community
(Beta = .051, P = .105). This indicates that time spent online does not make one more
community exclusive. It cannot be said that the amount of time spent on the lntemet leads
a person to redefine who he/she considers to be a part of their community.

Discussion and a Summary of the Findings

While this chapter has focused on the causes of cyber Balkanization, others have
already commented on its social consequences. Personal choice combined with social
fragmentation leads some users to constantly engage in politics online, while others may
never encounter any politics online via their favorite websites and bookmarks (Norris

2002). This is troublesome because the like-minded citizens of cyberspace may develop a

149

type of civic life that involves little or no exchange among those who hold different
opinions, reinforcing fragmentation and factionalism (Margolis & Resnick 2000).

The hypermedia of the web can create more social discourse and fragmentation,
preventing any centralized construction of consensus and unification (V acker 2000).
Additionally, a system of limitless choices is not necessarily in the interest of self-
government and democracy. When people’s preferences are either a product of
excessively limited options or when the preferences of these past choices leads to the
narrowing of their own horizons (Sunstein 2007). As these majorities grow within
communities, whether they be online or offline, minority beliefs retreat from political life
and fuels divisive national politics in a bottom-up fashion (Bishop 2008).

William Galston (2002) warns that we should not organize our lives around the
fantasy that community entrance and exit can always be cost free or that we can wall
ourselves off fi'om those who are different without paying a long-term social price. The
potential breakdown of common experiences and shared experiences via fragmentation
makes it more difficult for the nation to address social problems via misunderstanding
and enmity (Sunstein 2007). However, this need not be the future of political discourse in
the US. Balkanization, whether offline or online, can be reduced by more heterogeneity
in groups. Exposure to competing political arguments can reduce intolerance and feelings
certainty in the correctness of our own arguments (Stmstein 2007).

In this chapter, I have shown that cyber Balkanization is not a universal process
which is experienced among all Internet users. Rather, the personal characteristics and
motives of different users drive people to engage in variant levels of Balkanization.

Additionally, time spent online does not lead pe0ple to become more intolerant. Instead,

150

I up _____..- use "aw—“J

an identification of the conditions which motivate desires to seek homogeneity and
community selectivity can reduce political intolerance, aiding public discourse.
Admittedly, capturing a concept such as cyber Balkanization is not easy to do
using an online survey. The results in this chapter could be bolstered by using other
methods that could also measure this same concept (such as participant observation,
sometimes referred to as ‘Triangulation of methods’ (Denzin 1979). In doing so, research
could show other processes that make identities become salient online, and illustrate the
levels of hostility one faces within homogenous groups when voicing dissent online.
Similarly, experiments could be conducted to see whether people engage in selectivity as
the number of options are increased, or decreased in the presence of similar others, etc...
While this chapter looks to see how offline identities and environments overlap
with Internet activities, more research is needed to document other ways in which there
are connections between these two environments. The results here show that there are
relationships in online behavior that have offline consequences and vice versa, as these
two spaces seem to continually reinforce each other. This is a claim that future research
may be able to better illustrate. Additionally, future works can extend beyond politics to
show how social networks can become homogenized in all realms of social life and
organizations.
Findings Summary: Cyber Balkanization defined by Actions of the lntemet User
The findings within this chapter seems to support theorists such as Bill Bishop
(2008) who see political self-sorting being a result of close physical clustering of like-
minded people, taking place offline and online as well. The analysis here shows that

those living in politically homogenous communities in their physically proximate face-to-

151

face environments also like to be surrounded with like-minded others when they use the
Internet. When a person lives in an environment where the vast majority of primary
groups members share a political viewpoint, this person becomes an online cyber Balkan.

While it has been suggested earlier that cyber Balkanization may be higher for
people who have a collectivist Self-Construal due to a desire to maintain harmony and
avoid conflict, the results show that lntemet users who have a high level of collectivist
are less likely to engage in self-selectivity when discussing politics online than others.
The finding not surprisingly indicate that those who enjoy arguments are less likely to
engage in cyber Balkanizing behavior, and are more likely than other users to engage in
online debates or challenge their own views with others who feel differently.

Also, cyber Balkanization did not correlate significantly to any specific
demographic factors. An Internet user’s age, race, gender or education levels did not
make them any more or less likely to avoid dissenting viewpoints online. These findings
are consistent with my predictions, and have also found support from the work of Bruce
Bimber (2003) as well. Additionally, the results show that there is no relationship
between Cyber Balkanization and time spent online. This rejects notions that time spent
online will lead to some kind of natural clustering of like-minded people, and that the
lntemet “makes” a person less tolerant of dissention or more self-selective due to
exposure to the Internet over long periods of time each day.

Finally, there were no significant relationships between levels of trust and cyber
Balkanization. While it lmd been suggested earlier that distrust of others would lead
people to be more selective in regards to who they interact with, this was apparently not

the case. Being distrusting of the media, the government, or the general public did not

152

significantly affected levels of online selectivity and any avoidance of online debate.
Findings Summary: Cyber Balkanization as defined by self-selective community

In general, People who like to voice their opinions on the Internet while avoiding
criticisms seem likely to be opposed to criticism in their communities as well, whether
online or oflline. These findings suggest that Balkanizing behavior is strongly related to
people who dislike dissent and are motivated to avoid defending their viewpoints fi'om
others who question the validity of their claims; that these people engage in high amounts
of selectivity in whom they consider to be members of their communities and life-worlds.

This community definition of Cyber Balkanization has a negative relationship to
Collectivist Self-Construal, which is not surprising although it is not what I predicted to
find. It is possible that collectivists are more likely to be inclusive of others, and are less
likely to ostracize or exclude others from their sentiments due to holding differences in
political attitudes and values. In this sense, collectivists have a definition of community
which has more empathy and compassion, while seeing an importance to sharing a
connection to people even if differences in values are strong.

Also similar to earlier findings, there is no relationship between the amount of
time spent online and the tendency to self-select their commrmity based on shared values.
While some theorists may have been concerned that spending time online may be
destructive to communities and lead to a lack of civility towards others (Putnam 2000),
spending time online in and of itself seems to be insufficient in terms of self-selectivity
and intolerance of heterogeneity within their social networks and communities. The

lntemet is neither wholly ‘good’ nor ‘bad’ for communities. Of relevance here are the

153

personal characteristics of the users and the oflline environments that he/ she find
themselves within.

Finally, the tendency to self-select into like-minded communities was not found to
have any relationship to trust levels. Cyber Balkanization is not a phenomenon of the
angry or paranoid who suffer from high levels of distrust. The data shows that people
who trust the news media, the US. Government, and the general public are just as likely
to be Cyber Balkans as those who do not. Trust does not seem to be related to either

conceptualization of Balkanization in this study.

154

 

 

CHAPTERS '9
Findings, Limitations, and Directions for Future Research

Introduction

In the final chapter of this dissertation, my goals are to summarize the main
findings and to discuss some of the problems that were encountered. Four important
findings are identified and briefly discussed. Next, some of the limitations of this
dissertation will be discussed, largely revolving around methodological issues pertaining
to the online survey and content analysis interpretation. At the end of this chapter,
discussion will focus upon how these problems could be reduced in future research will
also specifying where future research interests should be directed.

Summary of Main Findings

The data analysis shows at least four main conclusions that contribute to our
understanding of computer-mediated political communication in the US. The first of
these findings is that difl'erent news sources dtflered significantly in terms of political
coverage. The US Presidential race of 2008 shows that mainstream online news sources
favor portraying the political landscape as a horserace, relying heavily on direct candidate
quotes and polling numbers as the staple of their daily political coverage. Nevertheless,
these sources were less likely to show overt political favoritism or bias towards a
candidate, and portrayed the election in more civil terms than bloggers did.

Blogs openly questioned the credibility of the candidates and questioned their
claims, but in a very predictable fashion. Conservative blogs overwhelmingly relied on
negative character portrayals of Barack Obama, and virtually never questioned the claims
made by the GOP. Likewise, liberal blogs tended to discuss personal character of the

candidates more than the issues, and in a direction which favored the Democratic Party.

155

More than any other sources, the blogs criticized mainstream news coverage and hurled
accusations of bias towards journalists and media organizations, which the candidates
strongly avoided doing on their websites.

The websites of the candidates focused almost exclusively on issues and personal
character. These sites were a constant battleground for both candidates, as they
continually attempted to construct public perception of themselves and their opponent.
Barack Obama spent more time discussing the issues than John McCain did on his
website, yet still made plenty of efforts to portray McCain as an undesirable choice. At
the same time, John McCain and the GOP Party website spent an enormous amount of
time trying to discredit the claims of Barack Obama and illustrate flaws in his character
and judgment.

It is easy to imagine that, given where an lntemet user received the bulk of their
online news, the beliefs and perspective of different users would diverge widely. Blog
readers probably feel more hostility and would probably exhibit more extremism in their
viewpoints than others. Whether reading a highly partisan blog is the cause or effect of
politically extreme attitudes is hard to determine, given that Internet users drawn to blogs
are likely the most partisan to begin with. Nevertheless, firrther research could better
document both the positives and negatives that political blogs contribute to levels of
political civility, and if these news sources provide an overall benefit to the public sphere.

The second major finding of this dissertation is that user identities and personality
variables aflect Internet usage. What this dissertation shows is that people who are
interested in politics use the Internet to find political news and to get involved. Likewise,

people who lack strong identification with political parties or are uninterested in politics

156

may still spend much time online. However, they are not anymore politically
knowledgeable or connected by doing so. Their attention is drawn elsewhere, satisfying
other needs. These findings are consistent with the “Uses and Gratifications” theory of
media usage discussed in chapter three. People use media instruments to fulfill their
personal interests. Because of this, it is nearly impossible to show that the Internet will
have an overall positive or negative efiizct on political discourse. Some users will become
more politically active and informed, while others will ignore politics altogether.

While specific identities related to political activity on the lntemet were found to
be important predictors of online usage, overall self-construal did not. There does not
seem to be any evidence here that individualists are more likely to use computers than
other people, or that collectivists are more or less likely to seek political connections on
the Internet. Although strongly held political identities and personal characteristics seem
to account for diflerent patterns of online behavior and attention, overall tendencies of
how one views him/herself seem to be irrelevant. Nevertheless, future research should
continue to identify these important personal characteristics, and document how different
types of lntemet users will vary in their usage patterns.

Third, the data also indicated that the total amount of time spent online showed
mixed results with regards to online political behavior. Chapter three showed that the
amount of time spent online had small but positive and significant relationships to
political activity levels or political knowledge seeking on the Internet. However, Chapter
Four indicated that time spent online did not have any significant relationship to cyber
Balkanization. It appears that high levels of Internet usage does not necessarily make

people less civil or more intolerant of opposing vieWpoints. Interestingly, chapter three

157

W_‘" :sjw J
ll n

found that there was a small but significantly positive relationship between political
liberalism and time spent online. Although these findings may be an anomaly based on
this sample, future research might better show whether the Internet is a catalyst of liberal
political activity in the US. Likewise, there may be identifiable personality differences
between conservative and liberals that might induce differential rates of online political
exposure patterns, such as having a greater interest in learning to use new communicative
technologies, for example. While these ideas are beyond the focus of this dissertation,
future research may better illustrate how conservatives and liberals vary in terms of how
much time they spend online.

Fourth, this dissertation shows that there appears to be links between online and
oflline balkanization. Patterns of selective exposure to information and the desire to seek
out communities of like-minded people seem to be based in offline living environments
and personality characteristics of the lntemet user. It does not seem accurate to say that
the lntemet act as a boon to local communities, nor is it universally separating all of its
users into a social world full of like-mindedness. However, people who live in strongly
like-minded communities ofiline, seem to desire the same in cyberspace. Personal
motivations and characteristics also seem to drive selective exposure patterns online. The
findings in chapter four indicate that those who don ’t like to argue and are closed-minded
are more likely to act as cyber Balkans, as are people who feel much anger towards
political opposition.

Again, the presence of the lntemet by itself is not a catalyst for Balkanizing
behavior patterns. Even though the worldwide web is a space where a user can maximize

personal preferences and must engage in at least some level of selectivity, these patterns

158

T

do not necessarily lead to decreased levels of civility in the public sphere. Extremism in
political attitudes can also be formed in offline environments and social circles, which the
Internet user brings with him/her when they use their computers. Overall, the data
suggests it is unfair to blame the Internet for rising levels of political extremism or
incivility. Findings in this dissertation show that extreme political attitudes are not related
to the amount of time spent online. Additionally, both personal characteristics and offline
social factors were shown to be related with selectivity patterns where peOple avoided
dissent in their online political communications. Yet, it is important to recognize that the
lntemet has the potential to reinforce these patterns of ofiline behavior. Futme research
should focus on identifying other offline environmental factors and personality
characteristics that could lead to patterns of selectivity and intolerance, and how these
variables could potentially interact with each other.

At this point, a brief discussion about the limitations of this dissertation should be
discussed. The next section is intended to point out some of the difficulties that were
encountered in obtaining the data, including some of the potential pitfalls involved in
measurement and sampling. During this discussion, I will point out ways in which these
problems could be averted in future research.

Limitations of the Dissertation

This dissertation used two different data sets: a collection of 537 online news
articles that were subjected to content analysis, and an online survey which had 624
respondents. For the sake of clarity, it should be pointed out that the content analysis

served as the basis of my findings in chapter two, while the survey responses were

159

analyzed in chapters three and four. The issues encountered concerning each data set will
be discussed separately here.

As with any content analysis research project, much of the data classification was
subject to the interpretation of the reader. First, there were a number of cases where
frames overlapped, and I had to rely on my judgment to classify the article. For example,
it was sometimes unclear whether an article focused mainly on defining the Obama
campaign positively or the McCain campaign negatively. Whenever multiple frames
were used within the article, I often relied on the headline or photos used in the article as
a sort of tie-breaker. My justification for doing so is that these images and text are likely
to impact the reader’s interpretation and cognitive processing. While this seemed to help
reduce uncertainty, other researchers could potentially interpret the information
differently.

Second, these articles were only a small sampling of all online content over a six
week time span, starting in early September and ending the night before the election. This
chapter did not intend to be an exhaustive examination of all political content posted to
the lntemet during this time period. Instead, I relied upon a selective sampling of the
most viewed political information posted online during the US Presidential election.
Additionally, some somees (such as cnn.com) offered a multitude of political content
simultaneously. I addressed my attention towards selecting the first political news story
that caught my eye when I accessed each site. It seems plausible that others would have
selected different stories, or browsed around each site to select the most in-depth story,
for example. Other analysis may have also tried to sample multiple stories within a

source that offered multiple stories, as another example. Nevertheless, I feel that my

160

selection of articles give a valid sampling of online news content during the election
season.

The online survey proved to be a difficult endeavor that led to a number of
concerns. First, the extensive number of variables I tried to measure resulted in a survey
which was probably too long. Although practice surveys indicated that the average time
spent filling out the survey was only about eleven minutes, I suspect that the length of the
survey depressed response rates, a concern I will revisit later. Additionally, my concern
about the length of the survey caused me to rely on fewer scale items and measures than I
would have preferred By trying to measure dozens of variables in an online survey, I
found myself trying to measure important concepts through the use of five or less
questions in virtually all cases. As I reduced the number of questions in a concern about
low response rates, I sometimes found myself left with few measurement scale items
which did not correlate well. This sometimes resulted in a couple of variables whose
measurement relied upon one question responses, as can be seen in appendix B. Variables
such as “Trust in the Government” and “Perceptions of living within a like-minded
community” were subject to this problem.

Variable measurement also proved to be difficult for two important concepts
within this dissertation: ‘Self-Construal’ and ‘Cyber Balkanization’. In regards to the
former, the Singelis (1994) self-construal scale was found to be multi-dimensional, which
led to difficulties in the interpretation and measurement of this important variable. This
scale is not recommended for future research, as responses to this scale turned out to be
very difficult to interpret, and of questionable validity. Measurement of the latter, cyber

Balkanization relied upon a number of untested survey questions that were written by the

161

 

author, as no well-known scale to measure this concept exists. While I don’t consider the
measurement of this variable to be severely flawed, better quantitative measurement of
cyber Balkanization (and other forms of Balkanization) requires the development of a
more reliable and sophisticated index measure.

Sampling also proved to be difficult on three accounts. First, facebook.com would
not permit a random sample to be obtained, stating that this would not be in the best
interests of its users. However, Facebook.com executives decided it would be okay for
me to sample within groups that I belonged to and within groups whose administrators
agreed to participate. While these administrators served as gatekeepers, I was forttmate
enough to persuade sixteen of them to participate in this study. Their generosity allowed
me to send invitations to 7,057 different facebook.com users to take my online survey.

Second, this study suffered from low response rates. Overall, there were 627
responses out of 7,057 invitations sent, or an 8.84% response rate. The refusal rate for
this survey is equal to 1 minus .0884, which equals 91.16% of the overall sampling
fi'ame. However, it is difficult to state these rates with any certainty because ( 1) there is
no guarantee that the emails sent to these individuals was ever received, and (2) people
join and quit facebook.com groups all the time, so there is no clear way of knowing that
the same people were contacted in the follow-ups. Furthermore, the survey was only
available online for one month (until I received at least 600 responses) so it is probable
that those who waited too long to respond or had not been on facebook.com during that
time period could not respond quick enough to participate. In addition, some of the

responses were ineligible for the survey (40 in all) because the respondent was either a

162

—T.

 

non-US citizen or under the age of eighteen. Given the limitations imposed upon me by
facebook.com, there was no way to avoid these kinds of complications.

Third, there seems to be response bias in the sample. Given the number of
questions in the survey and the low response rate discussed above, it seems likely that
this sample consists mostly of those who are the most interested in the topic area. Despite
the use of a personal invitation along with multiple follow-up contacts made by me,
responses were hard to come by. Given the nature of facebook.com as a place where
users like to avoid solicitors and privacy settings, perhaps these low response rates could
not have been avoided.

Nevertheless, I contend that this sample, while imperfect, contains a large number
of somewhat diverse Internet users who are representative of a much larger population.
While the response rates are low and are cause for concern, I point out that the sampling
procedure used here was more of a hybrid between pm'posive and random sampling
techniques, and note that the Internet really does not lend itself to pure random sampling
(Davis 1999). Overall, I feel that this dissertation was still able to adequately locate a
sizable and relevant population to sample. In future research, the hiring of an online
surveying company should be able to sample a more diverse population and be able to
produce higher response rates. In doing so, the results of future research may be better
able to attract those who are less interested in politics and are politically independent,
making the findings more applicable to the US population than they are now.

Directions for Future Research
While chapter two was useful in illustrating how different sources covered

politics differently, no emphasis was placed upon the receivers of these messages to see if

163

7
L

 

the messages were influencing these readers. Also, further research could be conducted 1
involving the selection of which sources lntemet users decided to attend to and which
reports they believed to be true. Experimental research methods could be useful for
showing the ability of online news reporters to persuade the public. The use of
experiments could also see how people complement their online news consumption to the
sources they see and read offline, illustrating commonalities and differences in the
patterns.

In future research, different sources of news could be pitted against one another
within experiments, seeing if some kinds of news sources are more persuasive than other
ones, and under what conditions are different sources of news preferred by the reader?
These kinds of questions remain unanswered, particularly when comparing stories
received by bloggers versus other news stories. For example, it seems plausible that
people read blogs for entertainment purposes and supplementing these perspectives with
mainstream news stories, which they still trust more in terms of reporting objective facts.

Also, while the findings in this dissertation point to some of the personal
characteristics and motives involved in online political activity and cyber Balkanization,
these findings are still preliminary and in need or further investigation. Future research
should also try to identify other personality factors, beliefs, values, and living conditions
which lead lntemet users to be self-selective in regards to politics and community. By
locating such influences, social psychologists and experts within the field of computer-
mediated communication can steer the public towards important improvements of the

public sphere.

164

 

 

Nevertheless, quantitative research will need to find better measurements for key t.
variables such as identity, self-construal, and cyber Balkanization. The development of
better measures for these elusive and abstract concepts would be very fruitful in terms of
the quality of future research. However, research pertaining to cyber Balkanization and
political identities can benefit via the application of qualitative methods as well. In order
to improve the validity of research in this area, qualitative research such as participant
observation could illustrate dimensions and relevance of these concepts despite the

limitations of measurement techniques that this field of research currently faces.

165

 

APPENDICES l
Apgndix A: Content Analysis Coding Sheet used in Chapter Two

C01: ID Number of the document (starting with the number 0001)
C02: Name of the News Source
C03: Code Identification of the news source

10= DailyKos.com 20= powerlineblog.com

ll= huffingtonpostcom 21= lnstapundit.com

12= crooksandliars.com 22== littlegreenfootballscom
13= americablog.com 23= michellemalkin.com
l4= talklefi.com 24= andrewsullivan.com
15= eschaton.com 25= redstate.com

l6= talkingpointsmemo.com 26= vodkapundit.com

l7= smirkingchimp.com 27= wizbangblog.com

18= digbysblogblogspotcom 28= drudgereporLcom

l9= wonkette.com 29= allamericanbloggemom
30= BamckObamacom

3 l== JoeBiden.com

32= Democrats.org

33= MoveOn.org

41= JohnMcCain.com

42= Vice-President Website
43= Republicanpartyorg
44= Townhall.com

 

51= Abcnews.com 61== WallstreetjournaLcom
52= Nbcnews.com 62= Newyorktimes.com
53= Cbs.com 63= Washingtonpost.com
71= Cnn.com 81= Yahoo.com

72= Msnbc.com 82= Msn.com

73= Foxnews.com 83= News.google.com

C04: Month of the Posting
9= September

10= October

1 1= November

C05: Date of the Posting

C06: Does the message attempt to define where an opposing party candidate stands, or to
assert where the supported candidate stand?

0= Does not attempt to define any candidate stand

l= Yes, opposing party defines where an opposing candidate stands

2= Yes, writer pundit defines where a supported candidate stands

3= Yes, multiple sources define the candidate

C7: Does the message allow the candidate do define his stance with Direct Quotes
0= no

166

l= Yes, Obama or Biden
2= Yes, McCain or Palin
3= Yes, Both Candidates

C08: what is the issue of concern?

l= War in the Middle East (issues)

1= Jobs/Economy (issues)

l= Environment (issues)

1= Foreclosure/Debt (issues)

2= Personal Character of Candidate

3 = Political Polling/Strategy (horserace)
4= Other

C9: Does the message try to refute (or respond to) another claim that some other issue
should be at the top of the presidential agenda?

0= No

l= Yes

C10: Is the message personally hostile towards the presidential candidate?
0= No
l= Yes

C11: If so, does the hostile message draw attention to a specific policy as well?
0= No
l= Yes

C12: Does the writer portray candidate an insincere or lying?
0= No
l= Yes

C13: What is the overall flame of the article?
1= Neutral or No Valence

2= Obama Campaign Positive

3= Obama Campaign Negative

4= McCain Campaign Positive

5= McCain Campaign Negative

C14: Does the writer seem skeptical of candidate or policy?
0= Not Skeptical

1= Skeptical towards the candidate

2= Skeptical towards the policy issue

3= Skeptical towards both candidate and policy

167

.5 ‘

rec-m... w. .0:
i
l

 

Apgndix B: Alphabetical Listing of Each Survey Item used

Age

Currently, what is your age?

Argumentativeness Scale (Infante & Rancer)

I enjoy a good argument over a controversial issue

(SA=2, A= l, NS=0, D = -1, SD = -2)

Arguing with people typically creates more problems for me than it solves
(SA=-2,A=-1,NS=0,D=1,SD=2)

I get an unpleasant feeling when I realize I am about to get into an argument
(SA=-2,A=-1,NS=0, D=I, SD=2)

I enjoy defending my point of view on an issue

(SA=2, A= l, NS=0, D = -1, SD = -2)

I consider an argument to be an exciting intellectual challenge

(SA=2, A= l, NS=0, D = -1, SD = -2)

---The responses from these five closed-ended questions were then summed

together to measure the variable. Values of this variable ranged from +10 to -10.

 

Certainty in Political Beliefs

My support for Democrats or Republicans tends to vary from one issue to the next.
(SA=2, A= l, NS=0, D = -1, SD = -2)

I don’t feel that I would ever support any other political party than the one I support now.
(SA=2, A= 1, NS=0, D = -l, SD = -2)

---The responses from these two closed-ended questions were then summed

together to measure the variable. Values of this variable ranged fiom +4 to -4.

Cyber Balkanization defined by Online Actions/Behaviors

I feel more comfortable when I am around people who think similarly to me .

(SA=2, A= 1, NS=0, D = -1, SD = -2)

When someone expresses a political view to me that is different from my own, it makes

me feel uneasy.

168

(SA=2, A= l, NS=0, D = -1, SD = -2)

When confronted to a political view that is different from mine, I prefer to talk about
something other than politics.

(SA=2, A= 1, NS=0, D = -1, SD = -2)

----The responses fi'om these three closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +6 to -6.

Cyber Balkanization defined by beliefs about a Community

I think that a community is a group of people that share your values whether they live
physically close to you or not

(SA=2, A= 1, NS=0, D = -1, SD = -2)

If a person does not share the values of a group, they should leave it voluntarily.
(SA=2, A= 1, NS=0, D = -1, SD = -2)

~---The responses from these two closed—ended questions were then summed
together to measure the variable. Values of this variable ranged from +4 to -4.

Dogmatism (Rohench Scale)

It is only when a person devotes him/herself to an ideal/cause that life becomes
meaningful

(SA=2, A= 1, NS=0, D = -1, SD = -2)

It is never necessary to restrict a person’s freedom of speech
(SA=-2,A=-1,NS=0,D=I, SD=2)

For most questions, there is many possible right answers
(SA=-2,A=~1,NS=0,D=1, SD=2)

Extremely tolerant people probably have few opinions of their own

(SA=2, A=1, NS=0, D = -1, SD = —2)

A group that accepts too much difference of opinion among its members cannot exist
very long

(SA=2, A= l, NS=0, D = -1, SD = -2)

----The responses from these five closed-ended questions were then summed

169

 

 

 

together to measure the variable. Values of this variable ranged from +10 to -10.

Education

Please indicate which of the following describes your current level of formal education.
__ Some High School (=0)

__ High School diploma (=0)

__ Some College (=0)

__ Associates Degree (=0)

__ Bachelors Degree (=1)

__ Some Graduate School (=1)

__ Masters Degree (=1)

__ PostaGraduate or PHD (=1)

-- Education was measured as a dummy variable.

Frequency of Time spent online
How much time per day do you spend online?
How much time per day do you spend yesterday?

---These two open-ended questions were than averaged together to measure the

variable.

Gender: (1 Categorical Measure)
What is your gender?

__ Male (= 1)

__ Female (= 0)

---Gender was used as a dummy variable.
Perceptions that user has Like-Minded Family

I think that most of my family shares the same political values as me
(SA=2, A= l, NS=0, D = -1, SD = -2)

170

 

 

 

Perceptions that user has Like-Minded Friends L
I think that most of my friends share my political values
(SA=2, A= 1, NS=0, D = -1, SD = -2)

Perception that user lives in a Like-Minded Community

I think that most of the people in the city I live in share my political values
(SA=2, A= l, NS=0, D = -1, SD = -2)

Most people in my community think the same way as me most of the time
(SA=2, A= l, NS=0, D = -1, SD = -2)

--The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged fi'om +4 to -4.

Political Activity Levels

I have signed an online petition during the 2008 Presidential election (Yes = I, No =0)

I have used the intemet to help myself participate in an offline political action (such as a
protest) (Yes = 1, No =0)

I have donated money ONLINE to a political campaign in 2008 (Yes = 1, No =0)

I have actively tried to persuade others THAT I MET ONLINE into changing their minds
during the presidential election (Yes = I, No =0)

I volunteered time to a Presidential campaign during the 2008 election (Yes = 1, No =0)

----The responses from these five closed-ended questions were then summed

together to measure the variable. Values of this variable ranged from +5 to 0.

Political Identity of Respondent
The rich are too highly taxed.

(SA=2, A= 1, NS=0, D = -1, SD = -2)
The federal government is too big.
(SA=2, A= l, NS=0, D = -1, SD = -2)

There would be fewer problems in our country if we would eliminate the separation of

church and state.

(SA=2, A= l, NS=0, D = -1, SD = -2)

171

 

—-.~

tram...- . _

Gays and lesbians should be able to marry or at least be treated as married under law if
they so desire.

(SA=-2,A=-1,NS=0,D=1, SD=2)

I believe that the economy performs best when government regulations are kept to a
minimum.

(SA=2, A= l, NS=0, D = —1, SD = -2)

The government has a duty to reduce poverty by enacting programs aimed to transfer
money from the wealthy in order to assist people of modest means.
(SA=-2,A=-1,NS=0,D=1, SD=2)

--The responses from these six closed-ended questions were then summed

together to measure the variable. Values of this variable ranged from +12 to ~12.

Political Information-Seeking

[I go online for political news. (Yes = I, No =0)

I read a political blog at least once a week (Yes =1, No =0)

I read or post comments to a political message board. (Yes = 1, No =0)

I have signed up to receive news email from a political organization (Yes = 1, No =0)

---The responses from these four closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +4 to 0.

Politically Interested

In general, I consider myself a person who is interested in politics
(SA=2, A= l, NS=0, D = --1, SD = -2)

When it comes to political issues, I have very strong Opinions.
(SA=2, A= l, NS=0, D = -1, SD = -2)

--The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +4 to —4.

Race
I would describe myself as being

172

 

 

________ Caucasian/White (=1) L
__ African-American/Black (=0)

__ Asian (=0)

__ Hispanic (=0)

__ Native-American Indian (=0)

_________ Middle Eastern (=0)

__ Mixed Race (=0)

_ Other (=0)

--- Race was used as a dummy variable.

Self-Construal: Collective (Singelis Self-Construal Scale)

(Each question is scored from +3=Very Much like me, to -3 = Not like me at all)
__I will stay in a group if they need me, even when 1 am not happy with the group.
___My happiness depends on the happiness of those around me.

________Jt is important to me to respect the decisions made by the group I am in.

______I respect people who are modest about themselves.

____My self-image largely depends on how close friends see me.

--The responses fi'om these five closed-ended questions were standardized, then
summed together to measure the variable. Values of this variable ranged from +15
to ~15.

Self-Construal: Independent (Singelis Self-Construal Scale)

(Each question is scored from +3=Very Much like me, to -3 = Not like me at all)
____I am comfortable with being singled out for praise or rewards.

_____I enjoy being unique and different from others in many respects.

______1 set a high value on traits that differentiate me from other people.

--Responses from these three closed-ended questions were standardized, then
summed together to measure the variable. Values of this variable ranged from +9
to -9.

173

 

Single-Issue Individual News Seeking

Instead of following all political news, I tend to follow only an issue that is important to
me.

(SA=2, A= l, NS=0, D = -1, SD = -2)

I tend to seek out a lot of news about a few issues that I follow passionately.

(SA=2, A= 1, NS=0, D = -1, SD = -2)

---The responses from these two closed—ended questions were then summed

together to measure the variable. Values of this variable ranged from +4 to -4.

Trust in the Government
You can trust the government in Washington D.C. to do the right thing most of the time.
(SA=2, A=1, NS=0, D = -1, SD = -2)

Trust in the News Media

Whenever the news media reports a story, I feel confident that their facts are accurate.
(SA=2, A= 1, NS=0, D = -1, SD = -2)

There is so much bias in the news that it is difficult to figure out what is really going on.
(SA=2, A= l, NS=0, D = -1, SD = -2)

“The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +4 to -4.

Trust in Other People

Most of the time, people can be trusted to do the right thing

(SA=2, A= 1, NS=0, D = -1, SD = -2)

In general, people will take advantage of you if they are given the chance to do so.
(SA=2, A= 1, NS=0, D = -1, SD = -2)

---The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +4 to —4.
User desires bolstering/support from similar others online

I go online to find information that I can use as ammunition in arguments against others.

174

if,“ ‘i xvi—F. _—.__ J

 

-._ ‘
a

(1=Agree, 0 = Disagree) L
I go online to find out about issues that are affecting people similar to me.
(1=Agree, 0 = Disagree)

---The responses from these two closed-ended questions were then summed

together to measure the variable. Values of this variable ranged from +2 to 0.

User wants to avoid criticism of others when expressing self

I use the internet to say what I want about politics without having to hear the criticisms of
others.

(1=Agree, 0 = Disagree)

The internet is a space where I can say whatever I am feeling without having to endure

censorship fiom others.

(1=Agree, 0 == Disagree)

--The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged from +2 to 0.

User wants to express anger about political opponents
I feel anger towards those who are voting against the candidate I support for president.

(1=Agree, 0 = Disagree)
I dislike those people who oppose my political values

(1=Agree, 0 = Disagree)

--The responses from these two closed-ended questions were then summed
together to measure the variable. Values of this variable ranged fi‘om +2 to 0.

175

 

BIBLIOGRAPHY

Abramowitz, Alan, Brad Alexander, and Matthew Gunning. 2006. “Don’t Blame
Redistricting for Uncompetitive Elections.” Political Science and Politics 39. Available
at http:// www.centerfogaolitics.org/crystalba1l/article/php?id=AIA2005052601.

 

Abrams, Dominic and Michael Hogg. 2001 . “Collective Identity: Group
Memberships and Self-Concept” in Blackwell Handbook of Social Psychology: Group
Processes, Edited by Michael Hogg and R. Scott Tindale. New York: Tindale Publishing,
pp. 425-460.

Adamic, Lada A. and Natalie Glance. 2005. “The Political Blogosphere and the
2004 US Election: Divided They Blog” in LinkKDD ’05 : Proceedings of the 3'4
international workshop on Link discovery; pp. 36-43.

Agger, Ben. 2004. The Virtual Self: A Contemporary Sociology. Oxford:
Blackwell Publishing.

Aikat, Debashis. 2000. "Cyberspace of the People, By the People, For the People:
Predominant Use of the Web in the Public Sector” in Understanding the Web: Social,
Political, and Economic Dimensions of the Internet Edited by Alan B. Albarran and
David H. Goff. Ames, IA: Iowa State University Press, pp. 23-48.

Alterman, Eric. 2004. What Liberal Media? The Truth about Bias and the News.
New York: Basic Books.

Altheide, David. 1985. Media Power. Beverly Hills, CA: Sage Publications.

Amichai-Hamburger, Yair. 2007. "Personality, Individual Differences and
lntemet Use" in The Oxford Handbook of Internet Psychology Edited by Adam N.
Joinson, Katelyn Y.A. McKenna, Tom Postmes, and Ulf—Dietrich Reips. Oxford: Oxford
University Press, pp. 187-204.

Anderson, David M. 2003. “Cautious Optimism about Online Politics and
Citizenship” in The Civic Web: Online Politics and Democratic Values, Edited by David
M. Anderson and Michael Cornfield. Lanharn, MD: Rowman and Littlefield.

Bagdikian, Ben H. 2004. The New Media Monopoly. Boston, MA: Beacon Press.

Bahnisch, Mark. 2006. “The Political Uses of Blogs” ” in Uses of Blogs, edited by
Axel Bruns and Joanne Jacobs, 139-150. New York: Peter Lang Publishing.

Bartels, Larry M. 2000. “Partisanship and Voting Behavior, 1952-1996.”
American Journal of Political Science 44 (1), 35-50.

176

rt: 1?. .

 

 

Baumeister, Roy. 1986. Identity, Cultural Change, and the Struggle for Self L
Oxford: Oxford University Press.

Bell, Daniel. 1999. The Coming of Post-Industrial Society. New York: Basic
Books.

Benkler, Yochai. 2006. The Wealth of Networks: How Social Production
Transforms Markets and Freedom. New Haven, CT: Yale University Press.

Benoit, William L. 2007. Communication in Political Campaigns. New York:
Peter Lang Publishing.

Bimber, Bruce. 1998. “The Internet and political transformation: p0pulism,
community, and accelerated pluralism”. Polity Volume XXXI (1), pp. 133-160.

Bimber, Bruce A. 2003. Information and American Democracy: Technology in
the Evolution of Political Power. Cambridge: Cambridge University Press.

Bimber, Bruce and Richard Davis. 2003. Campaigning Online: The Internet in US
Elections. Oxford: Oxford University Press.

Bishop, Bill. 2008. The Big Sort. New York: Houghton Mifilin Company.

Blumer, Herbert. 1969. symbolic Interaction: Perspective and Method. Berkeley:
University of California Press.

Brewer, Marilynn B. and W. Gardner. 1996. “Who is this ‘we’? Levels of
Collective Identity and Self-Representations.” Journal of Personality and Social
Psychology 71: 83-93.

Brock, Timothy and Joe Balloun. 1964. “Behavioral Receptivity to Dissonant
Information.” Journal of Personality and Social Psychology 6, 413-428.

Bruns, Axel. 2006. “The Practice of News Blogging” in Uses of Blogs, edited by
Axel Bruns and Joanne Jacobs, 11-22. New York: Peter Lang Publishing.

Bruns, Axel and Joanne Jacobs. 2006. Uses of Blogs. New York: Peter Lang
Publishing.

Burke, Peter and Don C. Reitzes. 1981. “The Link Between Identity and Role
Performance.” Social Psychology Quarterly 44: 83-92.

Callero, Peter L. 2003. “The Sociology of the Self.” Annual Review of Sociology
29: 115-133.

177

Cerulo, Karen. 1997. “Identity Construction: New Issues, New Directions.”
Annual Review of Sociology 23, 385-409.

Cooley, Charles H. 1998. On Self and Social Organization. Chicago: University
of Chicago Press.

Cornfield, Michael. 2004. Politics Moves Online: Campaigning and the Internet.
New York: Century Foundation.

Cramer, Kenneth L. and Sherry L. Grace. 2003. “The Elusive Nature of Self-
Measurement: The Self-Construal Scale versus the Twenty Statements Test.” Journal of
Social Psychology 143 (5): 649-668.

Crandall, Christian, Laurie T. O’Brien, and Amy Eshleman. 2002. “Adapting the
Self to Local Group Norms: Internalizing the Suppression of Prejudice” in The Social

Self: Cognitive, Interpersonal and Intergroup Perspectives. New York: Psychology
Press, 293-308.

Dahl, Robert A. 1972. Democracy in the United States: Promise and Performance
(2” Edition). Chicago: Rand McNally.

D’Alessio, David 2000. “Adoption of the World Wide Web by American
Political Candidates, 1996-1998.” Journal of Broadcasting and Electronic Media 44 (4),
556-568.

Davis, Richard. 2005. Politics Online: Blogs, Chatrooms, and Discussion Groups
in American Democracy. New York, NY: Routledge Press.

Davis, Steve, Larry Elin, and Grant Reeher. 2005. “The lntemet Helps Citizens
Become More Active in Politics” in The Internet: Opposing Viewpoints Edited by James
D. Torr. F armington Hills, MI: Greenhaven Press, pp. 35-46.

Deaux, K., A. Reid, K. Mizrahi, and K.A. Ethier. 1995. “Parameters of Social
Identity.” Journal of Personality and Social Psychology,_68: 280-291 .

Demo, David H. 1992. “The Self-Concept Over Time.” Annual Review of
Sociology 18, 303-323.

Dittrnar, Helga. 1992. The Social Psychology of Material Possessions. New York:
St. Martin’s Press.

Dowd, Matthew, Douglas B. Sosnick, and Ron Fournier. 2006. Applebee ’s
America. New York: Simon & Schuster.

Downie, Leonard Jr. and Robert G. Kaiser. 2003. The News about the News. New
York: Vintage Books.

178

 

Dunn, Robert G. 2008. Identifizing Consumption: Subjects and Objects in 3.—
Consumer Society. Philadelphia, PA: Temple University Press.

Dye, Thomas R. 1990. Who ’s Running America? Englewood Cliffs, NJ: Prentice
Hall.

Entrnan, Robert M. 1993. “Framing: Toward a Clarification of a Fractured
Paradigm.” Journal of Communication 43 (4): 51-58.

F iorina, Morris P., Samuel J. Abrams and Jeremy C. Pope. 2006. Culture War?
The Myth of a Polarized America. New York, NY: Pearson Education.

Foot, Kristen A. and Steven M. Schneider. 2002. “Online Action in Campaign
2000: An Exploratory Analysis of the US. Political Web Sphere.” Journal of
Broadcasting and Electronic Media 46 (2) 222-244.

Galston, William A. 2002. “The impact of the lntemet on civic life: An early
assessment” in Governance. com: Democracy in the Information Age Edited by Elaine C.
Kamarck and Joseph S. Nye Jr. Washington DC: Brookings Institution Press, pp. 40-58.

Galston, William A. 2004. "The Impact of the lntemet on Civic Life: An Early
Assessment" in The lntemet and Public Life Edited by Verna V. Gehring. Lanham, MD:
Rowman and Littlefield, pp. 59—78.

Galston, William A. 2005. “The Internet Fosters More Divisiveness in Politics” in
The Internet: Opposing Viewpoints Edited by James D. Torr. Farmington Hills, MI:
Greenhaven Press, pp. 47-57.

Gamson, William A. and Andre Modigliani. 1989. “Media Discourse and Public
Opinion on Nuclear Power: A Constructionist Approach.” American Journal of
Sociology, 95:1. pp. 1-37.

Gans, Herbert. 2003. Democracy and the News. Oxford: Oxford University Press.

Gecas, Viktor and Peter Burke. 1995. “Self and Identity” in Social Perspectives
on Social Psychology Edited by Karen Cook, Gary Alan Fine and James S. House.
Boston, MA: Allyn & Bacon, 41-60.

Gecas, Viktor. 2000. “Value Identities, Self-Motives, and Social Movements” in
Self Identity, and Social Movements Edited by S. Stryker, T.J. Owens, and RW. White.
Minneapolis, MN: University of Minnesota Press, 93-109.

Gecas, Viktor. 2001. “The Self as a Social Force” in Extending Self-Esteem

Theory and Research: Sociological and Psychological Currents Edited by TJ. Owens, S.
Stryker and N. Goodman. Cambridge: Cambridge University Press, 85-100.

179

Gergen, Kenneth J. 1991. The Saturated Self: Dilemmas of Identity in
Contemporary Life. New York: Basic Books.

Giddens, Anthony 1991. Modernity and Selfildentity: Self and Identity in the Late
Modern Age. Cambridge: Cambridge Press.

Goffman, Erving. 1959. The Presence of Self in Everyday Life. New York:
Doubleday.

Goldberg, Bernard. 2003. Bias: A CBS Insider Exposes how the Media Distort the
News. New York: Harper Publishing.

Goldberg, Lewis R. 1992. The development of markers for the Big-five factor
structure. Journal of Personality and Social Psychology, 59(6), 1216-1229.

Goldberg, Lewis. R1993. The structure of phenotypic personality traits.
American Psychologist, 48, 26-34.

Graber, Doris. 2005. Mass Media and American Politics. New York: CQ Press.

Green, Melanie C. 2007 . "Trust and Social Interaction on the lntemet" in The
Oxford Handbook of Internet Psychology Edited by Adam N. Joinson, Katelyn Y.A.
McKenna, Tom Postmes, and Ulf-Dietrich Reips. Oxford: Oxford University Press, pp.
43-52.

Habermas, Jurgen. 1989. The Structural Transformation of the Public Sphere: An
Inquiry into a Category of Bourgeois Society. Cambridge: Polity Press.

Hamilton, James. 2004. All the News that ’sfit to sell: How the Market Transforms
Information into News. Princeton, NJ: Princeton University Press.

Hampton, Keith N. and Barry Welhnan. 2002. “The Not So Global Village of
Netville” in The Internet in Everyday Life Edited in Barry Wellman and Caroline
Haythomthwaite. Malden, MA: Blackwell Publishers Ltd. pp. 345-371.

Hastie, Reid, David Schkade and Cass R. Sunstein. 2005. “What Happened on
Deliberation Day?” California Law Review 95, 915.

Haythomthwaite, Caroline and Barry Wellman. 2002. “The Internet in Everyday
Life: An Introduction” in The Internet in Everyday Life Edited by Barry Wellman and
Caroline Haythomthwaite. Malden, MA: Blackwell Publishers Ltd, pp. 3-44.

Haythomthwaite, Carolyn. 2007. "Social Networks and Online Community" in
The Oxford Handbook of Internet Psychology Edited by Adam N. Joinson, Katelyn Y.A.
McKenna, Tom Postmes, and Ulf-Dietrich Reips. Oxford: Oxford University Press, pp.
12 1-13 8.

180

 

 

Herek, GM. 1986. “The instrumentality of attitudes: Toward a neofimctional
theory.” Journal of Social Issues 42 (2): 99-114.

Hewitt, John P. 1989. Dilemmas of the American Self Philadelphia, PA: Temple
University Press.

Higgins, ET. 1987. “Self-Discrepancy: A Theory Relating Self and Affect.”
Psychological Review 94, 319-340.

Holmvall, Camilla M. and DR. Bobocel. 2008. “What fair procedures say about
me: Self-construals and reactions to procedural fairness.” Organizational Behavior and
Human Decision Processes 105(2): 147-168.

Holstein, James A. and Jaber F. Gubrium. 2000. The Self We Live By: Narrative
Identity in a Postmodern World. Oxford: Oxford University Press.

House, James. 1992. “Social Structure and Personality” in Social Psychology:
Sociological Perspectives Edited by Morris Rosenberg and Ralph H. Turner. New
Brunswick, NJ: Transaction Publishers.

Howard, Judith A. 2000. “Social Psychology of Identities.” Annual Review of
Sociology 26, 367-393.

Hull, Christopher C. 2006. “Online Organization: Dean, Kerry, and lntemet
Politicking in the 2004 Iowa Caucus” in The Internet Election: Perspectives on the Web
in Campaign 2004 Edited by Andrew Paul Williams and John C. Tedesco. Lanham, MD:
Rowman & Littlefield Publishers, pp. 57-66.

Iyengar, Sheena S. and Mark R. Lepper. 2002. "Choice and Its Consequences: On
the Costs and Benefits of Self-Determination" in Self and Motivation: Emerging
Psychological Perspectives Edited by Abraham Tesser, Diedrick A. Stapel, and Joanne V.
Wood. Washington DC: American Psychological Association, pp. 71-96.

Iyengar, Shanto and Richard Morin. 2006, May 3. “Red Media, Blue Media:
Evidence for a Political Litmus Test in Online News Readership.” Washington Post.

Jensen, Jakob Linaa. 2006. “The Minnesota E-democracy project: mobilising the
mobilised?” in The Internet and Politics: Citizens, Voters, and Activists Edited by Sarah
Oates, Diana Owen and Rachel K. Gibson. New York, NY: Routledge Press, pp. 39-58.

Jones, Steve. 1999. “Studying the Net: Intricacies and Issues” in Doing Internet
Research: Critical Issues and Methods for Examining the Net Edited by Steve Jones
Thousand Oaks, CA: Sage Publications, pp. 1-27.

181

Kamarck, Elaine C. 1999. “Campaigning on the lntemet in the Elections of 1998”
in Governenace. Com: Democracy in the Information Age Edited by Elaine C. Kamarck
and Joseph S. Nye. Washington, DC: Brookings Institution, pp. 99-123.

Kashima, Y., S. Yamaguchi, U. Kim, S-C Choi, M.J. Gelfand, and M. Yuki.
1995. “Culture, Gender, and Self: A Perspective fi'om Individualism-Collectivism
Research.” Journal of Personality and Social Psychology 69, 925-937.

Katz, Daniel. 1960. “The Functional Approach to the Study of Attitudes.” Public
Opinion Quarterly 24, 163-204.

Katz, James E. and Ronald E. Rice. 2002A. “Syntopia: Access, Civic
Involvement, and Social Interaction on the Net” in The Internet in Everyday Life Edited
in Barry Wellman and Caroline Haythomtlrwaite Malden, MA: Blackwell Publishers
Ltd., pp. 114-138.

Katz, James E. and Ronald E. Rice. 2002B. The Social Consequences of lntemet
Use. Cambridge, MA: NflT Press.

Kavanaugh, Andrea L. and Scott J. Patterson. 2002. “The Impact of Community
Computer Networks on Social Capital and Community Involvement in Blacksburg” in
The lntemet in Everyday Life Edited in Barry Wellman and Caroline Haythomthwaite.
Malden, MA: Blackwell Publishers Ltd., pp. 325-344.

Kaye, Barbara K. and Thomas J. Johnson. 2006. “The Age of Reasons: Motives
for Using Different Components of the lntemet for Political Information” in The Internet
Election: Perspectives on the Web in Campaign 2004, edited by Andrew P. Williams and
John C. Tedesco, 147-168. Oxford: Rowman and Littlefield.

Keen, Andrew. 2007. The Cult of the Amateur: How Today ’s lntemet is Killing
our Culture. New York: Doubleday.

Kitayama, 8., HR. Markus, H. Matsumoto, and V. Norasakkunkit. 1997.
“Individual and Collective Processes in the Construction of the Self: Self-Enhancement
in the United States and Self-Criticism in Japan.” Journal of Personality and Social
Psychology 72, 1245-1267.

Klotz, Robert J (2004). The Politics of Internet Communication Lanham, MD:
Rowman & Littlefield Publishers.

Kolko, Beth E. 2003. Virtual Politics: Policy and Community in an Electronic
Age. New York: Columbia University Press.

Komito, Lee. 1998. "The Net as a Foraging Society: Flexible Communities"
Information Society 14(2), 97-106.

182

1 h "1... t ‘

 

Kuhn, Manford H. and Thomas S. McPartland. 1954. “An Empirical Investigation
of Self-Attitudes.” American Sociological Review 19: 69-85.

Lakofi‘, George. 2004. Don ’t Think of an Elephant: Know Your Values and Frame
the Debate. White River Junction, VT.: Chelsea Green Publishers.

Lakoff, George. 2006. Thinking Points. New York, NY: Farrar, Strauss, and
Giroux.

Lasica, Joseph D. 2003. “Blogs and Journalism Need Each Other.” Nieman
Reports 57 (3): 70-75.

Leary, Mark R. 2002. "When Selves Collide: The Nature of the Self and the
Dynamics of Interpersonal Relationships" in Self and Motivation: Emerging
Psychological Perspectives Edited by Abraham Tesser, Diedrick A. Stapel, and Joanne
V. Wood. Washington DC: American Psychological Association, pp. 119-146.

Levine, Peter. 2004. "The lntemet and Civil Society" in The Internet and Public
Life Edited by Verna V. Gehring. Lanham, MD: Rowman and Littlefield, pp. 79-100.

Levine, Timothy, M.J. Bresnahan, H.S. Park, M.K. Lapinski, G.M. Wittenbaum,
S.M. Shearman, S.Y. Lee, D. Chrmg, and R. Ohashi. 2003. “Self-Construal Scales Lack
Validity.” Human Communication Research 29 (2): 210-252.

Lord, Charles, Lee Ross, and Mark Lepper. 1979. “Biased Assimilation and
Attitude Polarization: The Effects of Prior Theories on Subsequently Considered
Evidence.” Journal of Personality and Social Psychology 37 (1 1), 2098-2109.

Lydon, John E., D. Menzies-Toman, 1!. Burton, and C. Bell. 2008. “If-then
contingencies and the differential effects of the availability of an attractive alternative on

relationship maintenance for men and women.” Journal of Personality and Social
Psychology 95 (1): 50-65.

Manjoo, Farhad. 2008. True Enough: Learning to Live in a Post-F act Society.
Hoboken, NJ: John Wiley and Sons, Inc.

Margolis, Michael and David Resnick. 2000. Politics as Usual: The Cyberspace
“Revolution ”. Thousand Oaks, CA: Sage Publications.

Markus, Hazel R. and S. Kitayama. 1991. “Culture and the Self: Implications for
Cognition, Emotion, and Motivation.” Psychological Review 98, 224-253.

Mazzoleni, Giorgio. 1987. “Media Logic and Party Logic in Campaign Coverage:
The Italian General Election of 1983.” European Journal of Communication 2 (1): 55-80.

183

McCall, George J. and J.L. Simmons. 1978. Identities and Interactions. New
York: Free Press.

McChesney, Robert W. 2004. The Problem of the Media: U. S. Communication
Politics in the 21 “ Century. New York: Monthly Review Press.

McKenna, Katelyn Y.A. 2007 . "Through the lntemet Looking-Glass: Expressing
and Validating the True Self" in The Oxford Handbook of lntemet Psychology Edited by
Adam N. Joinson, Katelyn Y.A. McKenna, Tom Postmes, and Ulf-Dietrich Reips.
Oxford: Oxford University Press, pp. 205-222.

McLeod, Jack M., and Lee B. Becker. 1981. “The Uses and Gratifications
Approach” in Handbook of Political Communications edited by Dan D. Nimmo and
Keith R. Sanders. Sage Publications: Beverly Hills, CA. pp. 67-99.

Mead, George H. 1934. Mind, Self and Society From the Stantboint of a Social
Behaviorist. Chicago: University of Chicago Press.

Nardi, Bonnie, Diane J. Schiano, Michelle Gumbrecht, and Luke Swartz. 2004.
“Why We Blog.” Communications of the ACM 47 (12): 41-46.

Nezlek, John B., D. Menzies-Toman, K. Bmton, and C. Bell. 2008. “Emotions in
Everyday Social Encounters: Correspondence between Culture and Self-Construal.”
Journal of Cross-Cultural Psychology 39 (4): 366-3 72.

Nie, Norman H. and Erbring, Lutz. 2005. “Internet Use Decreases Social
Interaction” in The Internet: Opposing Viewpoint; Edited by James D. Torr. F armington
Hills, MI: Greenhaven Press, pp. 28-34.

Norris, Pippa. 2002. "Revolution, What Revolution? The lntemet and US
Elections, 1992-2000" in Governance. com: Democracy in the Information Age Edited by
Elaine C. Kamarck and Joseph S. Nye Jr. Washington DC: Brookings Institution Press,
pp. 59-80.

Norris, Pippa. 2004. "The Bridging and Bonding Role of Online Communities" in
Society Online: The Internet in Context Edited by Philip N. Howard and Steve Jones.
Thousand Oaks, CA: Sage Publications, pp. 31-42.

Oates, Sarah and Rachel K. Gibson. 2006. “The lntemet, civil society and
democracy: a comparative perspective” in The Internet and Politics: Citizens, Voters, and
Activists Edited by Sarah Oates, Diana Owen and Rachel K. Gibson. New York, NY:
Routledge Press, pp. 1-18.

Oppenheimer, Bruce. 2005. “Deep Red and Blue Congressional Districts” in

Congress Reconsidered Edited by Lawrence C. Dodd and Bruce Oppenheimer, 8"I
edition. Washington DC: Congressional Quarterly Press.

184

Owen, Diana. 2006. “The Internet and youth civic engagement in the United
States” in Sarah Oates, Diana Owen and Rachel K. Gibson The Internet and Politics:
Citizens, Voters, and Activists. New York, NY: Routledge Press, pp. 20-38.

Oyserman, Daphna and Martin Packer. 1996. “Social Cognition and Self-
Concept: A Socially Contextualized Model of Identity” in What ’s Social About Social
Cognition? Edited by J. Nye and A. Brewer. Beverly Hills, CA: Sage Publications, 175-
201.

Park, Robert 1940. “News as a Form of Knowledge” in On Social Control and
Collective Behavior, edited by RH. Turner, 32-52. Chicago: Chicago University Press.

Pavlik, John M. and Stephen S. Ross. 2000. “Journalism Online: Exploring the
Impact of New Media on News and Society” in Understanding the Web: Social, Political,
and Economic Dimensions of the Internet. Ames, IA: Iowa State University Press.

Pavlik, John V. 2008. Media in the Digital Age. New York: Columbia University
Press.

Perlmutter, David D. 2008. Blogwars. Oxford: Oxford University Press.

Polsby, Nelson W. 2004. How Congress Evolves: Social Bases of Institutional
Change. New York: Oxford University Press.

Prentice, Deborah A. 2001. “The Individual Self, Relational Self, and Collective
Self: A Commentary” in Individe Self Relational Self Collective Self Edited by C.
Sedikedes and MB. Brewer. New York: Psychology Press, 315-326.

Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American
Community. New York, NY: Simon and Schuster.

Quan-Haase, Anabel, Barry Wellman, James C. Witte, and Keith N. Hampton.
2002. “Capitalizing on the Net: Social Contact, Civic Engagement, and Sense of
Community” in The lntemet in Everyday Life Edited by Barry Wellman and Caroline
Haythomthwaite Malden, MA: Blackwell Publishers Ltd, pp. 291-324.

Rainie, Lee, Michael Cornfield, and John Horrigan. 2005. The lntemet and
Campaign 2004. Washington DC: Pew Interest & American Life Project and Pew
Research Center for the People and the Press. March 6. Available online at
www.pewintemet.org/

Rheingold, Howard. 1995. The Virtual Community. London: Minerva.

Rice, Ronald E., Adrian Shepherd, William H. Dutton, and James E. Katz. 2007 .
"Social Interaction and the lntemet: A Comparative Analysis of Surveys in the US and

185

Britain" in The Oxford Handbook of Internet Psychology Edited by Adam N. Joinson,
Katelyn Y.A. McKenna, Tom Postmes, and Ulf-Dietrich Reips. Oxford: Oxford
University Press, pp. 7-30.

Roberts, B.W. and R. Helson. 1997. “Changes in Culture, Changes in Personality:
the Influence of Individualism in a Longitudinal Study of Women.” Journal of
Personality and Social Psychology 72, 641-651.

Robinson, John P., Alan Neustadtl, and Meyer Kestnbaum. 2004. "Technology
and Tolerance: Public Opinion Differences Among lntemet Users and Nonusers" in
Society Online: The Internet in Context Edited by Philip N. Howard and Steve Jones.
Thousand Oaks, CA: Sage Publications, pp. 237-254.

Rodman, Gilbert B. 2003. "The Net Effect: The Public's Fear and the Public
Sphere" in Virtual Politics: Policy and Community in an Electronic Age Edited by Beth
E. Kolko. New York: Columbia University Press, pp. 9-48.

Rosenberg, Morris. 1979. Conceiving the Self New York, NY: Basic Books.

Rubin, Alan M. 2002. “The Uses-and-Gratifications P 've of Media
Efl‘ects” in Media Eflects: Advances in Theory and Research ( Edition) Edited by
Jennings Bryant and Dolf Zillman. Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Saco, Diana. 2002. Cybering Democracy: Public Space and the lntemet.
Minneapolis, W: University of Minnesota Press.

Sassenberg, Kai and Kai J. Jonas. 2007. "Attitude Change and Social Influence on
the Net" in The Oxford Handbook of lntemet Psychology Edited by Adam N. Joinson,
Katelyn Y.A. McKenna, Torn Postmes, and Ulf-Dietrich Reips. Oxford: Oxford
University Press, pp. 273-288.

Schneider, Stephen M. and Kirsten A. Foot. 2006. “Web Campaigning by
Presidential Primary Candidates” in The lntemet Election: Perspectives on the Web in
Campaign 2004, edited by Andrew P. Williams and John C. Tedesco, 21-36. Oxford:
Rowman and Littlefield.

Schudson, Michael. 2003. The Sociology of News. New York: WW Norton and
Company.

Schwartz, Barry. 2005. The Paradox of Choice Why Less rs More. New York:
Harper Collins Publishers.

Sedikedes, Constantine and Marilynn B. Brewer. 2001. Individual Self Collective

Self Relational Self Edited by C. Sedkiedes and M. Brewer. New York: Psychology
Press, 1-13.

186

Semetko, Holli A. and Patti M. Valkenburg. 2000. “Framing European Politics: A
Content Analysis of Press and Television News.” Journal of Communications 50 (2): 93-
109.

Serpe, Richard T. 1987. “Stability and Change in Self: A Structural Symbolic
Interactionist Explanation.” Social Psychology Quarterly 50, 44-55.

Shavitt, S. and MR. Nelson. 2000. “The social-identity frmction in person
perception: Communicated meanings of product preferences.” in Why we evaluate:
Functions of attitudes Edited by GR. Maio and J .M. Olson. Mahwah, NJ: Lawrence
Erlbaum Associates, pp. 37-57.

Sherif, Muzafer. 1935. “A Study of Some Social Factors in Perception.” Archives
of Psychology 187: 47.

Simon, Bemd and Claudia Kampmeier. 2001. “Revisiting the Individual Self:
Towards a Social Psychological Theory of the Individual Self and the Collective Self” in
Individual Self Relational Self Collective Self Edited by C. Sedikedes and MB. Brewer.
New York: Psychology Press, 199-218

Singelis, T. M. 1994. “The measurement of independent and interdependent self-
construals” Personality & Social Psychology Bulletin, 20 (5), 580-591.

Singer, Jane B. 2006. “Journalists and News Bloggers: Complements,
Contradictions, and Challenges” in Uses of Blogs, edited by Axel Bruns and Joanne
Jacobs, 23-32. New York: Peter Lang Publishing.

Skitka, Linda J. and Elizabeth Mullen. 2002. "Psychological Determinants of
Public Opinion" in The Social Psychology of Politics Edited by Victor C. Ottati, R. Scott
Tindale, John Edwards, Fred B. Bryant, Linda Heath, Daniel C. O'Connell, Yolanda
Suarez-Balcazar, and Emil J. Posavac. New York: Kluwer Academic/Plenum Publishers,
pp. 107-134.

Skocpol, Theda. 2003. Diminished Democracy: From Membership to
Management in American Civic Life. Norman, OK: University of Oklahoma Press.

Slater, Don. 1997. Consumer Culture and Modernity. Malden, MA: Blackwell
Publishing Inc.

Snow, David A. and Doug McAdam. 2000. “Identity Work Processes in the
Context of Social Movements: Clarifying the Identity/Movement Nexus” in Self Identity,
and Social Movements Edited by S. Stryker, T.J. Owens, and R.W. White. Minneapolis,
MN: University of Minnesota Press, 41-67.

Spears, R. and M. Lea. 1992. “Social Influence and the influence of the “social”
187

in Computer-Mediated Communication” In Contexts of Computer-Mediated
Communication Edited by M. Lea. London, UK: Harvester-Wheatsheaf, pp. 30-65.

Spears, Russell, Martin Lea, and Tom Postmes. 2007. "CMC and Social Identity"
in The Oxford Handbook of Internet Psychology Edited by Adam N. J oinson, Katelyn
Y.A. McKenna, Tom Postmes, and Ulf-Dietrich Reips. Oxford: Oxford University Press,
pp. 253-272.

Stets, Jan E. and Burke, Peter J. 2000. “Identity Theory and Social Identity
Theory.” Social Psychology Quarterly 63, 224-237.

Stokes, Ashli Quesinberry. 2006. “Discrediting Teresa: Wounded by Whispers on
the Web” ” in The Internet Election: Perspectives on the Web in Campaign 2004 Edited
by Andrew Paul Williams and John C. Tedesco. Lanham, MD: Rowman & Littlefield
Publishers, pp. 169-186.

Stromer-Galley, Jennifer and Andrea B. Baker. 2006. “Joy and Sorrow of
Interactivity on the Campaign Trail: Blogs in the Primary Campaign of Howard Dean” in
The Internet Election: Perspectives on the Web in Campaign 2004 Edited by Andrew
Paul Williams and John C. Tedesco. Lanham, MD: Rowman & Littlefield Publishers, pp.
1 1 1-132.

Stryker, Sheldon. 1980. Symbolic Interaction: A Social Structural Version. Menlo
Park, CA: Benjamin Cummings Publishing.

Stryker, Sheldon and Richard Serpe. 1994. “Identity Salience and Psychological
Centrality: Equivalent, Overlapping, or Complementary Concepts?” Social Psychology
Quarterly 57, 16—35.

Stryker, Sheldon and Peter J. Burke. 2000. “The Past, Present, and Future of an
Identity Theory.” Social Psychology Quarterly 63, 284-297.

Stryker, Sheldon, Timothy J. Owens and Robert W. White. 2000. “Social
Psychology and Social Movements: Cloudy Past and Bright Future” in Self Identity, and
Social Movements Edited by S. Stryker, T.J. Owens, and R.W. White. Minneapolis, MN:
University of Minnesota Press, 1-20.

Stryker, Sheldon. 2000. “Identity Competition: Key to Differential Social
Movement Participation?” in Self Identity, and Social Movements Edited by S. Stryker,
T.J. Owens, and R.W. White. Minneapolis, MN: University of Minnesota Press, 21-40.

Sunstein, Cass S. 2007. Republic. com 2.0: Revenge of the Blogs. Princeton, NJ:
Princeton University Press.

188

Sunstein, Cass R. 2009. Going to Extremes: How Like Minds Unite and Divide.
Oxford: Oxford University Press.

Taj fel, Henri. 1981 . Human Groups and Social Categories. London: Academic
Press.

Thompson, John B. 1995. The Media and Modernity: A Social Theory of the
Media. Stanford, CA: Stanford University Press.

Trammell, Kaye D. and Joshua Britton. 2005. “Gatewatching: The Impact of Blog
Content on the Mainstream Media.” Paper presented at the annual meeting of the
Association of Internet Researchers. Chicago, October.

Trammell, Kaye D. 2006. “The Blogging of the President” ” in The lntemet
Election: Perspectives on the Web in Campaign 2004 Edited by Andrew Paul Williams
and John C. Tedesco. Lanham, MD: Rowman & Littlefield Publishers, pp. 133-146.

Traudt, Paul J. 2005. Media, Audiences, Ejfects: An Introduction to the Stuay of
Media Content and Audience Analysis. Boston, MA: Pearson Education, Inc.

Turner, John C. 1999. “Some Current Issues in Research on Social Identity and
Self-Categorization Theories” in Social Identity: Context, Commitment, Content Edited
by N. Eleemers, R. Spears, and B. Doosje. New York: Blackwell Publishing, 7-34.

Uslaner, Eric M. 2004A. “Social Capital and the Net” in The Internet in Public
Life_Edited by Verna V. Gehring. Lanham, MD: Rowman & Littlefield Publishers, pp.
101-112.

Uslaner, Eric M. 2004B. "Trust, Civic Engagement, and the Internet." Political
Communication 21, 223-242.

Vacker, Barry. 2000. "Global Village or World Bazaar?" in Understanding the
Web: Social, Political, and Economic Dimensions of the Internet Edited by Alan B.
Albarran and David H. Goff. Ames, IA: Iowa State University Press, pp. 211-238.

Vallone, Robert R, Lee Ross and Mark R. Lepper. 1985. “The Hostile Media
Phenomenon: Biased Perception and Perceptions of Media Bias in Coverage of the Beirut
Massacre.” Journal of Personality and Social Psychology 49 (3): 577-586.

Van Alstyne, Marshall and Erik Brynjolfsson. 1997. “Electronic Communities:
Global Village or Cyberbalkans?” Sloan School, MIT, March, 1997.

Wachbroit, Robert. 2004. “Reliance and Reliability: the Problem of Information

on the lntemet” in The Internet in public life, edited by Verna V. Gehring, 29-41. Oxford:
Rowman and Littlefield.

189

Walker, Rob. 2008. Buying In: The Secret Dialogue Between What We Buy and
Who We Are. New York: Random House.

Weinberger, Mary. 2005. “The lntemet Increases Social Interaction” in The
Internet: Opposing Viewpoints Edited by James D. Torr. F armington Hills, MI:
Greenhaven Press, pp. 18-27.

Wellman, Barry. 1999. Networks in the Global Village. Boulder, CO: Westview
Press.

Whillock, Rita Kirk. 2000. "Age of Reason: The Electronic Frontier Confronts the
Aims of Political Persuasion" in Understanding the Web: Social, Political, and Economic
Dimensions of the Internet Edited by Alan B. Albarran and David H. Goff. Ames, IA:
Iowa State University Press, pp. 165-192.

White, Harrison C. 1992. Identity and Control: A Structural Theory of Social
Action. Princeton, NJ: Princeton University Press.

Wise, J. MacGregor.2003. "Community, Affect, and the Virtual: The Politics of
Cyberspace" in Virtual Politics: Policy and Community in an Electronic Age Edited by
Beth E. Kolko. New York: Columbia University Press, pp. 112-133.

Williams, Andrew P. and Lynda L. Kaid. 2006. “Media Framing of the European
Parliamentary Elections: A View from the United States.” In Campaigning for Europe:

Political Parties, Campaigns, Mass Media and the European Parliament Elections 2004,
edited by Michaela Maier and Jens Tenscher, 295-304. London: LIT.

Winograd, Morley and Michael D. Hais. 2008. Millennial Makeover: MySpace,
YouTube & the Future of A merican Politics. New Brunswick, NJ: Rutgers University
Press.

Wise, J. MacGregor.2003. "Community, Affect, and the Virtual: The Politics of
Cyberspace" in Virtual Politics: Policy and Community in an Electronic Age Edited by
Beth E. Kolko. New York: Columbia University Press, pp. 112-133.

Wood, Andrew F. and Matthew J . Smith. 2005. Online Communication: Linking
Technology, Identity, and Culture (2" Edition). London: Lawrence Erlbaum Associates.

Wright, Stephen C., Art Aron and Linda R. Tropp. 2002. “Including Others (and
Groups) in the Self: Self-Expression and Intergroup Relations” ” in Individual Self
Relational Self Collective Seledited by C. Sedikedes and MB. Brewer. New York:
Psychology Press, 343-3 64.

Yeung, Dannii Y., H.H. Fung, and FR. Lang. 2008. “Self-Construal moderates
differences in social network characteristics.” Psychology and Aging 23 (1): 222-226.

Zurcher, Louis A. 1977. The Mutable Self Beverly Hills, CA: Sage Publications.

190