Internet connectedness and its social origins: An ecological approach to communication media and social inequality

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content INTERNET CONNECTEDNESS AND ITS SOCIAL ORIGINS:
AN ECOLOGICAL APPROACH TO COMMUNICATION MEDIA AND
SOCIAL INEQUALITY
Copyright 2003
by
Joo-Young J. Jung
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
May 2003
Joo-Young J. Jung
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UMI Number: 3103911
Copyright 2003 by
Jung, Joo-Young Janice
All rights reserved.
®
UMI
UMI Microform 3103911
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, written by
under the direction o f h P C dissertation committee, and
approved by all its members, has been presented to and
accepted by the Director o f Graduate and Professional
Programs, in partial fulfillment o f the requirements fo r the
degree of
DOCTOR OF PHILOSOPHY
■4cey(jiA^ A , _
Director
Date L 'L ilA iP g-'y
Dissertation Committee.
f N
" f c c t P (2- k e s c U
, Chair
jg O
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ii
ACKNOWLEDGEMENTS
First of all, I thank God for His blessing in every aspect of my life.
I would like to take this opportunity to acknowledge the guidance and
support of many people during my doctoral study, research and writing of this
dissertation. None of this would have happened without consistent guidance, support
and advice of Professor Sandra Ball-Rokeach. Her supreme thinking, research and
teaching have been a continuous inspiration for my academic development in the
doctoral program. I also wish to express my gratitude to Professors Margaret
McLaughlin, Patricia Riley, William Dutton, Peter Monge, William Loges and
Timothy Biblarz for their guidance during my coursework, qualifying examination
and dissertation. An essential part of my graduate study was working with fellow
students. I wish to express particular appreciation to fellow doctoral students at the
Metamorphosis Project. It has been an invaluable asset with regard to thinking,
theorizing, research, and many more.
I am also greatly indebted to professors at Yonsei University in Seoul, Korea
where I learned invaluable lessons during my undergraduate years. I express
particular gratitude to Professors Heung-Soo Park, In-Hwan Oh, Jung-Woo Suh,
Young-Seok Kim, Yang-Soo Choi and Young-Chul Yoon.
Love and support of my family have held me up throughout my graduate
years living across the Pacific Ocean from my home. I thank my father, Ku-Hyun
Jung, and my mother, Eun-Kyung Park for giving me life in the first place, for
educating me in every aspect of my life and giving me intellectual and emotional
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
supports. They have always been my greatest role models and the best mentor. I
cannot express my love and respect to them enough. My brother Se-Young has been
a good and proud friend of mine. My in-law family, especially my father-in-law and
mother-in-law, Wann Yu and Yang-Ja Choi, has provided support and
encouragement during my dissertation-writing period. I thank them for their love and
support. Last, but not the least, my husband, Tong Yu, has always been there to rely
on, talk, and share everything. Although we have been apart in New York and Los
Angeles due to my study, the distance didn’t exist in our mind and heart. I love you
dearly.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS.................................................................. II
LIST OF TABLES.............................................................................VIII
LIST OF FIGURES..............................................................................X
ABSTRACT.........................................................................................XI
CHAPTER 1. COMMUNICATION TECHNOLOGY AND
INEQUALITY: INTRODUCTION........................................................ 1
Background....................................................................................................................... 1
General Assumptions...................................................................................................... 3
Overview of the Dissertation..........................................................................................4
Media, Information and Society: A Historical Review..............................................7
Internet’s Significance.................................................................................................. 10
Double Barriers to Communication Technology and Equality: Access and
Application...................................................................................................................... 12
First Barrier: Connecting to the Internet..................................................................15
Second Barrier: The Barrier in Utilizing Internet Resources..................................19
CHAPTER 2. MEDIA SYSTEM DEPENDENCY TO
COMMUNICATION CONNECTEDNESS: A REVIEW OF THE
EVOLUTION OF MSD THEORY AND A PROPOSAL FOR A
NEW CONCEPT................................................................................23
Media System Dependency Theory............................................................................24
Evolution o f Media System Dependency Theory: Overview....................................24
The Review o f Three Major Theoretical Studies.......................................................28
Beginning.................................................................................................................28
Development.............................................................................................................34
Clarification............................................................................................................ 40
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
V
MSD Theory Development in Empirical Studies......................................................44
Micro-Level Dependency Relations...................................................................... 44
Meso/Macro-Level Relations................................................................................. 47
Ambiguous Personal Environment.........................................................................48
Social/Political Environment................................................................................. 51
Limitations o f the Media System Dependency Theory............................................. 53
Communication Infrastructure Theory.....................................................................55
A Theoretical Framework...........................................................................................55
Communication Infrastructure's Departure from Media System Dependency 59
Communication Connectedness................................................................................. 61
Dimensions o f Connectedness................................................................................ 63
Summary......................................................................................................................... 64
CHAPTER 3. NEW COMMUNICATION TECHNOLOGIES AND
INEQUALITY: LITERATURE REVIEW...........................................67
Introduction.................................................................................................................... 67
Diffusion of Innovation.................................................................................................67
Knowledge Gap..............................................................................................................71
New Technology and Inequality Studies....................................................................73
Internet and the “Digital Divide” ............................................................................... 76
Limitations of Past Studies...........................................................................................78
Approaching the Digital Divide.................................................................................. 80
Antecedents of Digital Divide...................................................................................... 84
Socioeconomic Status..................................................................................................84
Age.................................................................................................................................86
Ethnicity....................................................................................................................... 88
Social Environment..................................................................................................... 92
Communication Environment: New and Old Media.................................................95
Summary......................................................................................................................... 97
CHAPTER 4. HYPOTHESIS AND METHODOLOGY.................. 99
Hypothesis....................................................................................................................... 99
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
v i
Scope o f Media and Specific Media Connections.....................................................99
Internet Access...........................................................................................................100
Internet Connectedness............................................................................................. 102
Methodology.................................................................................................................106
Data.............................................................................................................................106
Variables.................................................................................................................... 109
Internet Connectedness Index.............................................................................. 109
Internet Goal Scope and Intensity........................................................................114
Social Environment............................................................................................... 115
Technological Environment..................................................................................116
Media Scope...........................................................................................................117
Specific Media Connectedness............................................................................. 118
Socioeconomic Status............................................................................................118
Demographic Characteristics.............................................................................. 119
CHAPTER 5. RESULTS................................................................. 121
Frequencies...................................................................................................................121
Correlations.................................................................................................................. 125
Hypotheses and Research Questions Testing......................................................... 129
RQ1: Media Scope.................................................................................................... 129
RQ2: Specific Media Intensity................................................................................. 130
Television Connections.........................................................................................131
Newspaper Connections........................................................................................131
Internet Connections.............................................................................................131
Interpersonal Connections....................................................................................132
Internet Access...........................................................................................................132
HI: Socioeconomic and Demographic Variables...............................................133
H2: Media Scope................................................................................................... 133
RQ3: Specific Media Connections.......................................................................134
Final Model............................................................................................................134
Internet Connectedness.............................................................................................136
113: Socioeconomic Status and Demographic Characteristics..........................136
H4: Social and Technological Environments..................................................... 137
RQ4: Media Connections......................................................................................137
H5: Goal Scope and Intensity.............................................................................. 138
Final Model............................................................................................................138
CHAPTER 6. DISCUSSION AND IMPLICATIONS.................... 142
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
vii
Theoretical Contributions..........................................................................................142
Empirical Contributions.............................................................................................146
Implications of the Findings.......................................................................................149
Descriptive Findings................................................................................................. 149
Media Connections.................................................................................................... 151
Media Scope...........................................................................................................151
Specific Media Connections................................................................................. 153
Internet Access...........................................................................................................154
Internet Connectedness.............................................................................................155
Methodological Implications......................................................................................158
Limitations.................................................................................................................... 161
Policy Implications...................................................................................................... 162
Larger Issue: Communication Technology and Social Inequality.........................162
Policy Implications.................................................................................................... 164
BIBLIOGRAPHY---------------------------------------------------------------- 169
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Vlll
LIST OF TABLES
Table 1. Antecedents of Internet Access: Hypotheses and Research Questions— 101
Table 2. Antecedents of Internet Connectedness: Hypotheses and A Research
Question-------------------------------------------------------------------------------------------106
Table 3. Factor Loadings of Items in Internet Connectedness Index and its Sub-
Dimensions-------------------------------------------------------------------------------------- 114
Table 4. Median Number of Media Chosen for Each Goal----------------------------- 121
Table 5. Percent of Respondents who Chose Specific Media for Each Goal 121
Table 6. Internet Access, Home Computer Ownership, Time Spent Online, Places for
Going Online and Average Number of Places for Going Online-----------------------122
Table 7. Internet-Related Help Source------------------------------------------------------ 123
Table 8. Proportions of Family and Friends Using the Internet----------------------- 123
Table 9. Mean Degree of Importance of Goals for Going Online----------------------124
Table 10. Internet Activity Participation---------------------------------------------------- 125
Table 11. Correlations between Internet Related Variables----------------------------- 126
Table 12. Inter-Correlations between Variables in the Internet Connectedness Index—
128
Table 13. Correlations between Media-Related Variables------------------------------ 128
Table 14. Summary of Simultaneous Multiple Regression Analysis for Variables
Predicting Media scope (Standardized Coefficients)------------------------------------ 129
Table 15. Summary of Simultaneous Multiple Regression Analysis for Variables
Predicting Television Connection, Newspaper Connection, Internet Connection, and
Interpersonal Connection (Standardized Coefficients)---------------------------------- 130
Table 16. Summary of Three Logistic Regression Analyses for Variables Predicting
Internet Access— ---------- 135
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 17. Summary of Hierarchical Multiple Regression Analyses for Variables
Predicting Internet Connectedness (Standardized Coefficients)------------------------ 139
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
LIST OF FIGURES
x
Figure 1. Double Barriers to Communication Technology and Equality------------- 14
Figure 2. 1970s: Effects-Focused Conceptual Model of Media System Dependency—
------------- 24
Figure 3. 1980s: Elaboration on the Origins of Individuals’ MSD Relations 25
Figure 4. 1990s: Elaboration ofMeso-Level Dependency Relations and Cross-Level
Interactions---------------------------------------------------------------------------------------- 26
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
xi
ABSTRACT
This study examines issues relating to social inequality and communication
technology. Both theoretical and empirical missions are pursued.
Theoretically, the study reviews the evolution of media system dependency
theory (Ball-Rokeach, 1985, 1998; Ball-Rokeach & DeFleur, 1976) into
communication infrastructure theory (Ball-Rokeach, Kim, & Matei, 2001) over the
past three decades. How communication infrastructure theory originates and departs
from media system dependency theory is examined. Based on the two theories, the
author develops the concept of “communication connectedness” to illustrate the
multidimensional relationships that individuals form with communication media.
Four dimensions of communication connectedness include: access to technology;
scope of activity participations within a medium or across different media; intensity
of relationships; and centrality of communication media in daily lives.
Empirically, the four dimensions of connectedness are applied to the Internet
and operationalized to form the Internet Connectedness Index. With a dataset of 585
randomly selected telephone survey respondents, individual, technological, social
and communication factors that affect their Internet connectedness are considered
and tested under an ecological model.
The results indicate that the following factors have significant influence on
the Internet connectedness: socioeconomic status (income and education);
demographic characteristics (age and ethnicity); technological environment (e.g.,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
having a computer at home; the number of places for having Internet access); social
environments concerning the nature of social relationships people have in their
everyday lives (e.g., interactions with others in receiving and providing Internet-
related help); and the scope and intensity of Internet-related goals.
The findings reveal that even after people gain access to the Internet, the
ways they incorporate the Internet into their everyday lives differ, and that the
differences reflect social disparities that characterize individuals in the context of
their everyday lives. Mere access to the Internet cannot bridge social inequalities.
There must be further efforts to improve people’s abilities to build up their access to
the Internet to derive meaningful values that can help them enhance their well-being.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1
CHAPTER 1. COMMUNICATION TECHNOLOGY AND
INEQUALITY: INTRODUCTION
Background
New technology is intriguing, particularly communication technologies that
directly intrude into everyday life experiences of individuals and institutions. The
invention of the telegraph in the mid-19th century enabled people to distinguish
communication from transportation (Czitrom, 1982; Marvin, 1988). When the
telegraph was first introduced, “universal communication” became the key phrase. In
1838, for example, Samuel Morse wrote that it would not be long before "the whole
surface of this country would be channeled for those nerves which are to diffuse with
the speed of thought, a knowledge of all that is occurring throughout the land,
making in fact one neighborhood of the whole country" (As quoted in Czitrom, 1982,
p. 12). On the other hand, people also expressed pessimistic views about the
telegraph:
[It is] an improved means to an unimproved end... .We are in great haste to
construct a magnetic telegraph from Maine to Texas; but Maine and Texas, it
may be, have nothing important to communicate. ...We are eager to tunnel
under the Atlantic and bring the old world some weeks nearer to the new; but
perchance the first news that will leak through into the broad, flapping
American ear will be that Princess Adelaide has the whooping cough.”
(Thoreau, 1854, p. 36, quoted in Czitrom, 1982, p. 11)
Neither Morse’s optimism nor Thoreau’s pessimism was completely
appropriate. Today, neighborhood is still a smaller unit than a nation, and Texas and
Maine do communicate important issues via communication technologies. Rather, as
Czitrom (1982) states, “although its presence was not directly felt in everyday life,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
2
the telegraph eventually touched most people indirectly through the mass press it
helped create” (p. 29).
Since the telegraph, many more new technologies have been introduced to
human society. Similar arguments about them reflect hopes and fears occasioned by
new inventions (Czitrom, 1982; Marvin, 1988; Nye, 1994). For example, when the
motion picture appeared, many feared that it would destroy family interactions
(Belton, 1994; Czitrom, 1982; Sklar, 1994). When television appeared, people
predicted that radio would soon disappear (Bogart, 1956; Coffin, 1954; Douglas,
1999) or that true democracy would be realized (Bamouw, 1990; Smith & Paterson,
1998).
Among the hopes and fears that the advent of a new technology entails, its
influence on social inequality has been a frequent topic. In modem society, social
inequality has been an important social issue but, simultaneously, its complex and
multi-faceted nature makes it a continuing challenge. Efforts to resolve inequality in
many facets of society, e.g., reforming the educational system, providing equal job
opportunities, or providing subsidies to poor families, have been made, but with all
the different efforts to narrow the gap between those who have and those who do not,
social inequality continues to exist and the gap becomes even wider (Wolff, 2001).
Because social inequality is an enduring phenomenon, whenever a new device
comes into a society people often resort to predictions and rhetoric about the
possibility that the technology will resolve society’s inherent problems. For example,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
3
when radio was introduced, it was seen as a tool to overcome inequalities in terms of
geographic barriers in receiving news and entertainment information.
General Assumptions
Have past communication technologies resolved social inequalities? Can a
communication technology resolve social inequality? The investigator’s answer to
both questions is “no.” This dissertation provides a theoretical and empirical
rationale. However, before discussing theories of communication technology and
inequality, the investigator presents three general assumptions about communication
technologies and society as a baseline for the theoretical framework and hypotheses
testing throughout this dissertation. First, communication technologies are embedded
in a larger social system which has inequalities in terms o f income, education,
gender, or ethnicity. Society is an organic system that emerges, maintains, and
changes as various actors in the system take actions as individuals, groups,
organizations, institutions, nations, and beyond. As a new communication
technology enters the social system, it interacts with social actors and, sooner or later,
unless it fails, becomes embedded. Therefore, the effect of new communication
technology on society should not be assessed by considering technology in isolation
from, but as part of a social system, that is, in terms of how it does or does not
become successfully embedded.
Second, one new communication technology will not radically change the
inequality patterns in society. Because communication technologies are part of a
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
4
larger social system, the advent of one new technology is not likely to bring about a
radical change in the entire social system. The advent of a new communication
technology per se, for example, is not a solution to society’s existing disparities.
Third, although the previous assumption denied technology’s potential to bring
about a revolutionary change, it does not mean that a social system is static, or that
communication technologies do not make any difference. Communication
technologies, alone or embedded in social structures and processes, have the
potential to influence how actors act and interact in society. Various new and old
communication technologies shape how people access information that is important
for them to achieve their everyday goals, such as enhancing their career, socializing,
or spending leisure time.
The above three assumptions are general, functioning as a basis for
theoretical and empirical development in this dissertation. In addition, these
assumptions avoid the technological determinism of giving too much weight to
communication technologies while underemphasizing other components of society.
Set explicitly at the beginning of this dissertation, the three assumptions help keep
the theoretical and empirical development in this paper within the framework that
examines communication technologies within a larger social system.
Overview of the Dissertation
This study examines issues relating to social inequality and communication
technology. Both theoretical and empirical missions are pursued. Theoretically, the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
5
study reviews the evolution of media system dependency theory (Ball-Rokeach,
1985, 1998; Ball-Rokeach & DeFleur, 1976) into communication infrastructure
theory (Ball-Rokeach et al., 2001) over the past three decades. How communication
infrastructure theory originates and departs from media system dependency theory is
examined. Based on the two theories, the author develops the concept of
“communication connectedness” to illustrate the multidimensional relationships that
individuals form with communication media. Four dimensions of communication
connectedness include: access to technology; scope of activity participations within a
medium or across different media; intensity of relationships; and centrality of
communication media in daily lives.
Empirically, the four dimensions of connectedness are applied to the Internet
and operationalized to form the Internet Connectedness Index. With a dataset of 585
randomly selected telephone survey respondents, multiple factors that affect Internet
connectedness are considered and tested: socioeconomic status (income and
education); demographic characteristics (age and ethnicity); technological
environments (e.g., having a computer at home, the number of places for having
Internet access); social environments concerning the nature of social relationships
people have in their everyday lives (e.g., proportions of friends and family who go
online, interactions with others in receiving and providing Internet-related help);
connections to “older” communication channels (e.g., television, newspapers and
interpersonal communication); and the scope and intensity of Internet-related goals
that are important motivations for people’s incorporation of the Internet.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
6
This dissertation has six chapters. Chapter One (this chapter) presents a
general theoretical approach to the communication technologies and social inequality
issue, followed by a historical overview of the relationship between communication
media and social inequality. A “double barriers in communication technology and
equality” model is presented.
Chapter Two presents a theoretical framework by reviewing the evolution of
media system dependency theory (MSD) to communication infrastructure theory.
The concept of “communication connectedness” that comes out of the recent
development of communication infrastructure theory is presented.
Chapter Three is a review of past literature about communication
technologies and inequalities. Following the studies of older media and inequalities,
the author extensively reviews studies concerning factors affecting Internet and
inequality.
Chapter Four presents the hypotheses and the research methodology for
testing them. Data collection methods and the construction of variables and factors
are discussed.
Chapter Five presents results involving factors that affect people’s media
connections, Internet access and Internet connectedness. Multiple regression and
logistic regression analysis techniques are employed in the analyses.
Chapter Six draws conclusions from the theoretical and empirical
investigations developed in this dissertation. This chapter contains a general
discussion of the research findings in light of larger social issues.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
7
Media, Information and Society: A Historical Review
When the influence of communication media on social inequality is discussed,
the argument focuses upon the information resources that the communication media
are able to provide and the ways in which the media provide the resource. For
example, when the telegraph was invented, the possibility of “telecommunicating”
instantly triggered hopes of narrowing the gap between urban and rural areas, and
between poor and rich nations (Kieve, 1973; Lubrano, 1997). When the telephone
became popular, it was expected to unite the nation and connect people from all
different social sectors. For example, advertising copy from AT&T in 1915 states
that “the large task of the telephone is to connect communities and keep all the
people in touch, regardless of local conditions or distance” (Fischer, 1992, p. 163).
What kind of information resource is available to whom is a crucial factor
affecting existing disparities in society. The information resource is embedded in
many media genres, including not only “hard” knowledge such as news, academic
knowledge, or certain facts, but also “soft” knowledge such as entertainment
programs or guidelines for leisure activities. The meaning of the information,
therefore, is likely to differ among individuals, societies, and historical periods. As
Hobart and Schiffman (1998) state, “rather than attempting to find a single,
overarching definition of information, applicable across time and culture, we must
seek its unique meaning in each age, where technology and culture combine to
isolate different kinds of information” (p. 4).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The range and depth of information available to individuals or to certain
social groups evolved as social structures evolved and as communication technology
progressed. Before the newspaper was popularized, normal citizens did not have
direct access to political, economic, cultural or international issues that were beyond
their social circles or geographic areas. Even after the newspaper emerged as a mass
medium, those who did not have educational training and literacy to read newspapers
did not have direct access to the information they provided. When radio was
introduced and as the content evolved from mainly a “music box” to news and
connection of citizens to government, those who did not have the capacity to gain
access to the information in newspapers could receive information disseminated
verbally via radio. This was a drastic change in the kind of information resources
available to everyday people. In addition, the radio had an important advantage over
the printing press: providing access to information in real time. For example, in the
1930s and 1940s, everyday citizens listened to President Roosevelt’s series of
“fireside chats” as he delivered them. Also, the type of entertainment information
that was available to individuals on radio changed how people spent their leisure
time. For example, soap operas and comedy shows that were only available in
theaters became available in people’s households through the radio.
The advent of television again changed the type of information available to
people. The spread of television to average households in the United States in post
World War II was extremely rapid, increasing from 9% of the households in 1950 to
87% in 1960 (Brown & Bryant, 1989). Due to the wide use of television in US
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
9
households, television stations quickly shifted their target audience from a few early
adopters to average citizens, providing easy-to-understand and salient news that
caught people’s attention. Entertainment information also became an important genre
in television programs, extending those that had previously been broadcasted on
radio. Television became the primary way for everyday people to find out what was
going on in their society, get information on what to purchase, learn where to go for
entertainment or be entertained by television itself. As television matured, lower
income and lower educational groups had stronger connections to television than
their wealthier and more educated counterparts (Graham, 1999; Mullan, 1997; Petrie
& Willis, 1995), partly due to the primacy of television as an information source in
their everyday lives.
The advent of computers per se did not bring much change to the type of
information people accessed in their everyday lives. Rather, the computer served as a
prototype of the “information age” in academic and public discourse (e.g., Bell,
1973; Toffler, 1980). As the computer became popular in the mid to late 1980s,
disparities between those who had access to the computer and those who did not
started to be recognized as an issue that involved existing inequalities in larger
society (Dutton, Sweet, & Rogers, 1989; Steinfield, Dutton, & Kovaric, 1989). The
inequality discussions surrounding the computer were less about having access to the
information that the computer can provide to users, and more about the computer as
a tool for people to use and apply to create, process, and store information that was
important in their careers.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
10
The advent of the Internet changed the nature of information that computers
can provide to people. With the popularization of the World Wide Web in the early
to mid-1990s, the Internet became an essential feature of computers in providing
users access to information resources. The computer has become not only a tool in
itself that is used to accomplish work-related goals, but also a channel that connects
people to vast amounts of information. With the popularization of the Internet,
inequalities involved in accessing Internet technology have also become a prominent
issue of the time.
Intemet’s Significance
Inequalities in connecting to other technologies, such as compact disc (CD)
players, video cassette recorders (VCRs), cellular phones, or fax machines, have not
gained as much attention as the inequalities of connecting to the Internet. Why has
the inequality in accessing the Internet become such an important social issue? A
short answer is that resources available on the Internet are much more diverse and
richer than those provided by other technologies. CD players and VCRs provide
access to information, and cellular phones and fax machines, at this stage of their
development, mainly provide access to other people. In contrast, the Internet
provides direct access to information, people, services, and other technologies
(Dutton, 1999). It provides access to a diverse range of information, such as text,
images, video and audio (Dutton, 1999, p. 5). The Internet provides connections to
people by e-mail, chat rooms, web boards or Internet phones. It also makes available
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
various services such as shopping, banking and other transactions. Finally, it
connects people to other technologies such as teleconferencing equipment or web
cameras.
In addition, the ways in which the Internet delivers information to individuals
are more diversified compared to previous communication media. For example,
while traditional daily newspapers deliver information once a day, online newspapers
update information many times a day. In addition, the Internet operates as a platform
for users to communicate with one another synchronously. Chatrooms and instant
messagings are the examples. Internet is also flexible in the number of parties
involved in communications. Television, radio and newspapers mainly provide one-
to-many communications, while the telephone provides one-to-one communications.
The Internet provides one-to-one (e-mail), one-to-many (newspapers, commercial
websites) and many-to-many (bulletin boards) communications under its
infrastructure. Finally, Internet users can engage in interactive activities, such as
buying products, filling out forms or downloading documents or music. In sum, the
Internet gives individuals more control over where to connect, when to connect, and
whom to connect, which were more difficult and costly with previous
communication technologies.
Rich and diverse resources available on the Internet mean diverse
opportunities for individuals, communities, corporations and government to improve
their well-being. At an individual level, the Internet provides diverse groups of
people with educational and occupational opportunities (Bikson & Panis, 1999;
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
12
Castells, 1999; Dutton, Rogers, & Jun, 1987). For communities, the Internet provides
“socio-technological structure” to mobilize members to participate in community
activities (Castells, 1999, p. 35). Corporations use the Internet to lower the cost for
transacting various tasks (Malone, Yates, & Benjamin, 1987), and governments use
the Internet to enhance public services and encourage wider political participation
(Bennett & Entman, 2000; Norris, 2001; Wilhelm, 2000).
The various resources that the Internet provides have directly led to
predictions about the changes that the Internet will bring in society. Along with the
pessimistic visions that the Internet might widen the gap between the haves and
have-nots, many visions predicted that the Internet would narrow social inequalities
and give solutions to hitherto problematic social issues (e.g., Corrado & Firestone,
1996; Negroponte, 1995). Since its advent, has the Internet shown the potential of
narrowing social disparities? In the future, will the Internet resolve social inequalities
by providing easier access to diverse information to a diverse population?
Double Barriers to Communication Technology and Equality: Access and
Application
The author believes that technical potentials of communication technologies
will not directly translate to resolution of the “information gap” and thus to a
narrowing of social disparities. As the general assumptions suggested earlier in this
chapter, communication technologies are part of a social system where they interact
with various actors, institutions and other communication media. For example, when
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
13
a new communication technology, with new technical features that provide broader
and deeper access to information, enters the social system, various social forces that
have existed in society before the new technology is introduced shape the process of
dissemination and utilization of the communication technology (Bens & Mazzoleni,
1998; Dutton, 1999; Kim, 2001). In particular, the author proposes that two barriers
exist between the availability of information resources via various media and their
impact on resolving social inequalities. The first barrier is “connecting” to resources
available via a communication medium, and the other barrier exists in “utilizing”
those resources. Although a communication technology has the potential to make
various resources available to people and thereby become a steppingstone for
people’s social mobility, people are likely to face barriers, such as economic or
motivational disparities, in connecting to the resources provided by the medium.
Second, even after connecting to the resources and information, people are likely to
face another barrier in utilizing the resources for their various goals like career
development or political participation. In order for a communication technology to
narrow social inequalities, people should be able to overcome these two barriers.
Figure 1 visualizes the double barriers model.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
14
CONNECTING UTILIZING
Barriers Barriers
Income
Education
Age
Gender
Ethnicity
Individuals
Information
Resources in
Communication
Technologies
Television
Radio
Newspapers
Internet
Interpersonal
communication
Career
development
Socialization
Political
participation
Community
participation
Social Practice
Figure 1. Double Barriers to Communication Technology and Equality.
Applying the above double barriers model specifically to the Internet, one
notes that the issue goes beyond the popular discussion of a “digital divide” in
accessing the Internet. The majority of the digital divide discussion identifies
computer ownership or Internet access as a dependent variable, which is affected by
various socio-economic disparities in society (Novak & Hoffman, 1998; UCLA
Center for Communication Policy, 2001; US Department of Commerce, 1999, 2000,
2002). However, most of these discussions do not specify what it means to narrow
the digital divide. If everyone has access to the Internet, does it mean that the digital
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
15
divide has disappeared? What happens next? Does it narrow social inequalities? The
barrier model suggests that two barriers intervene between the Internet and its effect
on society: “connecting” to the Internet and “utilizing” Internet resources.
First Barrier: Connecting to the Internet
The first barrier concerns antecedent factors affecting “connection” to the
Internet. The potential of the Internet to enhance various opportunities is not a matter
of everyone having access to computers and the Internet. Inequalities regarding the
Internet mean more than access to technology. The ways in which people incorporate
the Internet should be taken into consideration (Ball-Rokeach, Gibbs, Jung, Kim, &
Qiu, 2000; Crigler, Just, & Greene, 2000; Jung, Qiu, & Kim, 2001; Lin, 1998; Loges
& Jung, 2001; Neu, Anderson, & Bikson, 1998; Norris, 2001; Wilhelm, 2000). The
question of how people incorporate the Internet into their lives involves not only
economic, but also educational and motivational factors.
In this approach, the digital divide can be defined as disparities in connecting
to Internet resources. The author conceptualizes “Internet connectedness” as
multidimensional ways in which individuals incorporate the Internet and the
resources provided by it (Jung et al., 2001). Four dimensions of connectedness are
access, scope, intensity, and centrality. Access is having a connection to the Internet.
Scope means the breadth of Internet activities people engage in, and intensity means
the degree of “helpfulness” of Internet resources in achieving one’s goals (Ball-
Rokeach, 1985, 1998). Centrality indicates the importance of the Internet, relative to
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
16
other communication media, in fulfilling everyday goals. More in-depth
conceptualization of the concept and its sub-dimensions are discussed in Chapter
Two.
The concept of “connectedness” is different from merely having access to or
spending a certain number of hours on the Internet. In examining the ways in which
people use the Internet, researchers have applied time measures in many studies
(Kraut, Patterson et al., 1998; Nie & Erbring, 2000; Putnam, 2000; Robinson,
Kestnbaum, & Neustadtl, 2002). Spending the same amount of time, however, does
not mean that people are using the media in the same way (Hawkins & Pingree,
1981; Jung et ah, 2001; Moy, Scheufele, & Holbert, 1999; Norris, 1996; Shah,
McLeod, & Yoon, 2001). First, the amount of time needed for people to obtain
specific resources on the Internet may differ. For example, a novice may need more
time to obtain a specific resource while an experienced person can accomplish the
same task more quickly. Second, time does not provide any information of the
breadth and depth of the activities that a person engages in on the Internet. One
person may spend a lot of time in one activity, while another person may engage in
several activities within the same amount of time. One study reports, for example,
that Internet newcomers spend as much time on the Internet as veterans, but many of
them limit their activities to entertainment and do not expand their activity scope to
engage in other activities (Rainie & Packel, 2001). In sum, the time measure does not
capture the ways in which people connect to Internet resources.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
17
When defined as a disparity in connecting to Internet resources, the digital
divide implicates a number of antecedent variables. In particular, three antecedent
factors are provided below as examples. First, the economic capacity to afford the
cost of buying computers and subscribing for Internet access is likely to affect
people’s connections to the Internet (Dutton et al., 1989; Thierer, 2000). Having
access to technology is a gateway for reaching information resources on the Internet.
In addition to the economic capacity, having computer-related skill is another
important disparity factor affecting the ways in which people connect to the Internet.
Many people who have computers or Internet access at home do not connect to the
Internet or mainly use e-mail because they do not have appropriate skills. Anderson
and Bikson (1998) specifically point out that people need “generic skills” to fully
utilize Internet resources (p. 1). Generic skills include cognitive abilities to
understand the logic of a hyperlink website structure, the ways in which data and
information can be structured, how to use generic tools such as search engines or
bookmark functions, and how to use video or audio media on the Internet (Anderson
& Bikson, 1998, pp. 2-3). Skill level is a crucial antecedent condition affecting
people’s connection to Internet resources.
People’s motivations for connecting to the Internet are also important.
Motivation to connect means first having goals that a person intends to accomplish
and second, perceiving the Internet as a resource that can fulfill the goals (Ball-
Rokeach, 1998). Lin (1998) included motivation as an important variable affecting
people’s adoption of computers. She emphasized that “perceived advantages and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
18
necessity will be the main reasons for adoption” of the Internet (p. 110). Flanagin
and Metzger (2001) explored the relationship between individuals’ motivations for
technology use and their choice of media for the different motivations. Among the
10 clustered groups of motivations, information-seeking, leisure needs and learning
were the main motivations for connecting to the Internet and to other mass media
modalities.
The UCLA Center for Communication Policy (2000) reports that one-third of
their Internet non-user respondents are simply “not interested” in going online (p. 22),
while the cost of going online is an important issue for only about nine percent of the
Internet non-users. Lenhart (2000) found that about sixty percent of those without
Internet access do not have motivations to go online. These motivational disparities
are not only keeping some people from going online, but they also are likely to shape
the scope and intensity of people’s Internet connections once they go online. People
who perceive resources as important, who have needs for innovation and who rate
the usefulness of the Internet high are more likely to connect to various resources
(Lin, 1998).
It should be emphasized that these three antecedent variables (economic
capacity, computer-related skills and motivations), which directly affect the ways in
which people connect to the Internet, are embedded in the larger social structure
where inequalities of income, education, occupation, age, ethnicity and gender exist.
For example, income is a strong factor affecting economic capacity, education level
affects the level of computer-related skills, and age and occupation are factors
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
19
affecting motivation. Therefore, social inequalities are likely to be reflected in the
ways in which people connect to the Internet (Jung et al., 2001). Making the
connection is, thus, the first barrier that intervenes between the various resources that
the Internet can provide to a diverse population and the real connections that people
have with those resources. It involves much more than having access to the Internet.
Second Barrier: The Barrier in Utilizing Internet Resources
Another obstacle intervenes between the Internet and its potential to narrow
social inequalities. Let’s be optimistic and assume that the first barrier with respect
to connections to the Internet resources has narrowed so substantially that the
disparities in scope and intensity of connections have almost disappeared. In order
for the Internet to narrow the inequalities in society, people still need the ability to
utilize these resources in appropriate venues to enhance their well-being. In other
words, the scope, intensity and centrality of people’s Internet connection are not ends
in and of themselves. They should be able to serve as independent variables that
contribute to narrowing inequalities. “Narrowing” inequalities means more than all
people utilizing the resources equally. It means that the disadvantaged people in
society are able to climb up the ladder to catch up with more privileged people. Even
if Internet resources are utilized equally between the advantaged and the
disadvantaged, the same gaps will exist and social inequalities will be maintained
rather than narrowed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
20
Barriers to utilization make it difficult for disadvantaged people to catch up
with the more privileged even after they connect to the Internet resources. Disparities
in utilizing Internet resources are shaped by people’s structural locations in society
(Ball-Rokeach, 1985), i.e., what position a person holds in the social structure. Three
major factors affect the utilization of Internet resources. First factor is a person’s
general education level. Having a connection to Internet resources is not sufficient as
a means to upward mobility. Resources have to be utilized in conjunction with a
person’s general educational resources. As long-term research on the knowledge gap
hypothesis suggested, higher education levels provide a better capacity for
understanding and retaining gained knowledge and for better recognizing the
relevance of the resources to achieve particular goals (Tichenor, Donohue, & Olien,
1980).
Occupation is another important factor that characterizes the structural
location of a person. Those who have a higher occupation status are more likely to
utilize Internet resources than those in a lower occupational stratum (Beeghley,
1978; Duncan, 1961). Even though people in a lower occupation level connect to
various resources, it is more challenging for them to utilize the Internet resources to
enter higher income jobs. A factory worker is not likely to become a software
entrepreneur by simply connecting to Internet resources because the structural
position of the person in society influences the ways in which the resources are
utilized (Amsden & Clark, 1999).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
21
In addition to education and occupation, a person’s social network is an
important intervening variable between the connection and the utilization of Internet
resources. In the process of utilizing and applying what a person has obtained on the
Internet, the social network of that person is crucial in helping him/her make better
use of the resources (Fulk, Schmitz, & Steinfield, 1990; Kraut, Rice, Cool, & Fish,
1998). The social network is important capital (Putnam, 2000), which facilitates
productive human activity by particular “social relationships, expectations,
obligations, and norms” (Alkalimat & Williams, 2001, p. 6; Putnam, 2000). Strong
social capital can be a mechanism to advantage those who are in a certain network,
but disadvantage those who are not. As Putnam (2000) stated, “People in
economically disadvantaged areas appear to suffer doubly. They lack the material
resources to get ahead, and they lack the social resources that might enable them to
amass these material resources” (p. 32). Bourdieu (1984; 1986) argued that having
“cultural capital,” which is largely determined by having certain family backgrounds,
is likely to affect the ways a person utilizes educational resources. To utilize
resources “in accordance with [a person’s] specific purpose, he or she must have
access to embodied cultural capital” (Bourdieu, 1986, p. 247). Social and cultural
capital is likely to intervene between having connections to Internet resources and
being able to fully utilize them to move upward in society. They form a barrier for
the Internet having the effect of narrowing social inequalities.
Two barriers, a barrier in connecting to the resources on the Internet and a
barrier in utilizing the resources obtained on the Internet, are both social forces that
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
22
shape who gets what kind of the information on the Internet and how that
information is utilized. These barriers emphasize the importance of social factors in
affecting the implication of a new medium when it enters a social system.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
23
CHAPTER 2. MEDIA SYSTEM DEPENDENCY TO
COMMUNICATION CONNECTEDNESS: A REVIEW OF THE
EVOLUTION OF MSD THEORY AND A PROPOSAL FOR A NEW
CONCEPT
Introduction
Whether people recognize it or not, the media environment of everyday life is
always changing. New communication technologies are having an impact on the
family environment, the workplace, and the maintenance of social relationships.
Nowadays, more and more media are available for people to get information and
communicate with others. For example, a person often talks with her friend on the
phone about an e-mail she has written. In the current communication environment,
therefore, focusing on one medium without putting it in context with other media is
inadequate (Flanagin & Metzger, 2001; Shah, McLeod et al., 2001). Studying one
medium in isolation from other media often leads to incorrect estimates of the role
the medium plays in people’s communication lives. Studying media in people’s
everyday lives, therefore, indicates that one medium should not be the focus of the
study. Instead, people’s everyday communication environments where they interact
with a number of media and other people are the focus of the study. An ecological
way of examining people’s communication systems is desirable.
This chapter reviews media system dependency (MSD) theory, which was
formulated and developed to explain the relationship between individuals, media and
society from ecological and multilevel perspectives. For each of the last three
decades (1970s, 1980s and 1990s), major theoretical works of MSD theory are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
24
discussed (Ball-Rokeach, 1985, 1998; Ball-Rokeach & DeFleur, 1976). Following
the thematic and conceptual discussion of these works, the chapter introduces
communication infrastructure theory, which grew out of MSD theory to explain the
21st century communication environment under a concrete and inclusive framework
(Ball-Rokeach et al., 2001). Based on the communication infrastructure theory, and a
response to the “dependency” concept limitations, the author introduces and
develops a “communication connectedness” concept. The communication
connectedness concept is multidimensional, as well as inclusive of interpersonal
communication and mass communication systems.
Media System Dependency Theory
Evolution o f Media System Dependency Theory: Overview
Each representing a decade, the following three diagrams demonstrate the
evolution of media system dependency theory. Subsequent to the three diagrams is
an explanation of the most important shifts in attention and evolution of thinking.
Media System
Audience
Societal System
Effects
Cognitive
Affective
Behavioral
Figure 2. 1970s: Effects-Focused Conceptual Model of Media System Dependency.
Adopted from Ball-Rokeach & DeFleur, 1976, p. 8
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
25
Within
Macro-
Level
Macro-
Level
Meso-
Level
Micro-
Level
Individual Media System Dependency
Relations
Structural Dependency
Relations
Media system
Economic system
Political system
Other systems
Social
Environment
Media System
Activity
1
W
f
Interpersonal
Network
Individual
Characteristics
Structural Location
Personal Goals
Figure 3. 1980s: Elaboration on the Origins of Individuals’ MSD Relations.
Adoptedfrom Ball-Rokeach, 1985, p. 499
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
26
Text
Personal
Goals
Personal
Environs
Social
Environs
Public MSD
Relations
Individual
Differences
Interpersonal
Network
MSD
Relations
Selectivity
Cognitive &
Affective Arousal
Behavior
Consumption
Organizational Structure
and Goals
Technological Form
Organizational Policies
and Procedures
Media System
Production
Characteristics
of Individuals’
MSD Relations
Variables
Intensity
Goal Scope
Referent Scope
Invariant
Resource scope
Structure
E
F
F
E
C
T
S
t t t
Macro- Meso- Micro-
Individual Interpersonal Organizational
System
Figure 4. 1990s: Elaboration of Meso-Level Dependency Relations and Cross-Level
Interactions.
Adopted from Ball-Rokeach, 1998, p. 18
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
27
Media system dependency theory underwent three important evolutions. First,
in explaining the individual’s media system dependency relations within a social
system, agents included in the models have changed. In the early work, audience,
media and society were the model’s main components. In her 1985 work, Ball-
Rokeach changed the micro-level unit of analysis from audience to individual. The
major change in the 1998 piece was an explicit attempt to include a meso-level agent
(the interpersonal network) along with micro- and macro-level agents.
Second, the individual’s media system dependency relations were given more
attention in the later works. In the 1976 work, the macro-level media and social
system were given more attention in the theory. Characteristics of individual-level
MSD relations were spelled out in more detail and more clearly in Ball-Rokeach’s
1985 and 1998 pieces. Particularly in the 1998 piece, she clarified the individual-
level media system dependency by contrasting it with the assumptions of uses and
gratification theory (Blumer & Katz, 1975).
Third, the focus of analysis changed. The 1976 piece focused on the effect of
the individual’s MSD relations on his/her cognitive, affective and behavioral changes,
reflecting the theoretical development at that time of “media effect” studies. In the
1985 piece, the focus became “determinants” of individual media system
dependency relations, i.e., which factors shape the characteristics of the individual’s
MSD relations. The 1998 piece described the ecology of the individual’s media
system dependency relations in an expanded model. Two major differences from
previous models were an active inclusion of the meso-level interpersonal network
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
28
into the model and specification of the media content flow as shaped by media
system production and individual consumption.
The Review o f Three Major Theoretical Studies
Beginning
After the focus on the hypodermic and powerful effect of mass media during
the World War II period and the popularity of the two-step flow model (Katz &
Lazarsfeld, 1955) from the 1950s to the mid-1960s, mass communication research in
the American academic tradition became markedly focused on individual-level
psychological processes and effects (Blumer & Katz, 1975; Katz, Blumer, &
Gurevitch, 1973-1974). As Katz, Gurevitch, and Haas (1973) described the
perspective at that time:
Media-related needs are not, by and large, generated by the media. Most
predate the emergence of the media and, properly, ought to be viewed within
the wider range of human needs. As such, they have always been, and remain,
satisfied in a variety of ways, most quite unrelated to the mass media.” (p.
180)
Media system dependency theory (Ball-Rokeach, 1985, 1998; Ball-Rokeach
& DeFleur, 1976; DeFleur & Ball-Rokeach, 1989) emerged from the discordance of
relying on psychological needs and gratifications in explaining people’s selection
and use of media.
As Ball-Rokeach and DeFleur (1976) stated:
The dependency model avoids a seemingly untenable all-or-none position of
saying either that the media have no significant impact on people and society,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
29
or that the media have an unbounded capacity to manipulate people and
society. It allows us to specify in a limited way when and why mass-
communicated information should or should not have significant effects upon
how audiences think, feel, and behave, (p. 19)
Sandra Ball-Rokeach and Melvin DeFleur, both of whom were trained in the
sociological discipline, developed the “media dependencies” concept based on the
belief that media and audiences are both integral parts of a larger social system. As
society gets more and more complex, the individual’s relationship with media cannot
be detached from other social organizations and systems. Durkheim (1933/1964)
once argued that requirements for shifting from traditional to modem society were
the capacities to deal with distance and heterogeneity. Durkheim’s argument
foreshadowed the advent of a communication mechanism, i.e., a media system that
can mediate dispersed and diverse individuals, institutions and societies. Media
system dependency theory builds upon Durkheim’s conceptualization of mass media
as an indispensable information system for society. It is different from
conceptualizing a media system as a persuasion system whose main concerns are to
persuade people with propaganda and campaign. As an information system, a media
system positions itself as an integral part of society where individuals, institutions
and other social systems cannot but connect to adapt to the changing social
environment.
From this perspective, Ball-Rokeach and DeFleur (1976) argued for the
necessity of employing an ecological and system perspective in studying
communication media and audiences. “If one hopes to account for changes in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
29
or that the media have an unbounded capacity to manipulate people and
society. It allows us to specify in a limited way when and why mass-
communicated information should or should not have significant effects upon
how audiences think, feel, and behave, (p. 19)
Sandra Ball-Rokeach and Melvin DeFleur, both of whom were trained in the
sociological discipline, developed the “media dependencies” concept based on the
belief that media and audiences are both integral parts of a larger social system. As
society gets more and more complex, the individual’s relationship with media cannot
be detached from other social organizations and systems. Durkheim (1933/1964)
once argued that requirements for shifting from traditional to modem society were
the capacities to deal with distance and heterogeneity. Durkheim’s argument
foreshadowed the advent of a communication mechanism, i.e., a media system that
can mediate dispersed and diverse individuals, institutions and societies. Media
system dependency theory builds upon Durkheim’s conceptualization of mass media
as an indispensable information system for society. It is different from
conceptualizing a media system as a persuasion system whose main concerns are to
persuade people with propaganda and campaign. As an information system, a media
system positions itself as an integral part of society where individuals, institutions
and other social systems cannot but connect to adapt to the changing social
environment.
From this perspective, Ball-Rokeach and DeFleur (1976) argued for the
necessity of employing an ecological and system perspective in studying
communication media and audiences. “If one hopes to account for changes in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
30
cognitive, affective, and behavioral aspects of people’s social realities brought about
by mass communicated information, one must take into account the
interrelationships between audiences, media, and society...individually, interactively,
and systematically” (p. 5).
They defined dependency as “a relationship in which the satisfaction of needs
or the attainment of goals by one party is contingent upon the resources of another
party” (p. 6). Based on the definition of dependency, the degree of media
dependency is defined as the “degree of audience dependence on media information
that is a key variable in understanding when and why media messages alter audience
beliefs, feelings, or behavior” (p. 5). The authors posed two theoretical hypotheses:
(1) “the greater the number and centrality of the specific information-delivery
functions served by a medium, the greater the audience and societal dependency on
that medium” and (2) “audience dependency on media information increases as the
level of structural conflict and change increases” (p. 7). That is, the media system’s
capacities to create, process and disseminate information, and the nature of social
environment are likely to affect the nature of individuals’ media dependency
relations.
The media system dependency perspective assumes that society goes through
periods of stability and change at different times. The more change and conflict exist
in the social environment, the more ambiguity individuals face. The ambiguity
makes people seek out information to make sense of what is going on in society, and
the media system is often an information-rich and necessary resource for that. As
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
31
people construct and reconstruct their social realities to reduce ambiguity, people’s
dependency relations with the media system intensify.
Ball-Rokeach and DeFleur characterized the relationship between audiences
and media as asymmetrical, media resources being more exclusive for audiences
than audience resources are for the media system. The exclusivity of media system
resources depends on (1) how scarce media resources are, (2) how essential those
resources are to individuals’ goal attainments, and (3) how critical it is for the
individual to attain a certain goal (e.g., a goal to make sense of an ambiguous social
environment vs. a goal to play on a boring afternoon) (Ball-Rokeach, 2002; Halpem,
1994). In the communication process between individuals or groups and mass media,
roles are divided, i.e., the media’s role of gathering, creating, processing and
disseminating content versus individuals’ limited role of selecting and disseminating
the content provided by the media to other individuals. Individuals are unaware of
the gathering and processing part of information production. For example, an
individual does not know what information a media institution is currently gathering
or how the institution is processing the information. The individual only knows what
the information is when reading newspapers or watching television, i.e., in the
disseminating stage. On the other hand, government is likely to be involved in the
information gathering and processing stage, providing the media system press
releases or other stories and influencing how the information is being presented in
the media.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
32
At the same time, the authors mentioned that the relationships between
society/media and audience are not unidirectional. As media and social systems
affect the audience’s dependency on media, the audience’s altered cognitive,
affective and behavioral conditions feed back to society and the media. The emphasis
on the “relationships” between media, society and audience are the main thrust of the
media system dependency theory that differentiates it from other theories, especially
from the more mainstream uses and gratifications theory of that era. The authors
argued that “proponents of [the] uses and gratifications approach examine how
audiences use the media to gratify similar information needs but do so by taking the
audience as the focal point of analysis, not the interrelationships between audience,
media, and society” (Ball-Rokeach & DeFleur, 1976, p. 8).
After presenting and explaining the main concepts and hypotheses of MSD
theory, Ball-Rokeach and DeFleur demonstrated the effects of media dependencies
on cognitive, affective and behavioral dimensions of people’s daily lives. The
authors suggested five cognitive effects that media dependencies are likely to bring:
construction of a social reality in largely ambiguous social situations, attitude
formation, agenda-setting, the expansion of people’s belief systems, and the impact
on values. In explaining the agenda-setting effect, the authors posed two interesting
questions: first, “why is there a considerable similarity in the agenda of concern
regarding certain types of topics among members of the media audience?” (p. 11),
and second, “in spite of such instances of similarity, why do members of the public
who attend to the media show numerous differences in their personal agendas of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
33
concern regarding media-presented topics?” (p. 11) In answering these questions
from an MSD perspective, Ball-Rokeach and DeFleur (1976) emphasized the
“interaction process” (p. 12) of agenda-setting where individuals share a social
position of having asymmetric dependency relations with the media system
(pertaining to the first question) which interacts with individuals’ diverse interests in
searching for media content (regarding the second question). That is, considerable
similarity in people’s media consumption exists because they share a similar
structure of media dependency relations under similar social, occupational and
economic strata. On the other hand, when it comes to which content elicits intense
relations, diversity in the personal agenda prevails.
As a concluding remark, Ball-Rokeach and DeFleur (1976) stated that,
The dependencies people have on media information are a product of the
nature of the sociocultural system, category membership, individual needs,
and the number and centrality of the unique information functions that the
media system serves for individuals and for society, (p. 19)
To the question of when the media system has strong and weak effects, media
dependency perspective suggests that when individuals have stable social realities,
the effect of media is less strong. In contrast, when individuals have unstable social
realities, media messages are more likely to affect individuals’ cognitive, affective
and behavioral activities. The authors saw contemporary social environments as
unstable due to the rapid rate of social change and the pervasiveness of social
conflict. Therefore, in contemporary society, the media system is likely to play a
central role in people’s everyday lives.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
34
Development
After the publication of the media dependency approach in 1976 (Ball-
Rokeach & DeFleur, 1976), the authors and other scholars further developed the
theoretical approach (Ball-Rokeach, 1985; DeFleur & Ball-Rokeach, 1982, 1989),
and conducted empirical studies on the consequences of media dependencies (Ball-
Rokeach, Rokeach, & Grube, 1984; Becker, Sobowale, & Casey Jr, 1979; Becker &
Whitney, 1980; Hirschburg, Dillman, & Ball-Rokeach, 1986). In the 1985 piece,
Ball-Rokeach focused on the “origins of individuals’ media system dependencies,”
(p. 486) that is, the antecedent factors that affect individuals’ media system
dependency relations. This paper examined individual’s media system dependency
as a dependent variable, rather than an independent variable.
First, individuals’ media dependencies are a consequence of the
interdependency between the media system and each of the other social systems.
“Individuals are bom into societies where the media system has, through its
resources and relations with other social systems, a range of
information/communication roles. It is that range of media roles that sets the range of
potential media dependencies of individuals” (p. 489). Ball-Rokeach called the
macro-level dependency relationships between media and other social systems
“structural dependency” (p. 490). Structural dependencies are mostly beyond the
control of individuals and meso-level social groups such as family, religious or
community organizations.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
35
In their 1976 piece, Ball-Rokeach and DeFleur referred to a micro-level unit
of analysis as “audience.” Ball-Rokeach (1985) shifted the micro unit of analysis
from “audience” to “individual” because “the mass audience does not act as a
coordinated unit vis-a-vis the media, nor does the mass audience, as such, control
resources or have shared goals that are necessary to meaningful analyses of
dependency relations” (p. 494). This change challenges the “media audience” as a
unit of analysis due to the diversified social and media environment where the
“audience” has become a more fluid concept that can change from one situation to
another, and from one period of time to another (See Neuman, 1991; Price, 1992). At
the individual level of analysis, the asymmetrical relationship between individuals
and the media system appears to be the central assumption of MSD theory. As Ball-
Rokeach (1985) stated, “[i]t is the media system that controls information resources,
and it is the media system that has relations with other social systems that shape the
dynamics and the content of individuals’ dependency on disseminated media
messages” (p. 488).
In explaining micro-level antecedents of media dependency relations, Ball-
Rokeach (1985) proposed the term “goal” as an adequate concept, making a
distinction from “need” as employed in the uses and gratification framework. She
stated that “[n]eeds connote both rational and irrational motives, conscious and
unconscious motives, and real and false interests. Goals, on the other hand, connote a
problem-solving motivation more appropriate to a theory of media behavior based
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
36
upon the dependency relation” (p. 494). That is, goals are the key dimension of
individual motivation preceding media dependency relations.
Ball-Rokeach proposed a typology of goals that affects individuals’ media-
system dependencies: understanding, orientation and play. Understanding is
specified as social understanding and self understanding. Social understanding is a
cognitive goal of trying to know how society and its institutions function at a given
time. Self understanding is the development of understanding about oneself through
the information offered by media and also from family or peers. For example,
individuals can gain knowledge about their health by watching health shows on
television. Orientation is specified as action orientation and interaction orientation.
Action orientation is a goal of getting to know how to behave in society. For
example, to gain information for voting or for purchasing a product is related to
action orientation. The interaction orientation goal is to acquire knowledge that is
necessary and suitable for interacting with other people. Play includes social play
and solitary play. The former is trying to engage in entertainment or to relax in the
company of other people. Examples of social play are going to a movie with other
people or watching television with other family members. Solitary play is the
entertainment or diversion of oneself from the more demanding and stressful parts of
life (Loges, 1992). Watching television alone or surfing the Internet are examples of
solitary play. These bi-dimensional three goals are exhaustive but not mutually
exclusive. That is, one person can pursue a social understanding goal at the same
time as engaging in social play.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
37
Although goals are an essential part of human motivation that affects
individuals’ media dependencies, they cannot fully explain the media dependency
relations because “a goal is an attribute and a dependency is a relation” (Ball-
Rokeach, 1985, p. 494). The dependency relations between individuals and media
system are shaped not only by individual goals, but also by the types of information
resources available in the media system and conditions of the social system.
Therefore, individuals with the same goals do not necessarily have the same media
dependency relations due to the “environmental factors” affecting media-dependency
relations (p. 495). Several antecedents that constitute the environment factors
affecting media system dependencies are discussed below.
First, the structural position of the media system in society affects the nature
of individuals’ media system dependencies. The structural dependency between the
media system and other social systems is a given social environment that affects
individuals’ media system dependencies. The social environment encompasses “all
environs that may bear upon individuals’ understanding, orientation, or play goals
whether they be international, national, community, or interpersonal” (Ball-Rokeach,
1985, p. 498). The effect of the social environment on individuals’ media
dependencies is shaped by two factors: the extent of the ambiguity or threat in the
social environment and how critical it is for individuals to define certain social
situations quickly. That is, the more ambiguous the social environment is, and the
more important a certain kind of social environment is for individuals, the more
intense their dependency relations with media are. This way of looking at the relation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
38
between social environment and media dependency parallels the knowledge gap
perspective (Donohue, Tichenor, & Olien, 1975), which argued that people’s
understanding of media messages is more homogeneous (1) when there is a
perceived conflict over a certain issue and (2) when an issue has a strong and
immediate impact on the local community. Both MSD and knowledge gap
perspectives take into account social environment conditions in examining
individuals’ relationships with media.
Not only the social environment, but also the activities of the media system
affect individuals’ dependency relations with media. The “message foci” of media
shape the availability of information that individuals can access. Media system
dependency theory assumes that the capacity of individuals to achieve their goals is
contingent upon the information resources of the media system—the resources being
media system’s capacities to (1) create and gather, (2) process and (3) disseminate
information (Ball-Rokeach, 1985, p. 487). The media system’s decision to select
certain stories and ignore others, and to produce certain genres and not others,
therefore, is an important force shaping individuals’ media dependencies.
Third, interpersonal discourse is another factor that shapes individuals’ media
dependency relations, particularly by affecting their everyday goals. Before engaging
in interpersonal communication, individuals are likely to seek certain goals
anticipating that the topic will be discussed in their interpersonal networks. In
addition, in the process of everyday conversation with other people, individuals are
likely to build up goals. For example, a person may go online to seek more
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
39
information about an organization that he or she heard during the conversation with
her colleagues. In many cases, individuals depend on media to achieve the specific
goals that emerge during their interaction with other people. On the other hand,
interpersonal discourse is also shaped by the message foci of the media system. For
example, the agenda of interpersonal communications is often shaped by the agenda
selected and disseminated by mass media.
Finally, the structural locations of individuals in society affect their access to
mass media and also to alternative information systems, such as experts or
organizations. Structural location includes “all the conventional stratification
variables, such as class, status, and power, and their consequences for proximity to
‘where the action is’ and the degree of investment in the outcomes of social action”
(Ball-Rokeach, 1985, p. 505). That is, who the person is, where the person lives and
what the person do affect his or her media dependency relations.
In the final paragraph of the paper, Ball-Rokeach differentiated her
framework from the typical notion of “functional analysis” which presupposes, in
most cases, stability of society and relies on dual notions of a strong media-weak
audience or a strong audience-weak media. Instead, she argued that “the powerful
audience of the uses and gratifications approach most likely coexists with the
powerful media that uses and gratifications [approach] rejects” (p. 503). The MSD
perspective is “an attempt to put forth the skeletal outlines of a sociological theory,
sociological in the sense that both macro and micro forces and their interrelations are
addressed...” (p. 507).
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
40
Clarification
From the 1980s to the 1990s, the media system dependency theory has been
cited in theoretical books (e.g., DeFleur & Ball-Rokeach, 1989; Littlejohn, 2001;
McQuail, 2000) and empirical studies (e.g., Ball-Rokeach et al., 1999; Boer &
Velthuijsen, 2001; Grant, Guthrie, & Ball-Rokeach, 1991; Loges, 1992; Morton &
Duck, 2000; Power, 1995; Skumanich & Kintsfather, 1998), which have directly or
indirectly applied MSD framework in supporting and conducting research. Many
studies, however, did not accurately explain the nature of individuals’ media
dependency relations. As Ball-Rokeach (1998) stated, “the tendency in the literature
to regard MSD as a macro theory that reduces to U&G [uses and gratifications] at the
micro-level suggests that past attempts to distinguish their theoretical approaches
have been too circumscribed” (p. 5). In the 1998 piece, Ball-Rokeach explained in
detail the micro MSD conceptualization as the main assumptions, and she compared
MSD variables with those of U&G. Ball-Rokeach went further than the 1985 piece
to discuss the theoretical origin and the building blocks of MSD theory vis-a-vis
U&G theory.
In the 1998 piece, Ball-Rokeach defined media system as “an information
system central to the adaptive conduct of societal and personal life” (p. 9). She
explicitly linked MSD theory to the ecological perspectives in sociology (Hawley,
1950; Park, 1922; Park, Burgess, & McKenzie, 1925), considering media as the
“interface between macro and micro units in community and city life” (p. 13). From
this conceptualization, Ball-Rokeach viewed mass communication and interpersonal
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
41
communication in the same ecology, creating, maintaining and changing
relationships with one another. By emphasizing this point, Ball-Rokeach reinforced
the main thrust of the MSD theory that structural dependencies in the macro level,
between media and political/economic system, inevitably affect the dependency
relations at the micro level, in interpersonal networks and in the relationship between
individuals and media.
Ball-Rokeach emphasized the cross-level links between micro-level and
macro-level as the core concept in the MSD relation, which was the main thrust in
the past works, but not explicitly framed as cross-level relations. Another difference
from the past work (Ball-Rokeach, 1985; Ball-Rokeach & DeFleur, 1976) was more
weight given to the micro-level agents’ role in the MSD relation. This does not mean
that the asymmetric relationship between individuals and media has changed, but the
nature of the micro-MSD relation was explained in a more detailed manner. Micro
level MSD relations are directly affected by types of goals that people seek to
achieve. In differentiating the characterization of micro-level consumption between
MSD and U&G, Ball-Rokeach argued that people’s relationships to media are “more
readily conceived as a characteristic that is open to change when it is treated as a
product of MSD relations than when viewed as a product of individual needs” (p. 23).
Needs are based on individuals’ psychological and socioeconomic origins, while
goals are not only based on psychological and socioeconomic origins, but also on the
personal and social environments that are constantly changing and affecting
individuals.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
42
Individuals’ goals are affected by personal environment, psychological
characteristics, and structural locations. Ball-Rokeach, as in the 1985 piece,
discussed structural locations of individuals as important factors that are likely to
affect individuals’ personal goals. Structural locations of individuals include
socioeconomic status, life cycle stage and lifestyle (p. 21). In addition to these
variables, a major structural location variable that affects MSD relations is “access to
alternative information systems” (p. 21).1 This brings up an important point that is
crucial in examining the incorporation of the Internet into the existing MSD relations.
If a person has a rich and broad communication network where different types of
information are accessible, access to the Internet is less likely to become a central
way of achieving goals, compared to a person whose available resources are limited.
In the latter case the person, when connected to the Internet, is more likely to form
an intense and broad relationship with the Internet where previously unavailable
information is available. Of course, skills and other resources required to obtain
necessary information on the new media mediate this hypothesis.
Ball-Rokeach (1998) went into details about core variables that characterize
macro and micro MSD relations: intensity, goal scope, and referent scope vary in
micro and macro relations, whereas structure and resource scope vary only in macro
relations (p. 19). Intensity is perceived exclusivity of resources for goal achievement.
For individuals, intensity is operationalized as perceived helpfulness of resources in
fulfilling personal goals. Goal scope is the range of goals implicated in dependency
relations. For individuals, goals refer to understanding, orientation, and play (Ball-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
43
Rokeach, 1985, 1998). The referent scope is the number of units involved in a
dependency relation. For the individual, this usually refers to the number of media
forms included in a dependency relation. At the macro-level, structure is the present
state of resource allocation in society, i.e., a structure of relations that ranges from
symmetrical to asymmetrical control over resources among different economic,
political and media systems (Ball-Rokeach, 1998, p. 19). Resource scope is the range
of resources-generating process that is involved in a relation. As stated earlier, media
system is involved in gathering, creating, processing, and disseminating of resources,
while only the dissemination resource is implicated in individuals’ media system
dependency relations.
In addition to discussion of the micro and macro MSD relations, Ball-
Rokeach extended the previous conceptualization of media system dependency
theory to incorporate meso-level dependency relations more actively by bringing
interpersonal network into the framework. Her conceptualization of interpersonal
network is different from that in the two-step flow model (Katz & Lazarsfeld, 1955)
where interpersonal communication is considered a buffer, enabling individuals to
maintain distance from mass media. Media system dependency theory emphasizes
that the interpersonal network can both constrain and facilitate the effects of media
on individuals. That is, interpersonal networks mediate the relationship between
individuals and the media system.
In her concluding section, Ball-Rokeach (1998) discussed the implications of
the Internet in the MSD framework. Arguing for the Internet’s evolutionary, rather
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
44
than revolutionary, incorporation into the existing system, she summarized that the
Internet would intrude on traditional relations “by being integrated into an expanded
media system that may expand the reach of understanding, orientation, and play
goals that individuals, groups, and organizations may attain through media
dependency relations” (p. 32). Media system dependency theory resonates with the
changing communication environment, but that is consistent with its ecological
framework.
MSD Theory Development in Empirical Studies
Micro-Level Dependency Relations
A majority of studies that directly or indirectly applied MSD theory
examined individuals’ media system dependency at a micro-level (Ball-Rokeach et
al., 1984; Becker & Whitney, 1980; Boer & Velthuijsen, 2001; Colman, 1990; Elliott,
1995; Grant et al., 1991; Loges, 1994; Loges & Ball-Rokeach, 1993; Manross, 1987;
Merskin, 1993; Morton & Duck, 2001; Power, 1995; Skumanich & Kintsfather,
1998; Waring, 1996). At the micro-level of analysis, the empirical referent of the
media system has usually been one or more of its forms, such as television,
newspapers, radio and magazines.
Ball-Rokeach, Rokeach, & Grube (1984) designed field experimental
research to study how “television induced long-range cognitive and behavioral
effects in the daily lives of adults in the privacy of their homes” (p. 151). Unlike
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
45
most other media effects studies that have already analyzed broadcasted television
programs, the researchers produced a 30-minute television program called the Great
American Values Test, which was broadcast simultaneously on the local CBS, NBC,
and ABC stations in three eastern Washington cities. By having the viewers self-test
their own values in their everyday lives, the program was designed to activate
viewers’ egalitarian and pro-environmental values subtly. Theoretically, under the
media system dependency framework, Ball-Rokeach et al. put together the typically
separated issues of selective exposure and message effects, proposing that while
individuals have considerable autonomy in designing media exposure, they are
situated in a media environment which determines the boundaries of message
availability (p. 15). In terms of selective exposure, the researchers found that media
system dependency was an important determinant of whether or not to view the
program, especially those who scored high on measures of social understanding and
self understanding dimensions in the media system dependency goals. At the same
time, the heightened attention of these people interacted with “uninterrupted viewing
to increase the likelihood of long-term value, attitude, and behavior effects” (p. 135).
Becker and Whitney (1980) examined different effects of television- and
newspaper-dependencies on people’s perceptions of government. They hypothesized
that newspaper dependency is positively associated with increased knowledge,
perceived comprehension and trust in government, while television dependency is
negatively associated with these variables in both local and national levels, when
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
46
controlled for age and education (pp. 100-101). Based on a telephone survey, the
results generally supported the hypothesis.
Grant, Guthrie, and Ball-Rokeach (1991) examined people’s television
dependency relations involving regular viewing of the home shopping network. The
authors first explained how television shopping as a genre changed the structural
relationships within the media system by directly connecting audiences with
merchandisers. Second, Grant et al. tested a model indicating the relationship
between individuals’ intensity and scope of television shopping dependency and
television shopping exposure, parasocial interaction with shopping hosts and
purchase behaviors. The authors found that the individuals’ dependencies on the
television shopping genre played a central role in their television shopping behaviors.
Colman (1990) examined changes in attitudes and beliefs after the broadcast
of a television movie concerning a woman’s use of holistic medicine and cancer. She
found that the intensity of television dependency relations in terms of the self
understanding goal was the best predictor of arousal to the content of the movie.
Loges and Ball-Rokeach (1993) examined the importance of “social
understanding dependency relations” (Ball-Rokeach, 1985; Ball-Rokeach et al.,
1984) in newspaper readership. The authors argued that demographic differences,
which have been the primary indicator of newspaper readership, do not explain as
much variance as dependency relations plus demographic variables do (Loges &
Ball-Rokeach, 1993, p. 603). Specifically, the authors found that the intensity of
social and self understanding dependency relations have significant positive effects
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
47
on the time people spend reading newspapers, when income, education, age, gender,
ethnicity and religion are held constant.
Meso/Macro-Level Relations
A few studies applied the media system dependency theory to meso/macro-
level relations and cross-level relations (Ball-Rokeach et al., 1999; Ball-Rokeach,
Power, Guthrie, & Waring, 1990; Fry, 1981). Ball-Rokeach, Power, Guthrie, and
Waring (1990) went beyond individual-level analysis to examine public opinion
formation on the abortion debate in the United States. The study examined the
parameters for individual opinion formation in the context of organizational
dependency relations between Pro-Choice and Pro-Life movements and the mass
media.
In another study, Ball-Rokeach and her colleagues (1999) used the media
system dependency theory to develop and implement an intervention program whose
goal was to change media production policies and practices (p. 229). Specifically, the
authors aimed at changing radio traffic report production policies so that radio traffic
reports did not reinforce an aggressive driving subculture. Rather than focusing on
the individual/consumption-level dependency relations, Ball-Rokeach et al.
examined the interconnectedness among the involved players, including radio
stations, traffic report production companies, governmental safety agencies, public
safety organizations and radio listeners. The intervention program was most
successful in introducing “personalization” into traffic reports, which included
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
48
references to people being involved in automobile crashes. The Ball-Rokeach et al.
study showed the applicability of the media system dependency theory in developing
social intervention programs that would likely be distinguished from conventional
intervention programs that mainly reach at-risk individuals.
Ambiguous Personal Environment
Several other studies examined the effects of an ambiguous personal
environment on individuals’ media system dependency relations (Hirschburg et al.,
1986; Kim & Jung, 2001; Loges, 1994; Nigg, 1982; Turner & Paz, 1986).
Hirschburg, Dillman, and Ball-Rokeach (1986) tested people’s media dependency in
the unique situation of Mt. St. Helens’ eruption in the state of Washington. The
researchers examined individuals’ “information-seeking activities” under a rare
situation of intense and pervasive ambiguity (Hirschburg et al., 1986, p. 118).
Hirschburg et al. (1986) argued that when ambiguity was experienced and when
interpersonal networks lacked information or expertise needed to resolve ambiguity,
“the media system [became] the major alternative information source.. .due to the
media’s structural position as an information system in our society” (p. 118). That is,
when the individuals faced an ambiguous social environment where they
continuously had to reconstruct their social realities in order to reduce ambiguity and
to adjust to the changing environment, which occurred in the case of Mt. St Helens,
people’s dependency on broadcast media (television and radio) became highly
intense, outweighing their dependency on interpersonal resources. Regardless of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
49
individuals’ demographic characteristics, people had intense media system
dependency relations for achieving their goals of understanding what was happening
and then finding out what they should do.
Nigg (1982) presented a “flow model of information exchange and
information-seeking pertaining to earthquake prediction and preparation,” examining
the interrelationships between formal and informal communication processes when
individuals are exposed to a threat of impending earthquakes (pp. 28-29). Based on
the ambiguity concept Ball-Rokeach (1973) discussed and the dependency model of
mass media effects Ball-Rokeach and DeFleur (1976) proposed, Nigg (1982)
explained five stages in which individuals become aware of the earthquake warning,
attend to media stories on the topic, engage in communications with other people and
consult experts on predictions. The author argued that the dynamic process by which
people seek out and exchange information is more important than merely focusing
on people’s media attention and retention results when studying the issue of
communication under an ambiguous situation.
Turner and Paz (1986) examined people’s dependencies on different media
sources for different kinds of information on the “earthquake threat” in 1976, 1977,
and 1978 when earthquake predictions in Southern California were reported in the
media (p. 99). Based on a field survey, a telephone survey and content analysis of
local media coverage of earthquake topics, the authors found that people mainly
depended on mass media (television, newspapers, radio) for earthquake information,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
50
and two-thirds of the respondents used interpersonal discussion to supplement the
mass media.
Loges (1992; 1994) examined the relationship between threat perceptions and
media system dependency relations. In survey data from two cities in the United
States, Loges found that those who perceived greater threat levels from crime, health
concerns, economic recession and environmental hazards had more intense media
dependency relations than others. Perceived threats were measured at personal and
social levels, and distinct types of threats, from economic problems to natural
disasters, were included.
Kim and the author of this dissertation (Kim & Jung, 2001) conducted
surveys in October 1999, two months before the turn of the century, when the fear of
“Y2K bugs,” the technological breakdown that might have happened on January 1,
2000 due to the inability of computer systems to recognize the year 2000, was a hot
social issue. In a path model based on 330 students from universities in Los Angeles
and Seoul, the authors found that when individuals felt personal and social threats
from the Y2K bugs, they “acted” to reduce ambiguities by intensifying their media
dependency relations (measured by intensity of Y2K-related goals), broadening the
number of media that they depend on and heightening their intentions for preventive
actions. Twelve modes of communication were included in assessing individuals’
media system dependencies (newspapers, national television, local television/cable
television, radio, talking with experts, talking with family, talking with friends,
magazines, books, movies, Internet, and catalogs or leaflets). Kim and Jung (2001)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
51
found that people formed intense dependency relations with mass media (newspapers,
television, radio) rather than interpersonal communications or narrow media
(magazines, books, catalogs/leaflets) when situated under ambiguous and threatening
situations like Y2K.
Social/Political Environment
Other studies examined people’s media system dependency relations in the
context of specific social and political climates (Halpem, 1994; Kim, 1982; Power,
1995; Rodriguez, 1982; Taylor, 1991). Halpem (1994) examined the effects of
individuals’ media system dependency on their political perceptions in Chile’s
authoritarian political system. He proposed that the absence of alternative sources in
an authoritarian political system was characterized by a lack of “functional
alternatives” for political information (p. 42). The lack of functional alternatives was
hypothesized to (1) increase people’s dependency on the available information
resources from the media system and (2) increase the influence of the media’s
messages on individuals (p. 45). Halpem found that the media’s degree of
exclusivity of information sources was a meaningful dimension that influenced
people’s pro-governmental political perceptions, when sociodemographic variables
and political affiliations were held constant.
Kim (1982) examined the relationship between individuals’ media
dependencies and their perceptions about political instability and mobility in the
context of 1970’s South Korea when the inflow of Western culture and the outbreak
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
52
of political protests occurred simultaneously. Dividing media into print and
electronic, and categorizing individuals into “elite” and “mass,” Kim concluded that
the elite exposure to the social environment, mainly through the printed media,
functioned as an important source of political instability and led to participation in
political protest. On the other hand, the masses, mainly dependent on television and
radio, remained politically immobilized (Kim, 1982).
Rodriguez (1982) examined how considerable media coverage of the
Philippine agrarian reform program affected audience perceptions of this issue.
Under the high degree of change and conflict in the Philippine agrarian situation,
individuals’ media dependencies were heightened, regardless of all demographic
variables.
Alman (1993) examined the relationship between the immigrant’s
acculturation into the United States and television advertising from the media system
dependency perspective. She found that immigrants (defined as foreign-born) who
had a lower cultural proximity to their country of origin culture, higher acculturation
motivation toward the host culture, and lower dependency on alternative information
sources (e.g., ethnic media) had a higher dependency on host television for fulfilling
acculturation goals. More importantly, a higher level of dependency on host
television was an important predictor for individuals’ use of the host television
advertising for acculturation purposes.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
53
Limitations o f the Media System Dependency Theory
Media system dependency theory has several theoretical and empirical
limitations. First, the term “dependency” tends to have connotative meanings of the
inherent asymmetric relationship between the two parties involved. For example,
when stated that person A has a dependency relationship with B, it is more likely to
be understood as A being subservient to B in their relationship. When the main thrust
of media system dependency is to examine the dynamic and changing nature of inter
relationships between media, society and individuals, naming the relationships as
“dependency relationships” is likely to obscure the theoretical clarity. This is why
many people misunderstood the ecological relationships defined in MSD theory,
especially the micro-level MSD relations as always being weak individuals
depending on a strong media system.
Second, MSD theory does characterize the asymmetric relationship between
individuals and media because of the degree and scope of control over resources that
are essential in the everyday goals of the involved parties. That is, individuals are
less likely to be an exclusive resource for the media systems in fulfilling their goals,
while the information provided by the media systems is likely to be the essential
resource for individuals. As stated earlier, the media system controls the
creating/gathering, processing and dissemination of information resources, while
only the dissemination of the information resource is implicated in individuals’
media system dependency relations. Therefore, the relationship between individuals
and the media system is assumed to be asymmetric. However, the nature of the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
54
asymmetrical relationship is changing. Under the transformation of new
communication technologies, the relationship between individuals and media
becomes a variable rather than a constant. That is, the dependency relations between
individuals and communication media are likely to become more diverse, expanding
the range of variations in individuals’ capacities to be involved in creating, gathering
and processing information. In such a media environment, the formerly assumed
asymmetric relationship between individuals and media may have to be critically
examined. For example, with a new medium like the Internet, individuals have the
potential to move from consumers to producers by creating or maintaining websites.
This does not mean, however, that on the Internet, individuals have become
empowered agents who can create and disseminate resources at the same extent as
the mass media system. Also, what it means to create stories on the Internet in
fulfilling everyday goals that people have is not clear at this point.
Third, in the MSD conceptualization, the role of meso-level agents has not
been sufficiently explicated. Although Ball-Rokeach’s recent work (1998) has
included meso-level interpersonal networks into the media system dependency
model, the overall focus of MSD theory has been on the relationships among
individuals, the media system and the social system, that is, between micro-level
individuals and macro-level media and social systems. The importance of meso-level
interpersonal network and social organizations is beyond question. In their everyday
lives, people are constantly connected to “meso-level” agents, such as community or
regional organizations or interest groups. Also, local or ethnically-targeted media are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
55
important meso-level agents that are embedded in people’s everyday lives.
Especially in the United States, which is geographically large and where cities and
states are quite distinctive from one another, the meso-level social system existing
between individuals and the macro social system should be considered as important
agents in examining media system dependency relations.
Finally, past MSD studies have not adequately explored empirical ways of
taking an individual’s personal environment into account. MSD theory hypothesizes
that personal environments surrounding individuals are likely to affect the intensity
and scope of goals individuals have in forming dependency relationships with media.
However, the relationship between personal environments and MSD relations is not
explicitly examined. In terms of the social environment, which is more macro-level
context than personal environment, several studies examined MSD relations under
certain social circumstances, such as natural disaster (Hirschburg et al., 1986), social
ambiguity (Loges, 1994), ideologically monopolized political environment (Halpem,
1994) or technological disaster (Kim & Jung, 2001). More immediate personal
environment involving family and friendship networks, work and organizational
contexts, and neighborhood environments were not included in the empirical studies.
Communication Infrastructure Theory
A Theoretical Framework
To be responsive to MSD theory limitations and to articulate macro, meso,
and micro relations more fully in a more specific social context, researchers
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
56
developed Communication Infrastructure Theory (CIT). The main hub where CIT is
being developed is the Metamorphosis Project at the University of Southern
California, under the supervision of Dr. Sandra Ball-Rokeach with other faculty and
doctoral students. The author has participated in the research effort since 1998.
As MSD theory’s main theoretical question is to uncover when media’s
power is strong or weak (Ball-Rokeach, 1985, 1998; Ball-Rokeach & DeFleur, 1976),
CIT seeks to uncover the often ignored but very important community configuration,
what they call “communication infrastructure” (Ball-Rokeach et al., 2001). CIT aims
to unveil the communication infrastructures in different social environments,
examine where and when it is strong or weak, and determine the difference a strong
or weak communication infrastructure makes. Communication infrastructure theory
seeks to understand how people reconstruct their social realities under the prevalence
of current world conflicts and under rapidly transforming social environments. The
concept of communication infrastructure pulls together all the available
communication resources in individuals’ social environments that can potentially be
used to achieve goals of understanding, orienting and relaxing in personal and social
worlds marked by change and conflict.
In a study that examined communication infrastructures of diverse urban
neighborhoods in Los Angeles, communication infrastructure is defined as “a
storytelling system set in its communication action context” (Ball-Rokeach et al.,
2001, p. 396). The storytelling system is a network of individual residents,
organizations and media in certain neighborhoods or communities. The ways in
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
57
which these “storytellers” connect and communicate with one another are crucial
factors assessing the viability of communication infrastructure in certain areas.
Using the same assumption as the MSD theory, the storytelling system
crosses multiple levels. The most important rationale for distinguishing micro, meso,
and macro levels of analyses is the variations in storytelling referents and imagined
audience (Ball-Rokeach et al., 2001). For example, macro-agents’ (e.g., network
television or mainstream newspapers) main referent covers a whole city, the nation,
and even the world, and their imagined audience is broadly conceived. On the other
hand, the main referent of a meso-level agent, for example a community newspaper,
is a particular part of a city, and the imagined audience is a particular group of
residents or business owners in the area. Community newspapers do report national
events, but their main focus is the implication of the events on the local area.
Ball-Rokeach et al. (2001) specified storytellers in each level. At a micro
level exist individuals living in a community. They are essential agents taking part in
the storytelling system. Local media and community organizations are main meso-
level agents. Local media refer to public media and ethnically-targeted media
covering certain neighborhood(s). Community organizations are any type of
organizations, including sports, recreational, cultural, ethnic, religious, political,
economic or educational organizations. These meso-level storytellers are given
prominent focus on the communication infrastructure theory as their activity and
participation in a storytelling system are crucial to the viability of the communication
infrastructure in their own right and for their linkage roles vis-a-vis macro- and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
58
micro-level storytellers. Finally, mass media (e.g., network television, national radio,
large newspapers), large organizations, and political, religious, and other institutions
that have production and dissemination resources for storytelling are included in the
macro-level (Ball-Rokeach et al., 2001).
Storytelling agents in multiple levels not only interact within a level, but they
also communicate across different levels. Individuals at the micro-level connect to
local media to find out what is going on in their community, and they participate in
organizational activities. Local television may connect to a local organization to
obtain a story for their news. The ideal storytelling system is broad (covering
referents from all three levels), intense (many stories from three levels) and
integrated (strongly connected network of agents) (Ball-Rokeach et al., 2001, p. 398).
The core of the storytelling system is the dynamic interplay, or what we call
“connectedness” among the agents, that affects the strength of the storytelling system.
In this sense, two types of connectedness exists: (1) on a system level, the
interconnectedness or integration of the storytelling network as a whole and (2) at
micro and meso levels, the connectedness between individuals/interpersonal network
and communication media (Ball-Rokeach, 2002). Further development in the
conceptualization of micro- and meso-level storytelling networks occurs in the
Metamorphosis Project (http://www.metamorph.org).
In MSD theory, personal and social environments are specified as
“communication action context,” which is resources in certain boundaries that are
likely to facilitate or constrain the viability of a storytelling network. The boundaries
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
59
are defined by shared conventions (e.g., major cross streets, incorporated areas, real
estate sections or geographic labels) (Ball-Rokeach et al., 2001). The communication
action context varies according to their openness and closedness. An open context is
one that encourages people and organizations to engage one another in
communication, whereas a closed context discourages such encounters (Ball-
Rokeach et al., 2001). For example, a community that has safe public spaces where
people can gather and talk with one another is considered to have an open context
that facilitates storytelling among residents. On the other hand, dangerous streets
discourage people from going for a walk or talking with neighbors on the street. Any
particular context has elements of both openness and closedness. Physical,
psychological, sociocultural, economic and technological dimensions are common
elements of communication action context.
Communication Infrastructure’ s Departure from Media System Dependency
In sum, the communication infrastructure (Cl) theory differs from media
system dependency (MSD) theory in four aspects. First, Cl theory is more inclusive
of all communication modalities available in people’s everyday lives. While the
main referent of MSD theory was the mass media system, Cl theory includes not
only macro-level mass media but also meso-level local and ethnic media.
Communication infrastructure also incorporates new and old media in its theoretical
framework. In addition, Cl theory gives much attention to interpersonal
communications as a focal storytelling network in individuals’ everyday lives.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
60
Second, the concept of the storytelling system regarding macro-, meso-, and
micro-level referents explicitly integrates narratives and sociological analyses within
a framework (Ball-Rokeach, 2002; Fisher, 1987). In MSD theory, the nature of
relations between individuals and mass media was mainly conceptualized as power
dependency relations in terms of the asymmetric resource allocation and goal
attainment. Under the Cl perspective, the storytelling narrative is one of the
important characteristics of the interconnectedness among the storytellers. The
relationship between individuals and communication media is not only a power-
dependency relation but also an ongoing communication process of creating and
sharing stories.
Third, in MSD, macro to micro effects of mass media were considered as
constraints on individuals’ agentic activities. In Cl theory, individuals are more
active participants in the storytelling system, interacting with macro- and meso-level
communication channels. Particularly, meso-level communication networks are
incorporated as “link” channels between individuals and the macro-level media
system. The meso-level communication networks, such as local and ethnic media
and interpersonal networks, are potential alternative mechanisms for individuals to
connect to when the information gained from mainstream mass media are not
sufficient or adequate. For example, when mass media portray certain ethnic group
inappropriately, individuals who belong to the ethnic group can connect to
interpersonal networks, community organizations or local media to overcome the
inappropriate portrayal.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
61
Fourth, Cl theory takes a more grounded approach in conceptualizing what
have been referred to as personal and social environments in MSD theory. The
“communication action context” specifies physical, social and cultural environments
that may facilitate or constrain the viability of the storytelling network. Therefore,
the very definition of the Cl theory links the social/personal environment and
everyday storytellers in a way that is more explicit and ecological in the levels of
analysis.
Communication Connectedness
Based on the media system dependency theory and communication
infrastructure theory, the author and her colleagues developed the term
“connectedness” at a micro-level which specifies variant and ecological relationships
that individuals form with various communication resources. The term was proposed
and used in several previous studies (Ball-Rokeach et al., 2000; Jung et al., 2001;
Kim & Jung, 2001; Loges & Jung, 2001), but this dissertation details the main
theoretical assumptions and sub-dimensions of the concept. Connectedness
encompasses the majority of theoretical assumptions implied in the “dependency”
relations. However, it differs from the dependency concept in several important ways.
First, connectedness encompasses a wider scope of resources in its referent, ranging
from interpersonal communication, new and old media, and small and big media
while the term dependency mainly focused on individuals’ relations to mass media.
The relationships between individuals and mass media have been the central focus in
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
62
the conceptualization of media dependency relations. However, other forms of
communication resources are as central in people’s everyday life goal attainment as
mass media. Also, a diversified media environment, including local media,
ethnically-targeted media and new media such as the Internet, requires a more
inclusive approach in examining people’s relations with communication resources.
Under the framework of the communication infrastructure theory, the connectedness
concept enables the examination of the relative importance of virtually all
communication modalities that are available in individuals’ everyday lives.
Second, connectedness is a more “neutral” term for characterizing
individuals’ relationships with communication resources, neutral in the sense that the
term itself does not connote any weight towards individuals nor to communication
media. The term dependency was often misunderstood as a personality characteristic,
e.g., a dependent person, and for that reason media system dependency theory has
sometimes been categorized as a “strong media, passive individuals” approach. The
important advantage of connectedness is that it connotes a relationship itself, rather
than implying an agent’s characteristic.
Third, individuals’ connectedness to certain communication media is
characterized as qualitative relations concretely defined by multiple dimensions:
access, scope, intensity and centrality. These ways of examining the relationships
that individuals form with communication resources is different from approaches
that mainly focus on whether people have access to certain media and how much
time they spend on the media. The author conceptualizes the four dimensions in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
63
next section, and reports empirical research of the relationship between “Internet
connectedness” and social factors in the next two chapters.
Dimensions o f Connectedness
Building on the qualitative relations between individuals and communication
media articulated in media system dependency theory and communication
infrastructure theory, the author defines four dimensions of connectedness that
implicate the quality relations between individuals and communication media: access,
intensity, scope and centrality. Access to a medium refers to a technology’s access,
which is a basic condition for forming a connectedness relation with the medium.
Intensity is the strength of the relations at a given time. The intensity is indicated not
only by the time spent with the medium but also by the perceived helpfulness of the
medium for achieving people’s everyday goals. For example, newspapers can be
perceived as extremely helpful for a person to find out what is going on in her
community, while television is perceived as less helpful for that goal. In this case,
the person has a more intense relation with the newspaper for that goal.
Scope is the breadth of the connectedness relations. It is examined both
across different media (media scope) and within a medium (activity/genre scope).
Media scope is the number of media forms that a person connects to when trying to
achieve a certain goal. Activity scope or genre scope is the number of activities or
genres that individuals engage in when connecting to a medium. For example, person
A mainly reads the sports section of a newspaper, while person B reads the political,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
64
economic and sports sections. In this case, B is considered to have a broader scope of
connectedness in her relation to newspapers.
Finally, the centrality of connectedness is a more subjective perception of
individuals about the importance of communication resources in their everyday lives.
Centrality implies an individual’s subjective appraisal of the relative importance of a
medium for achieving a certain goal in the context of other media. For instance, if a
medium is perceived to be exclusively central for achieving a goal, not being able to
use it would have a major effect on goal attainment (Halpem, 1994). On the other
hand, if a medium is a marginal resource for achieving a certain goal, or if the goal
can be easily achieved by connecting to other resources, the medium is less central
for the person, and the lack of the medium is not a critical factor for the person’s
connectedness pattern. This dimension situates a medium in the context of other
available communication resources in an individual’s communication infrastructure.
These four dimensions - access, intensity, scope and centrality - conceptualize
multidimensional connectedness relations that individuals form with communication
resources in their daily lives.
Summary
In this chapter, the evolution of the media system dependency theory, its
large-step expansion and its elaboration into the communication infrastructure theory
were examined. The work in this chapter was the first attempt to put together various
theoretical and empirical works published on media system dependency theory and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
65
communication infrastructure theory. First, across three decades, three important
theoretical developments of media system dependency theory (Ball-Rokeach, 1985,
1998; Ball-Rokeach & DeFleur, 1976) were thoroughly reviewed in terms of
changing conceptualization of agents included in the media system dependency
models, development of micro- and meso-level dependency relations, and changing
focus from mainly effects to the origins of the media system dependency relations.
Second, empirical studies that have applied and developed the media system
dependency theory were reviewed in four categories: micro-level relations,
meso/macro-level relations, ambiguous personal environments and social/political
environments.
Third, main definitions and framework of communication infrastructure
theory were introduced. Particularly, the ways in which communication
infrastructure theory owes origins to media system dependency theory and, at the
same time, how it departs from media system dependency theory were both
explained in this chapter.
Fourth, the concept of connectedness was explicated, stemming from
communication infrastructure theory and overcoming limitations of the
“dependency” concept. Connectedness reflects relational and multidimensional
conceptualization of the relationship between different storytelling agents.
In the next chapter, literature on communication technology and inequality,
in general, and the Internet and inequality, in particular, are reviewed, and the
limitations in the past literature are discussed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
6 6
1 When individuals cannot find information for achieving their goals, or when they refuse to
access a certain media system, people may form a coalition and create an alternative information
system. As Ball-Rokeach stated, “alternative media include oppositional information systems and
subcultural media” (p.25). One of the subcultural media that became a central referent in later
evolution of MSD theory is local and ethnically-targeted media (Ball-Rokeach et al., 2001).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
67
CHAPTER 3. NEW COMMUNICATION TECHNOLOGIES AND
INEQUALITY: LITERATURE REVIEW
Introduction
This chapter reviews past studies about new media (several of which have
become “old media”) and inequality in how individuals access a new communication
technology and develop relationships with it. Inequalities in diffusion and use of
media have not been popular research topics in the communications and media field.
Many studies focused on the effects of media, or what is called “media effects,”
assessing the degree of impact media have on individuals, organizations and society.
In particular, theoretical development in the area of communication media and
inequality has not been active. First, this chapter reviews two research lines that gave
explicit focus to the communication media and inequality issue: the “diffusion of
innovation” and the “knowledge gap” approaches. Second, this chapter reviews
empirical studies that addressed the issue of communication technologies (either new
or old technologies, or both) and inequality. Third, early studies on inequality in
Internet access are reviewed and the limitations of those studies are discussed.
Finally, recent studies that examined the digital divide issue from more diversified
approaches are reviewed.
Diffusion of Innovation
Diffusion of innovation has been one of the most active research areas in the
communication field since the early 1960s. Such studies concern the process of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
6 8
introduction, diffusion and adoption of a new technology in a society. Everett Rogers
(1983) framed the theory in two concepts: uncertainty and information. Uncertainty
indicates a lack of future predictability, which motivates individuals to seek
information. One kind of uncertainty, Rogers stated, is generated by “innovation,”
defined as “an idea, practice, or object that is perceived as new by an individual or
another unit of adoption” (p. xviii). When a new technology is introduced to society,
individuals and organizations are presented with a new alternative for solving their
everyday problems. However, the characteristics of the new alternatives are not
exactly known to individuals and, therefore, they are motivated to seek further
information about the innovation to reduce their uncertainty about it. In the process
of seeking information to reduce uncertainty, interpersonal communications with
close friends and family members are crucial in assessing a new innovation’s role.
Diffusion is “the process by which an innovation is communicated through certain
channels over time among the members of a social system” (Rogers, 1983, p. 5).
Therefore, Rogers characterized the diffusion of innovations process as “essentially a
social process in which subjectively perceived information about a new idea is
communicated” (Rogers, 1983, p. xix).
Four main elements in the diffusion of innovations research are innovation,
communication channels, time and a social system. Innovation is mostly discussed as
a new technology (Rogers, 1983). Communication channels include mass media and
interpersonal channels. Time involves an individual’s mental process of first getting
to know the innovation and then forming attitudes about it. Time also indicates the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
69
“innovativeness” of an individual, categorized into innovators, early adopters, early
majority, late majority and laggards. In addition, the rate of adoption among
members in a social system involves time. Finally, a social system is “a set of
interrelated units that are engaged in joint problem solving to accomplish a common
goal” (p. 37), where a social and communication structure may facilitate or constrain
the diffusion of innovations in the system.
Various academic fields conducted diffusion of innovations research,
sociology, education, public health, anthropology and marketing (Rogers, 1983).
One early concern of communication researchers was the diffusion of news events
carried by the mass media (Deutschmann, Ellingsworth, & McNelly, 1968;
Greenberg, 1964) and the other was communication processes involved in diffusing
medical practices and medicines. In Coleman, Katz, and Menzel (1966)’s seminal
study, the authors found that people tend to resist medical changes and new
innovations in treatment and care. They concluded that people resist change in
medical practice without considering whether or not the change will be beneficial
because they resist change of new innovations in general. Many scholars have
studied diffusion of technological innovations in the United States (Dillman, 1985;
Okolica & Stewart, 1996; Rogers, 1995; Speter, 1996; Van Slyke, 1997) and in
developing nations (Athreye, 1999; Besley & Case, 1993; Feder, Just, & Zilberman,
1985; Kyle Robison & Crenshaw, 2002; Maksabedian, 1980; Midgley, 1984; Stone,
2002; West, 1983; Xu, 2000).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
70
Among the antecedents of individuals’ adoption and use of a new innovation
is socioeconomic status. One study evaluated the effect of socioeconomic status on
people’s use of a new technology. Case (1982) examined the impact of the “Green
Thumb,” a computer-based information system that delivers frames of weather,
market and other information on demand to a farmer on his home television set. The
US Department of Agriculture provided Green Thumb boxes to Kentucky farmers at
no cost for fifteen months. Therefore, socioeconomic status did not affect the access
stage. However, the degree of use of the Green Thumb by farmers was related to
socioeconomic status. Farmers with larger farms, i.e., those who were better off,
valued the information from the Green Thumb and relied on it more for their
decision making than the farmers with smaller farms. This example shows that even
when a technology is given to everyone, if a person does not have an ability to see its
value and make use of it, then the technology is incorporated differently in the
person’s everyday life. Also in the early 1980s, Rogers (1983) examined
socioeconomic status of Kentucky farmers who owned computers at home. All these
farmers and those who expected to adopt a computer at home were well educated
and owned a large farm.
The diffusion of innovation approach emphasizes the importance of existing
social systems in shaping the ways in which a new innovation is incorporated into
society. According to Rogers (1983), when a social system is already highly
stratified, the diffusion of a new technology, especially when it is relatively high-cost,
may lead to even greater inequality for three reasons: (1) a high socioeconomic status
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
71
group has greater access to information that leads to the awareness of a new
technology, (2) the high socioeconomic status group has greater access to
information from peers for evaluating the new technology and (3) the group
possesses greater resources required for adopting the new technology (p. 412). The
in-depth theoretical development of the diffusion of innovation theory and its
emphasis on the diffusion of a technology as a social process have made the theory
influential in the adoption and diffusion of new communication technologies.
Knowledge Gap
Another line of research whose main concern has been inequality in people’s
access to media-provided information is the knowledge gap theory (Donohue, Olien,
& Tichenor, 1987; Ettema & Kline, 1977; Gaziano, 1995; Gaziano & Gaziano, 1999;
Tichenor, Donohue, & Olien, 1970). Its basic idea is that as the infusion of mass
media information into a social system increases, groups of the population with
higher socioeconomic status tend to acquire this information at a faster rate than the
lower status groups, so that the gap in knowledge between these groups tends to
increase rather than decrease (Tichenor et al., 1970). Tichenor and his colleagues
believed that knowledge gaps are functional for the existing power holders to
maintain their privileged positions in society. They claimed five reasons contribute
to the widening knowledge gap: (1) communication skills differences between SES
levels, (2) higher level of stored knowledge in higher SES, (3) greater contact with
experts and people in higher SES, (4) higher SES being less susceptible to selective
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
72
exposure and (5) information in mass media that is geared toward higher SES (e.g.,
public affairs programs).
In 1977, Ettema and Kline published a now well-known piece in which they
focused on individual differences in perceiving and understanding knowledge.
Ettema and Kline (1977)’s key difference from Tichenor et al.’s definition of
knowledge gap is that inequalities are based more on situational and motivational
perceptions of knowledge than on social system stratification. They proposed a
revised hypothesis:
As the infusion of mass media information into a social system increases,
segments of the population motivated to acquire that information and/or for
which that information is functional tend to acquire the information at a faster
rate than those not motivated or for which it is not functional, so that the gap
in knowledge between these segments tends to increase rather than decrease.”
(p. 188)
They also believed that a “ceiling effect” from artifacts, messages and audiences
exists and is likely to limit the widening knowledge gap between “haves” and “have-
nots” (pp. 197-199).
Tichenor et al.(1970)’s and Etteman and Kline (1977)’s hypotheses lead to
quite a different interpretation when examining the inequalities in adopting and using
a new communication technology. According to Tichenor et al’s perspective,
people’s adoption and use of new media are likely to follow the inequality pattern
existing in the current social system. As those from the higher SES group adopt and
use the new medium more effectively, the knowledge gap between them and lower
SES group is likely to widen. According to Ettema and Kline, on the other hand,
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
73
whether the new medium is functional for different individuals in terms of their
motivation to adopt and use the medium is a more important factor than SES
differences in the knowledge gap acquisition (Bogart, 1956) from the new medium.
Both perspectives have been important theoretical guidelines for the later research on
media and inequality (e.g., Chew & Palmer, 1994; Cohen, Salomon, & Nikjamp,
2002; Eveland & Scheufele, 2000; Gaziano, 1983; Holbrook, 2002; Kwak, 1999;
Mulugetta, 1987; Speight, 1999; Suss, 2001; Viswanath & Finnegan Jr., 1996).
New Technology and Inequality Studies
Compared to a few theoretical approaches, more empirical research has been
done recently on the adoption and use of communication technologies. Lin and
Jeffres (1998) studied factors influencing multimedia cable technology adoption. The
authors predicted that demographic variables would have weak associations with the
likelihood of its adoption. On the other hand, they predicted a substitution effect of
the cable technology over older technologies, and thus hypothesized that lighter
media users, along with less satisfied media users, will be more likely to adopt the
new multimedia cable technology. In addition, individuals who express stronger
needs to keep up with new communication technology were predicted to be more
receptive to new cable technology. Based on the telephone survey of 301
respondents, the result refuted the authors’ prediction that demographic variables
would have weak association with the adoption rate. Strong and moderately strong
relationships were found between marital status, higher income and higher
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
74
educational level and the likelihood of adopting the cable technology. The
substitution effect of the cable technology was found for network television services.
Intention to keep up with new technology was a stronger predictor of cable adoption
than the perceived need for owning it.
Also from the uses and gratification perspective, but taking a more ecological
approach than Lin and Jeffres (1998), Cohen, Levy, and Golden (1988) examined the
VCR (videocassette recorders) in households with six “old” media, including
newspapers, radio, television, books, cinema, and records/audio tapes. The authors
compared the gratification levels of these media for twelve social and psychological
needs. Cohen et al. found that although the VCR was often used, it was not given any
special priority to gratification level. VCRs received about middle ranking as a
means of need satisfaction (Cohen et al., 1988). In addition, VCRs were perceptually
indistinct from other older media, such as television and cinema. Overall, the authors
concluded that VCRs did not create a revolution in the ways people communicate at
home.
Steinfield, Dutton and Kovaric (1989) provided a framework for examining
home computing. Staying away from the technological deterministic view of the
effect of computers on individuals and families, Steinfield et al. provided two
dimensions in describing computer use: amount of time and variety of use. They
assumed that personal computing would evolve in a natural progression from the
limited use of a few applications to a heavier use of many. They identified social
status, personal attributes (e.g., values, attitudes), socio-cultural settings (social
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
75
network) and technical features as factors shaping use patterns (amount of time and
diversity of use), and also suggested areas that computer use was likely to have
impact on, including education, family, work from home, household routines and
civil liberties. The framework that Steinfield et al. suggested is likely to be
applicable not only to computer studies, but also to other information and
communication technologies.
Dutton, Sweat and Rogers (1989) empirically tested a similar model. Dutton
and his colleagues found that education and occupation do not directly affect early
adoption of home computing but instead act indirectly via income. They concluded
that education and occupation are “necessary but not sufficient conditions for
adoption.. ..[IJncome appears to be the sufficient condition” (pp. 266-267).
Attewell and Battle (1999) examined the effects of home computers on
adolescents’ school performance. They found that although having a home computer
was associated with higher test scores after controlling for family income and social
capital, children from higher socioeconomic status (SES) homes achieved larger
educational gains from home computers than did lower SES children. In addition,
given the same technology, boys performed better than girls, and Whites gained far
more than ethnic minorities. Attewell and Battle showed that even among those with
access to the same computer technology, existing social disparities are likely to
affect how technology is being used and what people gain from it.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
76
Internet and the “Digital Divide”
Although many studies focused on communication technology and inequality
before the Internet, numerous other studies have been conducted since the
widespread diffusion of the Internet. The term “digital divide” has been a widely
discussed public, academic and policy issue within the last decade. Several factors
boosted its social focus. First, the exponential expansion of the World Wide Web
drastically increased the awareness of and rate of access to the Internet, making it a
prevalent medium in people’s everyday lives. Second, the policy direction of
Clinton/Gore administration to connect the whole nation in a “national information
superhighway” brought public and academic resources to bear on studying and
improving inequality in uses and utilization of the Internet. Third, due to the
interactive nature of Internet technology and the abundance of information and
communications available on the network of networks, the Internet’s potential to
lower the barriers for less advantaged groups of people in society to gain access to
information and, simultaneously, the need to make sure that those people do not lag
behind in using the new technology have both heightened the importance of
resolving inequalities in accessing and using the Internet.
Several research efforts have led the digital divide discussion. The United
States’ National Telecommunication and Information Administration of the
Department of Commerce (1995; 1998; 1999; 2000; 2002) issued nationwide reports
based on Current Population Survey (CPS) data, including demographic pictures of
computers and Internet users. These reports demonstrated disparities in terms of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
77
income, education, age, gender, ethnicity and geography regarding access to the
computer and the Internet. A set of reports from the Pew Internet and American
Project (http://www.pewintemet.Org/j examined various aspects of people’s Internet
access and use, such as issues of gender, music, trust, privacy, Internet non-users,
health, Internet commerce, teenagers, ethnic groups, education, online communities
and online job hunting. The Benton Foundation
(http://www.benton.org/Resources/home.htmlj has advocated and reported on the
importance of digital divide issue and has been engaged in enhancing Internet and
computer connectivity in low income communities in the United States.
In academia, Thomas Novak and Donna Hoffman at Vanderbilt University
have studied race and socioeconomic disparities in Internet access and Internet
commerce issues from the mid-1990s (Hoffman & Novak, 1996, 2000; Novak &
Hoffman, 1998). The authors found that a significant portion of ethnic difference in
computer ownership and Internet access is explained by income and education.
Another well-known study, so called HomeNet study at Carnegie Mellon University
(Kraut, Patterson et al., 1998), reported negative effects of Internet use on social
involvement and psychological well-being, but later their longitudinal study of the
same respondents found that these negative effects had dissipated. Rather, they found
positive effects of Using the Internet on social involvement, well-being and
communications (Kraut et al., 2002). The UCLA Center for Communication Policy
(http://ccp. ucla.edu/j released three reports in 2000, 2001, and 2003 providing
profiles of Internet users and non-users on five subjects: basic demographic and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
78
socioeconomic profile, media use and trust, consumer behaviors online,
communication patterns and Internet’s effect on society.
Limitations of Past Studies
Despite the large number of reports and studies devoted to the digital divide
issue, most stopped at the preliminary results about the inequality in access to the
Internet technology. Although many studies reported disparities in Internet access
with regard to income, education, age, gender, ethnicity and geography, the
descriptive results did not provide sufficient information to intervene on behalf of
those who lag behind in Internet access. This is also due to the conclusion of several
studies that the “divide” is narrowing and will continue to narrow (e.g., Nie &
Erbring, 2000; Thierer, 2000; US Department of Commerce, 2002).
Second, even when studies looked beyond a simple access/non-access
disparity to examine different usage patterns among Internet users, many studies
isolated individuals’ Internet use from their social and communication environments.
Several studies, for example, relied solely on the amount of time individuals spent on
the Internet to assess its effect on individuals’ lives. For example, the Carnegie
Mellon University study of Pittsburgh residents reported that the Internet has the
potential of making people socially isolated, lonely and depressed (Kraut, Patterson
et al., 1998). Based on the assumption that “the Internet could change the lives of
average citizens as much as did the telephone in the early part of the 20th century and
television in the 1950s and 1960s” (p. 1), Kraut et al. found that greater use of the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
79
Internet was associated with a decline in communication with family members in the
household, a decline in the size of people’s social circles and an increase in their
depression and loneliness. Their striking conclusion, however, was theoretically and
methodologically problematic. In addition to the overly deterministic perspective
about technology’s potential to change people’s social and psychological patterns, as
shown in the quote above, the researchers did not allow the possibility that people
can communicate on the Internet with people they already know offline, and that
online communications can serve to supplement or extend offline social relations.
That is, Kraut et al. did not consider other social factors that may affect the ways
people incorporate the Internet into their everyday lives.
Nie and Erbring (2000), from their nationally recruited sample, concluded
that the more time people spend with the Internet, the more contact they lose in their
social environment. Among their “regular Internet users,” defined as those who
spend more than 5 hours a week on the Internet, 26% spent less time with family and
friends either in person or on the phone and 8% attended fewer social events than
before. Based on their preliminary analyses, Nie and Erbring predicted that “the
Internet could be the ultimate isolating technology that further reduces our
participation in communities even more than television did before it” (p. 19). They
saw that the Internet is an individual activity that requires more personal engagement
and attention than other media and is more likely to isolate people. Nie and Erbring
drew conclusions about the effect the Internet had on social isolation based on the
sole measures of hours of Internet use and hours of engaging in other activities. This
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
80
approach does not sensitively reflect how the Internet is being incorporated into
established communication environments.
Third, most studies did not consider older ways of communicating when
examining Internet inequalities. When the Internet was introduced to society and
people’s everyday lives, it was a new medium, but not the only available medium.
People had other ways of communicating, ranging from mass media to interpersonal
communications. When examining the nature and kind of disparities in people’s
connections to the Internet, it is important to examine what kind of communication
media have been available and where the Internet fits in the matrix of
communications in people’s everyday lives (Ball-Rokeach et al., 2000; Flanagin &
Metzger, 2001; Jung et al., 2001).
Approaching the Digital Divide
Several recent studies went further to overcome past limitations. Castells
(1999) argued that diffusion is not enough to reverse information inequality. He
thought that “informationalization”—the reliance of economy, politics and culture on
information processing in global networks via information and communication
technology—and “dualization”—the divide between “high-value making group” and
devalued group—are intertwined in current society (p. 28). To resolve the disparities,
he argued that information technology should serve as an access to jobs and income
generation by enhancing social and political resources to educate people to make use
of the technology in current society (p. 35). Van Dijk (1997) suggested four
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
81
obstacles to information society access: no access to computers and networks; lack
of basic skills and computer fear; insufficient user-friendliness of the resources
available on the Internet; and insufficient and unevenly distributed usage
opportunities of software and services available in computers and the Internet (pp. 2-
6). He argued that these obstacles magnify the effects of socio-cultural, material, and
political disparities on information inequality.
Ball-Rokeach and her Metamorphosis Project colleagues (2000) at the
University of Southern California argued that possessing computer technology or
Internet access does not directly translate into a “global village” where people of
different backgrounds come together to connect with one another. From their survey
and focus groups of seven new and old immigrant groups in the Los Angeles area
(for detail, see www.metamorph.org), they found that (1) a significant divide existed
among different socioeconomic and ethnic groups regarding computer ownership
and Internet access and (2) the ways in which people used the Internet largely
followed ethnic lines where people mostly connected with their friends and family
via the Internet or visited websites of their own countries of origin (in case of new
immigrants), rather than made “new friends” or visited websites of other cultures
(Ball-Rokeach et al., 2000).
Newhagen and Bucy (2001) approached “access” from multiple routes. They
defined access as “meaning, and involving not just the physical presence of
computers, but includes having an adequate connection to the Internet as a system, as
well as the user’s social capital and cognitive ability” (p. 3). They argued that what is
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
82
available on the Internet becomes “information” when it has meaning to users.
Therefore, true access to the Internet means access to content, rather than the
technology per se. Lievrouw (2000) too pointed out that not only is access to
telecommunications systems important, but also the access to Internet content is
crucial in achieving social and political participation via the Internet.
Hargittai (2002) pointed out study shortcomings that only viewed the binary
disparity between those who use the Internet and those who do not, and examined
disparities in people’s online skills, or what she calls the “second-level digital
divide.” By assigning search tasks to a random sample of Internet users, she found a
considerable difference in whether people could find various types of content on the
Internet and how long it took them to find it. LaRose, Eastin & Gregg (2001) tested
mediating effects of self-efficacy on the relationship between Internet use and
stress/depression. They found that those who have higher self-efficacy coped with
the stress and difficulties of using the Internet better than those who have lower self-
efficacy.
Patterson and Wilson (2000) and Bonfadelli (2002) both cautioned the
possibility of the Internet exacerbating society’s already existing inequality.
Consistent with the double barriers model introduced in Chapter 1 of this dissertation,
Patterson and Wilson (2000) argued that inequalities exist in two interfaces between
individuals and information technologies, the “upstream interface” and “downstream
interface” (p. 85). At the upstream, researchers ask how individuals with certain
demographic characteristics do or do not gain access to hardware and information
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
83
technology services. At the downstream, subsequent patterns of societal stratification
in education, health, wealth and income intersect with asymmetric access patterns in
the upstream interface to create further inequality in making use of the technology (p.
85). Therefore, the authors argued that even equal access to technology does not
resolve social equality.
Bonfadelli (2002), also argued that the “double digital divide” existed in
current Internet access and usage pattern in Europe (p. 65). From the knowledge gap
perspective, Bonfadelli argued that educational gaps in access to the Internet
widened between 1997 and 2000. In addition to access gaps, further gaps exist in the
Internet’s content-specific use. People with higher educational levels use the Internet
more actively and more for information-oriented purposes, while those with lower
educational levels mainly use the Internet for entertainment-related functions. This
pattern was also found in other studies (Jung et al., 2001; Shah, Kwak, & Holbert,
2001). Bonfadelli (2002) cautioned against the techno-centric view in
conceptualizing digital divide and urged for the need to study what people really do
when they go online. As he argued,
[RJesearch has to be based on a user perspective looking at technology as
embedded and shaped socially. Thus, inequality is not only a matter of
inadequate supply of content or not having access to Internet. Rather, it
means that inequality has to be conceptualized as structural and must be
studied with a focus on social processes of mediation, (pp. 81-82)
The author and her colleagues (Jung et al., 2001) argued that the digital
divide is not a single-faceted gap, but a multi-faceted divide reflected in different
aspects of people’s relationships with the Internet. They found that significant
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
84
variations exist among people of different income, education, age, gender and
ethnicity in ways they connect and use the Internet.
Studies discussed in this section approached the digital divide issue from a
more qualitative and contextual approach than did earlier studies.
Antecedents of Digital Divide
This section reviews past digital divide studies according to identified factors
that influence disparities in people’s Internet access and use.
Socioeconomic Status
Socioeconomic status is usually indicated by income and education. From
early digital divide studies to recent studies, socioeconomic status has been the most
commonly used indicator of the digital divide. The National Telecommunications
and Information Administration of the Department of Commerce has annually
reported socioeconomic disparities with regard to Americans’ computer and Internet
access and use since 1995. The Pew Internet and American Project, Rand
Corporation, The UCLA Center for Communication Policy (2001), and Stanford
Institute for the Quantitative Study (Nie & Erbring, 2000) have all mainly reported
socioeconomic disparities in access rate and time spent online. These reports agreed
that disparities in Internet access exist in terms of income and education.
Beyond access/non-access rates, Howard, Rainie, and Jones (2001),
according to their logistic regression analyses, found that higher education is
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
85
associated with higher e-mail use and higher Internet use for financial, political and
government information. People with more education are also more confident about
using online banking and performing other financial transactions online. The US
Department of Commerce’s 2002 report indicated that a higher income level is
associated with engaging in a wider range of activities on the Internet. The
proportion of Internet users in the highest income level (household income higher
than $75,000) exceeds other income groups in eight of the sixteen online activities
surveyed.
The author and her colleagues (2001) examined the effects of socioeconomic
variables on the “Internet connectedness index,” a multidimensional measure of
people’s Internet connections beyond access. They found that higher income, higher
education and greater youth are significantly associated with higher scores on the
Internet connectedness index. Hindman (2000), using national survey data from two
time periods, showed that the effects of income, education and age on the
“information technology index” were stronger in 1998 than in 1995, indicating
growing gaps over time in terms of these variables (p. 553). Results from General
Social Survey Demographics indicated that time spent on e-mail and other Internet
activities increases as income and education levels increase. When income and
education are analyzed together in multiple regression analyses, education turns out
to be a much stronger predictor of Internet use than income (General Social Survey,
2000).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
8 6
Age
In many recent digital divide studies, age has consistently been shown as the
strongest, or at least one of the strongest variables, for explaining disparities in
Internet access and use (Adler, 1996; Cody, Dunn, Hoppin, & Wendt, 1999; Lenhart,
2000; Loges & Jung, 2001; Nie & Erbring, 2000; UCLA Center for Communication
Policy, 2000; US Department of Commerce, 1999, 2000, 2002; White & Weatherall,
2000).
According to the Pew Internet report (Lenhart, 2000), just 13% of those over
65 have Internet access, compared to 65% of those under 30. More and more young
people are using the Internet, according to Department of Commerce’s 2002 report,
with teenagers having the highest Internet access rate among all age groups. People
over 65 show a more than 40 percentage point drop in their rates of Internet access
compared to those under 25 (Nie & Erbring, 2000). The Pew Internet Report
(Lenhart, 2000) also found that the economic factor does mediate the age difference,
but it does not override the difference. A 65-year-old living in a household with more
than $75,000 in income is three times less likely to be online than a 25-year-old at
that same economic level. At the other end of the economic ladder, a 25-year-old in a
household with a $25,000 income is twice as likely to be online as a 65-year-old at
that same economic level.
Within the same educational level, the percent of Internet connectors is
higher for people between 25 and 49 (82% of college graduates and 58% of high
school graduates), compared to those 50 or over (61.3% of college graduates, 39% of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
87
high school graduates). The US Department of Commerce reported that among those
aged 50 or older, 66% of college graduates and higher have Internet access, while
21% of high school graduates and less do. This wide gap is not as prominent in other
age groups (e.g. in ages 24-49, 88% of college graduates and 49% of those with a
high school education or less have Internet access).
Concerning motivation and attitude towards the Internet, people older than 65
reported much less agreement with questions regarding the Internet, such as “The
Internet would help me find things.” and, “I’m missing out by not being online.” than
did people younger than 30. Seniors were about half as likely as younger respondents
to agree with such statements (Lenhart, 2000).
The scope of activities engaged in online was narrower for elders than
younger age groups (Lenhart, 2000; Loges & Jung, 2001; US Department of
Commerce, 2002). Those 55 and older were least likely to use the Internet for many
categories, according to the Department of Commerce, such as online banking,
participating in chat rooms or list servs, viewing television or movies, trading online
and playing games. On the other hand, this group was more likely (43%) than other
age groups to check health information online and was as likely as younger age
group to use e-mail. Online shopping was particularly common among Internet users
in the 25-34 age group.
According to Loges and the author (Loges & Jung, 2001), who examined the
effect of age on multidimensional Internet connectedness, age has a significant
association not just with Internet access but also with a tendency to pursue a more
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
8 8
narrow range of personal goals online and with a pattern of connecting to the Internet
from a smaller range of places. Nonetheless, Loges and Jung found that older
respondents claim that the Internet is central to their lives just as younger people do.
The General Social Survey (2000) reported a similar result that once seniors (age 55
or over) are online, they report the same extent of usage hours on e-mail and other
activities as younger age groups.
Ethnicity
Compared to previous studies of “old” communication technologies, most
empirical research on the Internet included ethnicity as one of the main independent
factors in studying adoption and usage patterns (Novak & Hoffman, 1998; Rios &
Gaines Jr., 1998; Spooner, 2001; Spooner & Rainie, 2000; US Department of
Commerce, 1998, 1999, 2000, 2002). Some of these studies specifically tested and
confirmed a unique effect of ethnicity on individuals’ connection to computers or the
Internet. Department of Commerce reports (1998; 1999; 2000) showed that ethnicity
is a significant cultural factor in the digital divide after controlling for income and
education. Based on the same CPS data, an analysis by Bikson and Panis (1999)
confirmed the Department of Commerce result that ethnicity has a unique effect on
people’s Internet access beyond socioeconomic factors. In her dissertation, Babb
(1998) presented a finding that African Americans and Latinos are less likely to own
computers than Whites, after controlling for income and education. From extended
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
89
focus group data, Wilhelm (2000) also observed a significant unique effect of
ethnicity on the level of Internet connection.
Going beyond ethnic differences in Internet access, several studies examined
ethnic differences in how the Internet is incorporated into people’s everyday lives.
The Pew Internet Project’s report, Hispanics and the Internet, suggested that on-line
Hispanics are more likely than on-line Whites to search for information about books,
movies and other activities; more likely to download and listen to music; and more
likely to go online just for fun (Spooner & Rainie, 2000). They are as likely as
Whites to look for a job, get health and medical information, and get housing
information on-line. These results suggest that Hispanics, once online, are as broadly
and intensely connected to the Internet as Whites. On the other hand, Wilhelm
(2000) found that after gaining access to a technology, African Americans and
Hispanics have not been able to incorporate the technology in a way to move them
up the social ladder, especially because of the lack of Internet content that are
targeted at these minority groups. He stated, “If minorities are perceived as
marginalized actors in the dominant society’s euphoria over information technology,
then they too will not see the value in cultivating certain skills and training” (p. 62).
Howard and his colleagues (2001) found that significant differences exist in
the kinds of online activities different ethnic groups engage in. Compared to White
respondents, Asian Americans are more likely to engage in stocks, bonds, and
mutual fund transactions, and to make travel arrangements on the Internet. On the
other hand, they are less likely to research hobbies online than their White
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
90
counterparts. African Americans are much less likely to have used e-mail than other
ethnic groups, but they are more likely to have done fun things online, such as
playing games and checking sports scores. They are also more likely to have looked
for religious or spiritual content online.
In contrast to the studies showing persisting ethnic effects on the digital
divide, other studies reported the trend of “closing ethnic divides.” Unlike their own
previous reports, the Department of Commerce’s 2002 report titled “A Nation
Online: How Americans Are Expanding Their Use of the Internet” suggested that the
ethnic divide is closing. The report emphasized that African American and Latino
populations have shown much higher increases in adoption rates than White and
Asian American populations. A Stanford University study (Nie & Erbring, 2000)
reported that ethnic differences vanish after controlling for income and education.
Along these same lines, Novak and Hoffman (1998) found income and education
explain a significant portion of ethnic difference. The gap between Whites and
African Americans with regard to PC ownership is explained by income, while the
gap between these two ethnicities in terms of work-related computer access is
explained by education.
Although ethnicity was included in many studies as an important factor
affecting the digital divide, not many studies provided an in-depth focus on what the
ethnic divide means. For example, why Latinos, when compared with Whites in the
same income and education group, have a lower Internet access rate has not been
adequately investigated. Also, ethnicity was addressed in a crude category, either
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
91
White/Black or White/Black/Latino and, at most, White/Black/Latino/Asian
Americans. It is difficult to find a common factor that affects disparities in Internet
access when quite different ethnic groups are categorized together, for example,
various Asian Americans. A study that this researcher participated in with other
colleagues in the Metamorphosis Project examined the effects of “geo-ethnicity,” an
ethnic group living in a certain geographic area, on Internet connectedness (Kim,
Jung, Ball-Rokeach, & Loges, 2002). We assessed seven geo-ethnic samples in
seven different Los Angeles area neighborhoods, and found that a unique
communication environment surrounding each geo-ethnic group in different
neighborhoods shapes the breadth and depth of people’s Internet incorporation in
their everyday lives. All the surveys, focus groups and interviews in the study were
conducted in the respondents’ preferred language, which differentiates this study
from a majority of other ethnicity studies which only use English to interview
respondents. This approach, still being developed in the Metamorphosis Project (e.g.,
Kim & Jung, 2002), allows researchers to examine in-depth what it means to be
Korean-American, for example, living in Koreatown in Los Angeles, and how it
differs from Chinese-Americans living in Monterey Park, California, rather than
merging them, with other Koreans and Chinese living in other areas, all together
under the category of Asian-Americans.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
92
Social Environment
Scholars have conceptualized the value of the social environment in people’s
lives as “social capital,” a resource embedded in our social ties that can be
transformed to other types of resources such as human capital or economic capital
(Burt, 1992; Coleman, 1998; Loury, 1997; Putnam, 1995, 2000). According to
Bourdieu (1986), social capitals are resources or profits to which individuals have
access as a result of their membership or participation in groups such as families,
parties or associations. Burt explained that these resources do not flow equally in
each set of human relations and therefore individuals in a position to capitalize on
opportunities are those who have access to personal contacts that can provide
valuable information, knowledge or expertise.
Several studies highlighted the importance of the social environment in
introducing and distributing new technologies in society. For example, studies from a
diffusion of innovation perspective, as discussed earlier, differentiate public
communication and interpersonal communication where the former leads to the
awareness of a technology and the latter plays a critical role in the process of making
decisions on whether or not to adopt a technology (Hobday, 1994; Rogers, 1983;
Williams, Rice, & Rogers, 1986; Williams, Strover, & Grant, 1994). Research on the
adoption of technologies in organizational settings also found that the interpersonal
network in organizations is one of the most important factors in shaping the way a
technology is adopted and used (Bouwman & Van de Wijngaert, 2002; Fulk et al.,
1990; Schmitz & Fulk, 1991).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
93
A group of European researchers examined new media in the context of
social environment and, in particular, family environment (Haddon, 2000;
Livingstone, 1998; Livingstone & Bovill, 2001; Murdock, Hartmann, & Gray, 1992;
Silverstone & Hirsch, 1992; Suoninen, 2001). A cross-national project led by Sonia
Livingstone (1998) examined the importance of “contextualizing new media in
relation to both pre-existing media practices and the broader contexts of young
people’s lives” (p. 435). Suoninen (2001) conducted a qualitative study about
children and teenagers’ media use in the context of peer group relations. She
observed situations where children use media together with their friends, going to
friends to use media that are not available at their homes, talking with friends about
media, swapping media with friends and perceiving media as a status symbol.
Among the few studies that have examined the influence of social
environment on Internet access and use,1 1 Matei and Ball-Rokeach (2001) found that
those who have a larger social network in offline interpersonal relations tend to make
more friends online. Based on a social network perspective, Bimie and Horvath
(2002) found that frequency of contact with family and friends has positive effects
on the frequency of Internet contact. Shah, Kwak, and Holbert (2001) examined the
relationship between the Internet and social capital. They found that Internet use,
categorized by social recreation, product consumption, financial management and
information exchange, is significantly associated with social capital variables, such
as civic engagement, interpersonal trust and life contentment, when demographic
variables and other forms of media use are held constant.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
94
Wilhelm (2000) argued from what he calls a “quasi-sociological perspective”
that the ways in which one’s family members, friends, neighbors or other social
contacts perceive the value of the Internet and the computer affect the person’s
adoption and use of the Internet (p. 65). Fong and his colleagues examined the effect
of three different social characteristics (individual, household and neighborhood) on
people’s Internet use in Canada and the United States (Fong, Wellman, Kew, &
Wilkes, 2001). They found that, in addition to the individual’s socioeconomic status,
household characteristics such as household size, the presence of children and
neighborhood characteristics such as the proportion of immigrants and rural-urban-
suburban locations are important factors affecting people’s Internet access. A recent
report by a Silicon Valley-based organization found that despite the overall high
Internet access rate among youth across different social groups, the ways in which
students perceive the Internet’s importance to their future careers are highly
influenced by whether or not they are located in a social environment where it is
relatively easy to get Internet-related support whenever needed from people around
them (Joint Venture, 2002).
Kiesler, Zdaniuk, Lundmark and Kraut (2000) examined the process in which
a family member becomes the “family guru” in terms of computers and Internet use.
The authors found that those with the most computer skills, confidence or
enthusiasm became the most involved Internet users and this involvement led to their
making requests to external support sources and solving problems, leading to more
Internet usage. Highly involved Internet users often became the family guru upon
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
95
which other family members relied for help when having computer and Internet
problems (p. 345). Kiesler et al.’s study emphasized the importance of immediate
family members in the process of adopting and using computers and the Internet.
Similarly, a recent Pew Internet and American Life Project report showed that 70%
of US parents have Internet access, compared to 53% of non-parents (Allen & Rainie,
2002). Also stated in the Pew report is that parents are generally more enthusiastic
about a new technology and are more open to technological changes than non
parents (Allen & Rainie, 2002). The evidence indicated that immediate family
environment, i.e., having a child or children at home, affects the ways in which the
adults perceive and use the Internet.
Communication Environment: New and Old Media
In light of the vast number of published studies on communication media, not
many studies examined the interaction between different types of media. Most
studies addressed one communication medium. Among the studies that did look at
the interaction between old and new media, the most common topic was examining
the relationship between television viewing and newspaper readership, mostly
reporting the decline of readership since the advent of television (e.g., Belson, 1961;
Peiser, 1999; Van Eijck & Van Rees, 2000). However, results of many studies
indicated that television’s effects on newspaper readership were minimal (Belson,
1961; Peiser, 1999).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
96
Recently, several studies examined the interactions between computer
technology and older communication media. Examining new and old media in
households, Johnsson-Smaragdi, d’Haenens, Krotz and Hasebrink (1998) found that
television still occupies a dominant position in the lives of children and teenagers. At
the same time, the authors found that young people spend more time on PCs or PC
games than on print media. The UCLA Center for Communication Policy (2003)
reported that Internet users watched television 5.4 hours less per week than non-users
in the year 2002. In addition, television viewing tended to decline as Internet
experience increased; more than twice as many of the very experienced users than
new users said that they spend less time watching television since using the Internet.
With regard to goals for using different media, Johnsson-Smargiagdi et al.
(1998) found that young people watch television mostly to fulfill their entertainment
goals, while they use PCs for a wider variety of reasons, including school work,
communication and entertainment. Similarly, Beentjes, Koolstra, Marseille and van
der Voort (2001) found that young people in Europe spend the most time with
television, followed by audio media and electronic games, while book and
newspaper reading ranked lower. These studies and others (d'Haenens, 2001; Kayany
& Yelsma, 2000; Perse & Dunn, 1998) that looked at new and old media in
households, however, mostly stopped at description of results without establishing a
framework for examining the interaction patterns between new and old media.
One study (Atkin, 2001) tried to establish an ecological framework for
examining children’s television viewing in the new media environment. Atkin (2001)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
97
applied Bronfenbrenner (1990)’s model of ecology in contextualizing the new media
environment of children. Five components of the model are mesosystem, exosystem,
microsystem, macrosystem and ontological system, respectively referring to family
uses of new media; media environment and usage pattern; parental mediation in the
new media environment; social status influences on media adoption and use; and
children’s psychological competencies for development. Although the model was not
applied in an empirical study, Atkin (2001) provided an ecological and multileveled
picture of new and old media in current social and communication environments.
Summary
This chapter reviewed past studies on communication media and inequality,
focusing on recent Internet and inequality studies in terms of socioeconomic status,
age, ethnicity, social environment and media environment. Most studies introduced
in this chapter approached the issue of the Internet and inequalities by focusing on
only one aspect of these different social factors that affect the digital divide. Few
took an ecological approach to examine different social factors within a theoretical
or empirical framework. That is, the ways in which different factors, such as
demographics, social environment and media environment, interact with one another
in affecting individuals’ Internet connections were not adequately explored.
The next chapter introduces research questions and hypotheses, based on
media system dependency theory, communication infrastructure theory and past
studies, that apply an ecological approach in examining the influence of different
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
98
factors on people’s media connections, Internet access and multidimensional
connectedness to the Internet.
1 1 On the other hand, many recent studies emphasized the Internet’s central role as a
facilitator of social relationships (Baym, Zhang, & Lin, 2001; Flanagin & Metzger, 2001; Lee & Chan,
2001; Lee & Kuo, 2002; D.V. Shah et al., 2001; Wellman, Hasse, Witte, & Hampton, 2001). These
studies found that the Internet plays a positive role in maintaining and extending the existing social
relationships. For example, Lee and Chan (2001) found that the Internet facilitates adolescents’
communication with friends, the majority of them arranging face-to-face social gathering via e-mails.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
99
CHAPTER 4. HYPOTHESIS AND METHODOLOGY
Hypothesis
Based on the theoretical framework of media system dependency theory
and communication infrastructure theory and building on past Internet and
inequality studies discussed in Chapter 3, the researcher developed the following
research questions and hypotheses.
Scope o f Media and Specific Media Connections
In media system dependency theory, communication media have been
conceptualized as social institutions that control creation, processing and
dissemination of resources that individuals need to solve their everyday goals. The
scope of media available to an individual when he or she needs to connect to a
resource is largely influenced by his or her structural location (Ball-Rokeach, 1985,
1998). That is, the person’s social groups (socioeconomic and demographic
indicators) are likely to affect which media forms to access, and the breadth of his
or her accesses. Also, traditional audience research identified different media
consumption patterns as a function of socioeconomic and demographic factors
(Alasuutari, 1999; Ball-Rokeach & Cantor, 1986; Graham, 1999; Lauf, 2001;
McQuail, 1997; Mullan, 1997; Neuman, 1991; Petrie & Willis, 1995; Webster,
1986). Under a changing media environment, this study examines the associations
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
100
between socioeconomic/demographic variables and the breadth and depth of
individuals’ connections to communication media.
RQ1 What kinds of disparities exist among different socioeconomic and
demographic groups in the scope o f communication media that people
connect to for achieving everyday goals? Does a certain group of people
have a broader scope of media that they connect to than others?
RQ2. What kinds of disparities exist among different socioeconomic and
demographic groups in connections to specific communication media,
including television, newspapers, interpersonal communication and the
Internet? For example, which socioeconomic group has higher connections
to television than other groups?
Internet Access
Based on previous studies of the unequal access to the Internet, discussed in
Chapter 3, the researcher proposes hypotheses 1 and 2, and research question 3.
H I: Socioeconomic status (income and education) and demographic
characteristics (age, gender and ethnicity) affect Internet access in the
direction that higher income, higher education, younger, male, and Whites
are more likely to have Internet access than their counterparts.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
101
H2: The scope of people’s connection to mass media forms has a positive
effect on Internet access, controlling for socioeconomic status and
demographic characteristics. That is, the greater the number of media a
person has connections to, the more likely he or she has access to the
Internet.
RQ3: What relationships do people’s connections to traditional
communication media (television, newspapers and interpersonal
communication) have to their Internet access, when socioeconomic status
and demographic characteristics are held constant?
Table 1. Antecedents of Internet Access: Hypotheses and Research Questions
HI H2 RQ3
SES
Age
Gender
Ethnicity
Media
Connectedness
Scope
Television
Connection
Newspaper
Connection
Interpersonal
Communication
Connection
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
102
Internet Connectedness
Three hypotheses and one research question are proposed when Internet
connectedness is examined as a dependent variable. First, an individual’s structural
location (Ball-Rokeach, 1985,1998) is an important micro-level antecedent that
affects connectedness. Structural location is an individual’s position in a society in
relation to other people, as well as his or her social, economic and cultural
orientations. As reviewed in Chapter 3, a person’s socioeconomic and demographic
orientations play important roles in shaping his or her connectedness relations.
Educational background, income, occupation, age and gender are important
indicators of a person’s structural location in society. In addition, an important
variable that characterizes a person’s structural location, especially in US
metropolitan cities, is a person’s ethnic background (Ball-Rokeach et al., 2001;
Kim et al., 2002). Ethnicity involves the intermingled factors of language, cultural
networks and cultural interests that are crucial when studying people’s
communication connectedness.
H3: Socioeconomic status (income and education) and demographic
characteristics (age, gender and ethnicity) affect Internet connectedness
such that persons with higher income, higher education, who are younger,
male, and White are more likely to have higher scores in Internet
connectedness than their counterparts.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
103
A person’s structural location is also characterized by family, friends, co
workers and other people around the person (Fulk et al., 1990; Jung, Kim, Lin, &
Cheong, 2002; Putnam, 2000; Schmitz & Fulk, 1991; Wilhelm, 2000). An
individual’s interpersonal networks are likely to affect the person’s Internet
connectedness by shaping communication goals, providing support in forming a
connectedness relation to the Internet and delivering social pressure to use the
medium (See Chapter 3).
Another contextual factor likely to affect individuals’ Internet
connectedness relations is the technological environment present in a person’s
communication action context. The technological environment includes having a
household computer, length of Internet experience and the number of places the
person accesses the Internet. What kinds of communication media a person owns or
subscribes to at home is an important environmental indicator of connectedness.
For example, a person who has access to the Internet at home, work and the library
is likely to have different connectedness relations to the Internet from another
person who only has Internet access at work. In addition, how long a person has
used certain media is an important indicator for characterizing connectedness
relations (Hargittai, 2002; Horrigan & Rainie, 2002).
H4: Social environment (degree of interactions with others in receiving or
providing Internet-related help, proportion of family members who have
Internet access and proportion of friends who have Internet access) and
technological environment (having a computer at home, number of places
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
104
for available Internet access and number of years for using the Internet)
have positive effects on Internet connectedness, controlling for
socioeconomic status and demographic characteristics.
In addition to socioeconomic status, demographic characteristics and social
and technological environments, an individual’s general media connections are
likely to shape his or her Internet connectedness. This relationship is especially
important when studying new media. An individual’s connectedness to a new
medium is likely to be shaped largely by the connectedness relations the person has
had with existing media. For example, one can compare two people who recently
adopted the Internet. Person A, who had intense and broad connectedness relations
with television, radio and newspapers, would incorporate the Internet into his or her
life differently from person B, who mostly had an intense connectedness relation
with television. The kind of communication resources available to an individual is
likely to influence his/her choice of a new medium to achieve goals. In other words,
an individual’s connectedness to the Internet is always in the context of the
person’s connections to other media.
RQ4: What kind of relationships do people’s connections to communication
media, including television, newspapers and interpersonal communication,
have with their Internet connectedness, when socioeconomic status,
demographic characteristics, technological environment and social
environment are held constant?
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
105
Finally, individuals’ motivations affect their connectedness relations with
the Internet. Individual motivations, as described in MSD theory and also in the
connectedness concept, are mainly characterized by goals (Ball-Rokeach, 1985,
1998; Jung et al., 2001; Loges, 1992). As explained in media system dependency
theory, goals of understanding, orientation and play are the most essential
motivations in people’s everyday lives. Understanding is the goal of trying to find
out what is going on in society. The orientation goal is to acquire information to
behave, i.e., to engage in an activity. A play goal is to entertain oneself, either by
solitary play such as watching television or playing games, or by social play such
as going to a movie with a friend or chatting online. The importance of certain
goals versus others and the breadth of goals that people have are likely to shape
Internet connectedness.
H5: Internet goal scope and intensity have positive effects on Internet
connectedness, when socioeconomic status, demographic characteristics,
technological environment, social environment and media connections are
held constant.
Table 2 includes hierarchical regression models involving hypotheses 3-5
and research question 4.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
106
Table 2. Antecedents of Internet Connectedness: Hypotheses and a Research
Question
H3 H4 RQ4 H5 1
SES
Age
Gender
Ethnicity
■
Social and
technological
environments
B
Media
connectedness
BH
B
Internet goal
scope and
intensity
B
Methodology
Data
This study was based on a telephone survey conducted in the Los Angeles
area as part of ongoing research by the Metamorphosis Project at the Annenberg
School for Communication at the University of Southern California. Since 1997,
the Metamorphosis Project has been examining the transformation of urban
communities under the forces of globalization, new communication technologies
and population diversity ( 'http://www.metamorph.org'). Dr. Sandra Ball-Rokeach is
the principal investigator of the project. The researcher played central roles in the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
107
research design, data gathering and data analyses as a research assistant for four
years.
The Metamorphosis Project employed an in-depth and multi-level research
design to examine the communication infrastructure in a prototypical 21st century
urban area, Los Angeles. Eight ethnically diverse residential areas located within
10 miles of the Los Angeles Civic Center were studied between 1998 and 2002. In
the first seven areas, one ethnicity that shaped the tone and character of the area
was selected to draw a sample. The study areas and target ethnicities included
White/plurality Protestants in South Pasadena, White/plurality Jewish on the
Westside, African-American in Greater Crenshaw, Korean origin in Koreatown,
Chinese origins (Mainland, Taiwan and Hong Kong) in Monterey Park, Mexican
origin in East LA and Central American origins in Pico Union.
Glendale, which is an ethnically diverse area located 8 miles from the Los
Angeles Civic Center with multiple ethnicities living there, was the eighth area
studied to better understand cross-cultural/ethnic differences. The major Glendale
ethnicities were Armenian origin, White and Mexican origin (U.S. Census Bureau,
2000). The analyses in this dissertation were based on 585 Glendale respondents
selected by random digit dialing (RDD).U 1 Consistent with the previous
Metamorphosis Project strategy of administering surveys in the respondents’
preferred languages, surveys in Glendale were conducted in multiple languages:
English, Spanish and Armenian. A reputable survey research firm, California
Survey Research Services (CSRS), worked with the research team to conduct the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
108
survey interviews. The survey was programmed for Computer Assisted Telephone
Interview (CATI) administration. The CSRS supervisory staff trained bilingual
interviewers to administer the survey.
The forty-five minute telephone survey included questions asking respondents
about belonging to their community and their interactions with their neighbors, a
variety of questions about people’s media-related behaviors, social contacts, travel
patterns and socio-demographic characters. The survey response rate in Glendale
was 54% when calculated by dividing the number of completed interviews by the
number of theoretically eligible phone numbers. Eligible phone numbers were
calculated by examining the total number of study phone numbers excluding phone
numbers for which eligibility could not be determined, inappropriate/duplicate
phone numbers, non-qualified household phone numbers (e.g., outside the study
area) and the estimated number of initial refusals not likely to qualify for the study.
Forty-two percent of Glendale respondents were males. Median age was 40.
White respondents constituted 33% of the total sample. 27% were Latino and 36%
were Armenian. Forty-two percent of the total respondents had college graduate
degrees or higher educational achievement. Median income was between $35,000
and $45,000.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
109
Variables
Internet Connectedness Index
The researcher and her colleagues in the Metamorphosis Project have
developed the Internet Connectedness Index (ICI) since 1999 (Jung et al., 2002;
Jung et al., 2001; Loges & Jung, 2001). We have examined large disparities in
terms of having or not having Internet access in different areas that we have studied
but, at the same time, as more and more people have gone online, we saw the need
for an indicator to measure disparities “beyond access” in the ways in which people
incorporate the Internet into their everyday lives.
The first ICI was published in August, 2001 (Jung et al., 2001) with nine items:
• Number of years of home computer ownership
• Number of tasks accomplished through Internet connections
• Number of Internet access sites
• Number of goals served by online connections
• Number of activities participated in while online
• Frequency of participation in online interactive activities (e.g., chat rooms)
• Evaluation of the overall effect of the Internet on personal life
• How much the person would miss his/her computer if it vanished one day
• How much the person would miss the Internet if it vanished one day
In order to produce compatible scale items, each variable was multiplied by
a value to create a common factor of 12. For example, home computer history, a 4-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
110
point scale, was multiplied by 3, while goal scope, a binary scale, was multiplied
by 6. ICI scores were then calculated by taking an overall average, with ICI scores
ranging from 1 to 12. The reliability alpha was .71 for this nine-item scale.
In Jung et al.’s study (2001), ICI and time spent online measures were each
correlated with socioeconomic and demographic variables. The result indicated that
the ICI has linear relationships with socioeconomic and demographic variables,
while the time spent online measure did not show any linear relationship with the
variables. That is, those who were in higher income and education group, younger,
and male were more likely to have higher scores on the ICI (Jung et al., 2001).
Despite the ICI’s innovative and multidimensional nature as a theoretically
grounded measure, it had several limitations. Theoretically, the “connectedness”
concept had not been fully established when the first version of ICI was developed
to be able to identify what constitutes Internet connectedness and what constitutes
antecedent variables influencing connectedness. That is, a clear rationale for
including different items in the index and the coherence among the items were
lacking. For example, after the researcher conducted a more comprehensive review
of past literature and the examination of the development of communication
infrastructure theory from media system dependency theory, number o f years o f
home computer ownership and number o f Internet access sites items were
considered more appropriate to be included in the technological environment that
shaped Internet connectedness, rather than connectedness itself. Also, the number
o f goals served by Internet connections item was more appropriate as an immediate
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
I l l
antecedent variable for Internet connectedness according to media system
dependency theory (See Chapter 2).
In terms of a methodological limitation, nine items in the ICI were added
together without any sub-factors. Although the ICI is a significant step forward in
accounting for the multidimensional nature of Internet connectedness, the ways in
which different items in the index were clustered were not further explored. In
addition, the index was constructed by manually standardizing and adding the items
without deriving factors from them.
Other methodological limitations were found in the ways questions were
asked. In the process of revising the survey for Glendale, the author and her
colleagues made several modifications in response to the limitations. For example,
questions about Internet-related goals were modified from one question, allowing a
respondent to choose multiple answers (“What are the most common reasons why
you participate in these online activities? Is it...”) to multiple questions and asking
degrees of importance of each goal (“On a scale from one to five, where one is ‘not
important at all’ and five is ‘very important,’ how important is the Internet for
you...”). This modification was made to detect more thoroughly the differences in
the intensity and scope of Internet-related goals that individuals have. The Internet
activities question was also modified in the same way from “What Internet or web
activities do you participate in, other than e-mail?” to “Now I’m going to ask you
about some different things you might do online. Have you ever used the Internet
for.. and then listed 11 different Internet activities.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
112
In response to limitations mentioned above and reflecting the revised format
of questions, the author modified the Internet Connectedness Index in this
dissertation. The new Internet Connectedness Index composed of five items: scope
o f Internet activities; intensity o f Internet activities; time spent on the Internet;
computer miss; and Internet miss. The scope of Internet activities and the intensity
of Internet activities were based on a set of questions of “Now I’m going to ask you
about some different things you might do online. Have you ever used the Internet
for....” The items provided were chatting/IRC/ICQ or instant messenger; e-mail;
playing games; listening to music; participating in mailing lists of listserves;
maintaining a personal website; making Internet phone calls; participating in
MUDs/MOOs; reading messages from Newsgroups/Usenet/Bulletin Board (BBS);
surfing the web; and watching a webcast. The items were purposefully chosen to
reflect different skills that are required to participate in the activities. As soon as a
respondent answered “yes” to an item, a follow-up question was asked. “On a scale
from one to ten, where one is ‘not important at all,’ and ten is ‘very important,’
how important is this activity to you?” This question was asked for all the items
that respondents answered “yes” in order to assess the intensity of their engagement
in the activity. Number of affirmative answers to the eleven activities was counted
to form the scope o f Internet activities variable ranging from zero to eleven. The
intensity o f Internet activities variable was formed by averaging a respondent’s
score on the importance of activities scale ranging from one to ten.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
113
Time spent on the Internet was based on a question of “Including email,
approximately how many hours did you spend on the Internet last week?”
Respondents answered in an open-end format. The computer miss item was derived
from the question, “Imagine you woke up tomorrow to find that you could no
longer use any computer. On a scale from ‘1’ to ‘10’ where ‘1’ means you
wouldn’t miss it at all because your daily life could proceed as normal and ‘10’
means you would miss it an extreme amount, how much would you miss being able
to use a computer?” The Internet miss item was asked, “Imagine that you woke up
tomorrow to find that you could no longer use the Internet. Using a 10-point scale
where ‘ 1 ’ means you wouldn’t miss it at all because your daily life could proceed
as normal and ‘10’ means you would miss it an extreme amount, how much would
you miss being able to go online?”
A factor analysis was conducted to examine whether the five variables
considered for the Internet Connectedness Index loaded together (Table 3). All five
variables loaded together with high factor loadings (scope of Internet activities
= .81, intensity of Internet activities = .79, time spent online = .89, computer miss
= .82, Internet miss = .81). The reliability of the Connectedness Factor measured by
Cronbach alpha was .79. In addition, another factor analysis was conducted to
examine whether the five variables loaded into three factors (scope, intensity, and
centrality dimensions of Communication Connectedness conceptualized in Chapter
2). The author set the number of factors as 3 in the command, and ran the factor
analysis with a Varimax rotation. Three factors were derived that were consistent
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
114
with the theoretically proposed dimensions; scope o f Internet connectedness (scope
of Internet activities = .85); intensity o f Internet connectedness (intensity of Internet
activities = .58, time spent online = .87); and centrality o f Internet connectedness
(computer miss = .88, Internet miss = .80).
Table 3. Factor Loadings of Items in Internet Connectedness Index and its Sub-
Dimensions
Variables
Internet
Connectedness
Index
Scope
Dimension
Intensity
Dimension
Centrality
Dimension
Scope of Internet
activities
.81 .85
Intensity of
Internet activities
.79 .58
Time spent on the
Internet
.89 .87
Computer miss .82 .88
Internet miss .81 .80
Internet Goal Scope and Intensity
Internet goal scope and intensity was based on the questions “On a scale
from one to five, where one is ‘not important at all’ and five is ‘very important,’
how important is the Internet for you....” The items provided were “to stay on top
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
115
of events that you care about,” “to get in touch with people when you’re looking
for a job,” “to express yourself,” “to get your work done,” “to play or amuse
yourself,” “to stay on top of what is happening in your community,” “to gain skills
for career development,” “for social reasons like making friends” and “to get
advice on how to deal with other people such as doctors and health professionals.”
A respondent’s answers to the eight items (ranging from 1 to 5) were added and
averaged to form the total intensity scale ranging from 1 to 5. For the scope
variable, answers of a respondent for the same eight items were dichotomized into
“0” for not important at all and “1” for the rest of the categories in the scale. Then,
all scores for the eight items were added to form a scope variable ranging from 0 to
8. These two variables—intensity and scope—were included in a factor analysis to
derive a goal scope and intensity factor (factor loading: intensity=.946,
scope=.946).
Social Environment
Social environment was measured by four questions. First, two questions
were asked regarding people’s interactions with others in receiving and providing
Internet-related help. First question was, “When you are trying to figure out how to
do something new on the Internet or on your computer, how easy is it for you to get
help? On a scale from 1 to 10, where 1 is “very difficult” and 10 is “very easy,”
how easy is it to get help?” The other question was, “How likely is it that others
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
116
would approach you for help solving Internet or computer related problems? On a
scale from 1 to 10, where 1 is “very unlikely” and 10 is “very likely,” how likely is
it?” These two questions were included in a factor analysis to derive the
interactions with others in receiving and providing Internet-related help factor
(factor loading: receiving help = .76, providing help =.76).
Proportions of family and friends who were connected to the Internet were
asked for family and friends separately by two questions: “Thinking about your
family members whether they live with you or not, how many of them use the
Internet?” and “Thinking about your friends, whether they live in Glendale or
somewhere else, how many of them use the Internet?” The response set given was
“all of them,” “most of them,” “some of them,” and “none of them.” The responses
were reverse coded to have none of them as “1” and all of them as “4.” Running a
factor analysis with the two questions derived a proportion offamily and friends
connected to the Internet factor (Factor loading: family = .80, friends =.80).
Technological Environment
Technological environment was measured by three questions. The home
computer ownership question asked, “Do you have any type of personal computer,
including laptops in your home?” Second, the number ofplaces for Internet access
indicated where a respondent went online. Categories included home, school, work,
a community center or organization, a public library, a cybercafe or Kinkos, and
others. Multiple answers were accepted. Categories were trichotomized into home,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
117
school/work and others. A respondent’s affirmative answers to the three categories
were added to form a variable ranging from 0 to 3. Third, years o f Internet use was
based on the question, “How many years have you used the Internet?” Respondents
gave open-ended responses. The three variables—home computer ownership,
number of places for Internet access and years of Internet use—were put together in
a factor analysis. The three items loaded well into a factor (factor loadings: home
computer ownership=.70, number of places for Internet access=.78 and years of
Internet use=.68), and a technological environment factor was created.
Media Scope
Media scope was measured based on the media system dependency
questions in the survey.1 V Four similar questions were asked regarding the
respondents’ preferred media when trying to achieve four different goals in their
everyday lives. Questions read, “What are the most important ways for you to stay
on top of what’s happening in your community?”, “What are the most important
ways you get information to make decisions about the products you buy?”, “What
are the ways that you usually relax and have fun?” and “What are the most
important ways that you get medical information for yourself or for your family?”
Media forms given as a response set included interpersonal communication,
television, radio, newspapers, Internet, books, magazines, movies, etc. Multiple
answers were accepted. Affirmative answers to media forms were added for each
goal. The four variables (number of affirmative answers to each of four goals) were
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
118
included in a factor analysis to derive a media scope factor (factor loading:
community issues= 73, product buys-.81, having fun=.75, and health
information=.75).
Specific Media Connectedness
Media connectedness to three different communication forms (television,
newspapers and interpersonal communication) were calculated based on the media
system dependency questions described in the above media scope section. The
television connectedness variable was derived by adding the number of affirmative
answers to television for each of four goals. For example, if a respondent chose
television for the community issues goal and having fun goal, the person’s
television connectedness score was 2. The same procedure was applied to
newspapers connectedness and Interpersonal connectedness variables. These
variables ranged from 0 to 3.
Socioeconomic Status
Socioeconomic status was measured by income and education. Income data
were obtained by asking for the household income for the previous year. The seven
response categories were as follows: less than $20,000, $20,000-$35,000, $35,000-
$45,000, $45,000-$60,000, $60,000-$75,000, $75,000-$100,000 and more than
$100,000. The highest grade or year of school that the respondents completed was
used to indicate educational level. Seven response categories were provided: eighth
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
119
grade or less, some high school, high school graduate, some college or technical
school, college graduate, some graduate study and graduate degree. A
socioeconomic status factor was derived by running a factor analysis with income
and education where the two variables both had high factor loadings (income=.82,
educations 82).
Demographic Characteristics
Demographic characteristics were indicated by age, gender and ethnicity.
The respondent’s age on his or her last birthday was asked in an open-ended format.
Male gender was coded 1, and female was coded 0. Ethnicity was asked, “How do
you usually describe your ethnicity?” Given the large number of Armenians and
Latinos in the area that the data were derived, Armenian and Latino were included
in the regression models as dummy variables.
III The well-known RDD sample advantages of bias reduction are particularly evident in the
Los Angeles area. It is estimated that 50 percent of Los Angeles household phone numbers are
unlisted. Thus, the financial and procedural advantages of employing phone (or other) directories,
while considerable, are outweighed by the advantages of the RDD procedure. This is particularly the
case when the research design objective is to gain access to representative samples of geographically
located area residents (Matei, Ball-Rokeach, Wilson, Gibbs, & Gutierrez Hoyt, 2001).
I V An individual-level MSD scale was first introduced in the seminal research, the Great
American Value Test (Ball-Rokeach et al., 1984), a field experimental study where the researchers
arranged a real television program broadcast called The Great American Value Test, and assessed
viewers’ values, attitudes and behavioral changes. Media dependency was tested by constructing an
1 l-item scale, which was designed to measure each of the conceptual categories of individuals’
media system dependency goals; understanding (social and self), orientation (action and interaction),
and play (social and solitary). Questions were asked regarding frequency of media use: “How often
do you use television to... (achieve personal goals)?” The response categories were Often,
Sometimes or Never.
Following these studies in 1980s, subsequent efforts were made in early 1990s by a group
of researchers at the Annenberg School for Communication at the University of Southern California
to improve the MSD scale. Two major efforts were made. First, question/response format was
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
120
changed. The “frequency of use” was not considered consistent with the theoretical assumption of
the MSD theory. As Ball-Rokeach, Grant, and Horvath (1993) emphasized, the MSD concept
“implicitly suggests that it is the qualitative nature of media relations that we must capture. Quantity
measure should.. .be less sensitive indicators because they tell us little about the goals that motivate
individuals to establish media relations” (p.8). Therefore, questions in the MSD scale were changed
to “In your daily life, how helpful is television (or other media) to ... (achieve specific goals)?” The
response format was changed to a five-point likert-scale, ranging from “not at all helpful” to
“extremely helpful.”
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
121
CHAPTER 5. RESULTS
Frequencies
Before going into hypotheses testing, descriptive frequenci es of variables
included in the analyses are shown in Table 4 to 10 below. (Total N = 585)
Table 4. Median Number of Media Chosen for Each Goal
To stay on top of what is happening in the community
To get information on products to buy
To have fun or relax
To get medical information
M: Median, SD: Standard Deviation
Table 5. Percent of Respondents who Chose Specific Media for Each Goal
To stay on top
of what is
happening in
the
community
To get
information
on products to
buy
To have fun
or relax
To get
medical/health
information
Interpersonal
Communication
25% 34% 33% 20%
Television 55% 32% 33% 16%
Newspapers 40% 19% 3% 8%
Radio 9% 4% 7% 2%
Internet 14% 24% 7% 19%
CD or Cassettes 7%
M = 1, SD = .85
M = 1, SD = .85
M = 1, SD =1.12
M = 1, SD = .85
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
122
Table 6. Internet Access, Home Computer Ownership, Time Spent Online, Places for
Going Online and Average Number of Places for Going Online
Internet access 67%
Home computer ownership 72%
Median number of years for using the
Internet
Median = 4,
SD = 2.87
Mean Time Spent on the Internet a Week Mean = 12 hours,
SD = 47.68
Places for going online
Home 57%
Work 35%
School 14%
Public library 7%
Community center or organization 2%
Cyber cafe 2%
Median number of places for going online Median = 2,
SD = .99
SD: Standard Deviation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
123
Table 7. Internet-Related Help Source*
Family 26%
Friends 22%
Technical support 11 %
Online help 9%
Co-workers 7%
Customer service 4%
Books and references 2%
*Multiple answers were accepted. The median number of help sources are M = 1
(Standard Deviation = .67)
Table 8. Proportions of Family and Friends Using the Internet
Family members using the Internet Friends using the Internet
All of them 24% 30%
Most of them 28% 35%
Some of them 36% 26%
None of them 8% 3%
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
124
Table 9. Mean Degree of Importance of Goals for Going Online
Staying on top of events that you care about M = 3.58, SD= 1.40
Getting work done M = 3.53, SD = 1.65
Getting in touch with people when looking for a job M = 3.16, SD = 1.74
Gaining skills for career development M = 2.89, SD = 1.60
Expressing yourself M = 2.48, SD = 1.57
Playing or amusing yourself M = 2.46, SD= 1.45
Getting advice on how to deal with other people M = 2.43, SD= 1.46
Staying on top of what is happening in the community M = 2.30, SD = 1.45
Social reasons like making friends M = 1.70, SD = 1.14
Range: 1-5
M: Median, SD: Standard Deviation
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
Table 10. Internet Activity Participation *
E-mail 94%
Surfing the web 78%
Reading messages from newsgroups/Usenet/Bulletin Board (BBS) 58%
Listening to music 57%
Chatting/IRC/ICQ or instant messenger 52%
Playing games 37%
Participating in mailing lists of listserves 25%
Watching a webcast 25%
Maintaining a personal website 22%
Making Internet phone calls 17%
Participating in MUDs/MOOs 1%
*MuItiple answers were accepted. Median Internet activities participated in: M
(Standard Deviation = 2.86)
Correlations
Table 11 shows correlations between the Internet-related variables.
Correlations
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 11. Correlations between Internet Related Variables (N = 385)
Subscale 1 2 3
10 11 12 13 14
1. Internet
activity
scope
09 38*** 29*** 41*** 08 32*** 26*** iq*** 40*** 07 ig*** ^g*** 51*
2. Internet
activity
intensity
3. Time
spent online
4. Miss
computers
5. Miss the
Internet
29*** 4 9 *** 16** 16** 13** ig*** 05 15** 41*** 38***
2g*** 4 9 *** i q 30*** 34*** 06 38*** 00 10* 31*** 31^
55*** 12** 21*** 22*** 08 34*** 17*** 15** 28*** 41s 1
19*** 25*** 25*** 11* 41*** 10* 18*** 44*** 43*
6. Home
computer
ownership
7. Number
of places for
Internet
access
8. Years of
Internet use
25*** 12** -0 1 21*** 21*** 26*** 11* 38***
22*** 02 27*** _ 03 21*** 14** 20***
11* 32*** 15** 18*** 06 134
to
CN
Table 11. (Continued)
Subscale \ 2 3 _____ 4_____ 5_______6_____ 7
9. Receive
Internet-
related help
10. Provide
Internet-
related help
11 .
Proportion of
family
connected to
the Internet
12.
Proportions
of friends
connected to
the Internet
13. Internet
goal intensity
8 9 10 11 12 13 14
14** 10* 14** 19*** 12*
.02 .09 .31*** 3 7***
.28*** .03 .18***
12* .25***
7^***
14. Internet
goal scope
128
Table 12 shows inter-correlations among the variables in the Internet
connectedness index, and table 13 shows correlations between media-related
variables.
Table 12. Inter-Correlations between Variables in the Internet Connectedness Index
Subscale 1 2 3 4 5
1. Internet activity scope
n=385
.09
3g*** 29*** 4 ^***
2. Internet activity intensity —
2g*** 29*** 49***
3. Time spent online —
3g*** 49***
4. Degree of missing
computers
5. Degree of missing the
Internet
.66***
*** p < .001
Table 13. Correlations between Media-Related Variables
Subscale 1 2 3 4
Scope of connections to — 23*** .50*** .42***
media
TV connections — -.04 -.15***
Newspaper connections — .06
Interpersonal connections
*** p < .001
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
129
Hypotheses and Research Questions Testing
RQ1: Media Scope
What kinds of disparities exist among different socioeconomic and
demographic groups in the scope of communication media that respondents connect
to for achieving everyday goals? Media scope factor is regressed on socioeconomic
and demographic variables in a multiple regression (Table 14). The socioeconomic
(SES) factor, composed of income and education, has a significantly positive effect
on media scope factor (B - .12, p < .01); those who are of a higher socioeconomic
status have a broader scope of media that they connect to for achieving their
everyday goals. Age and gender do not have a significant effect on media scope.
Being an Armenian has a significantly negative effect on media scope, when White
is set as a reference group (B = -.26, p < .001). Armenians in Glendale are likely to
have a narrower scope of media that they connect to compared to Whites. Latinos
also have a narrower media scope when compared to Whites, but the difference is
not significant (B = -.08, p < .10).
Table 14. Summary of Simultaneous Multiple Regression Analysis for Variables
Predicting Media Scope (Standardized Coefficients)
Variable B
Socioeconomic status .12**
Age .02
Gender -.07
Latino -.08
Armenian -.26***
R-square .09
N = 514
** p < .01.***p <. 001.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
130
RQ2: Specific Media Intensity
What kind of disparities exist in the patterns of people’s connectedness to
specific communication media? Connectedness to each of four communication
media (television, newspapers, Internet and interpersonal communication) is
regressed on socioeconomic and demographic variables (Table 15).
Table 15. Summary of Simultaneous Multiple Regression Analysis for Variables
Predicting Television Connection, Newspaper Connection, Internet Connection and
Interpersonal Connection (Standardized Coefficients)
Variable Television
Connection
Newspaper
Connection
Internet
Connection
Interpersonal
Connection
Socioeconomic
status
_ 24*** 12** 20***
.03
Age .09*
19***
-.26***
1
to
*
*
Gender
(male=l,
female=0)
-.04 -.02 .10* -.06
Latino .01 -.02 -.14** -.06
Armenian .04 - 22*** -.16** -.13*
R-square .07 .11 .15 .03
N = 514
* p < .05. ** p < .01. *** p < .001.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
131
Television Connections
First, the television connection variable is regressed on socioeconomic and
demographic variables. Socioeconomic factor and age have significant effects on
television connectedness, but in the opposite direction. The socioeconomic factor has
a negative effect on television connectedness (B = -.24, p < .001), while age has a
positive effect on television connection (B = .09, p < .05). That is, those who are in
lower SES groups and those who are older are more likely to rely on television for
achieving their goals. Gender and ethnicity do not have significant effects on
television connection.
Newspaper Connections
In terms of newspaper connection, SES, age and ethnicity have significant
effects. The SES factor (B = .12, p < .01) and age (B = .19, p < .001) have positive
effects on the newspaper connection, while being an Armenian has a negative effect
on newspaper connectedness (B = -.22, p < .001) when compared to White. For
newspapers, those in a higher SES group, older and Whites are more likely to read
newspapers than other groups.
Internet Connections
The Internet connection is significantly influenced by all the variables
included in the equation: SES, age, gender and ethnicity. Socioeconomic status has a
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
132
positive effect on Internet connection (B = .20, p < .001) and younger age is
associated with stronger Internet connection (B = -.26, p < .001). Being male has a
positive effect on Internet connection (B = .10, p < .05), and being Latino and being
Armenian have both negative effects on Internet connection (B = -.14, p < .01, B = -
.15,p < .01, respectively), when White is set as a reference group. Overall, being in a
higher SES group (higher income and education), younger, male and White, the
respondent is more likely to connect to the Internet for achieving everyday goals.
Interpersonal Connections
In terms of connecting to interpersonal communication for achieving
everyday goals, age and ethnicity have significant effects. Age has a negative effect
on interpersonal connection (B = -.12, p < .01). That is, younger people are more
likely to engage in interpersonal communication when trying to achieve their goals.
Being an Armenian has a negative effect on interpersonal connection (B = -.13, p
< .05), which indicates that Whites are more likely to engage in interpersonal
communication for achieving goals compared to Armenians.
Internet Access
Factors affecting whether people have Internet access or not are analyzed
(Table 16). Logistic regression analysis is employed because the dependent variable,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
133
Internet access, is a dichotomous variable, where “1” means having an Internet
access, and “0” means not having the Internet access.
HI: Socioeconomic and Demographic Variables
In model 1, the Internet access variable is regressed on socioeconomic and
demographic variables. SES (b = 1.49, p < .001), age (b = -.07, p < .001), and
ethnicity (Armenians: b = -.90, p < .01) have significant effects on Internet access.
People with higher SES, who are younger and White are more likely to have Internet
access than their counterparts. The Bayesian Information Criterion (BIC)V of this
model is -186.30.
H2: Media Scope
In model 2, the media scope factor is added to the equation, thus testing the
effect of media scope on Internet access, controlling for socioeconomic and
demographic variables. Media scope does not have a significant effect on Internet
access (b = .10, p > .05). The BIC of model 2 is -180.50, which is higher than that
for model 1, indicating that the addition of media scope does not make a significant
contribution to predicting Internet access.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
134
RQ3: Specific Media Connections
In the third logistic regression, the effects of connections to three different
communication media on Internet access are tested, controlling for socioeconomic
and demographic variables. Among television, newspapers and interpersonal
connections, television connection has a significantly negative effect on Internet
access (B = -.35, p < .01), but the other two do not. That is, those who are more
connected to television are less likely to gain Internet access when other
socioeconomic and demographic variables are controlled. The BIC of this model is -
179.73, which is higher than the BIC of model 1. Considering that three variables are
added (television, newspaper and interpersonal connections), the incremental
improvement in model chi-square is not significant to make this model better than
model 1.
Final Model
Based on the significance of variables in the three models and the BIC scores,
a final model with six variables was formed. Variables included are SES, age, gender,
Latino ethnicity, Armenian ethnicity and television connection. This model has the
lowest BIC among all the models (-188.71) and is thus selected as a final model. For
this final model, exponents of bs (odd ratio corresponding to one unit change) are
reported in Table 16. According to the exponents of bs, those who have higher
income and education are over four times as likely to have Internet access as those in
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
135
Table 16. Summary of Three Logistic Regression Analyses for Variables Predicting
Internet Access
Variable Model
1 2 3 Final
Model
Exp (b)
of Final
Model
Constant
4 24***
4.22*** 4.65*** 4.62***
SES and
demographics
Socioeconomic
status
2 49*** 47***
^ 29*** 1 43***
4.16
Age
_ 07*** 07*** 07*** 07***
.94
Gender .21 .22 .17 .17 1.18
Latino -.12 -.10 -.09 -.1 .91
Armenian
_ 9Q**
-.85** -.77*
_ g7**
.42
Media scone
Media scope .10
Media connection
Television
connection
-.35** -.33** .72
Newspaper
connection
.29
Interpersonal
connection
-.08
Model chi-square 218.16*** 218.73*** 230.70*** 226.94***
Degree of freedom 5 6 8 6
BIC -186.30 -180.50 -179.73 -188.71
N = 585, * p < .05. ** p < .01. *** p < .001.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
136
lower income and education groups. Younger people are 6% more likely to have
Internet access than older people. Armenians are about half as likely to have Internet
access compared to Whites, or conversely, Whites are more than twice as likely to
have Internet access when compared to Armenians. Those who rely on television to
achieve their goals are about 30% less likely to have Internet access than those who
do not rely as much on television when socioeconomic and demographic variables
are held constant.
Internet Connectedness
What factors influence post-access inequalities in people’s Internet
connectedness? The Internet Connectedness Index (explained in the methodology
section) is regressed on various factors in four nested multiple regression models
(Table 17). Respondents included are those who use the Internet (n = 384).
H3: Socioeconomic Status and Demographic Characteristics
In the first model, where socioeconomic status and demographic variables are
included, socioeconomic status (income and education) (B = .11, p < .05) and age (B
= -•31, P < .001) have significant effects on Internet connectedness. Higher SES and
younger people have higher scores in the Internet connectedness index, i.e., having a
broader scope and more intense connectedness to the Internet. Gender and ethnicity
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
137
do not have significant effects on Internet connectedness. The BIC of this model is -
1.17.
H4: Social and Technological Environments
In model 2, where variables of technological and social environments are
added, proportions of family and friends who are connected to the Internet (B = .11,
p < .05), receiving and giving Internet-related help (B = .29, p < .001) and
technological environment (home computer ownership, Internet access place scope
and years of using the Internet) (B = .27, p < .001) have positive effects on Internet
connectedness, when socioeconomic and demographic characteristics are controlled.
Age remains a significant variable in this model. The BIC of this model is -27.86,
which is considerably lower than the BIC of model 1, indicating that the addition of
technological and social environments has a significant contribution to predicting
Internet connectedness.
RQ4: Media Connections
Connections to television, newspapers and interpersonal communication are
added to the equation in model 3. None of these communication variables show a
significant effect on Internet connectedness. The BIC of this model is -22.72, which
is higher than model 2. This model does not show a better fit than the previous model.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
138
H5: Goal Scope and Intensity
Finally, the Internet-related goal scope and intensity factor is included in the
equation. When all the variables included in the previous models are controlled, the
goal scope and intensity factor has a highly significant effect on Internet
connectedness (B = .48, p < .001). That is, those who have more intense and broader
motivation to seek goals on the Internet are more likely to have higher connectedness
to the Internet. In this model, socioeconomic status, age, ethnicity, receiving and
giving Internet-related help and technological environment have significant effects
on Internet connectedness. That is, along with a higher goal scope and intensity,
higher SES, younger, being White, receiving and giving Internet-related help to/from
others and having a better technological environment have a positive association
with having higher scores in the Internet connectedness index. The BIC of this model
is -54.71, significantly lower than the previous model. The goal scope and intensity
factor significantly increases the proportion of Internet connectedness explained by
the regression model.
Final Model
Based on the results of model 1 to 4, a final model is derived. Variables
included are socioeconomic status, age, gender, Latino ethnicity, Armenian ethnicity,
receiving and giving Internet-related help, the technological environment factor and
the goal scope and intensity factor. Significant factors in the final model are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
139
socioeconomic status (B = .12, p < .05), age (B = -.12, p < .05), Latino ethnicity (B =
13, p < .01), Armenian ethnicity (B = -.15,p<.01), receiving and giving Internet-
related help (B = .16, p < .001), technological environment (B = .19, p < .001) and
Internet-related goal scope and intensity (b = .48, p < .001). The BIC of this model is
-66.04, which is lower than the full model (model 4). This model is selected as a
final model.
Table 17. Summary of Hierarchical Multiple Regression Analyses for Variables
Predicting Internet Connectedness (Standardized Coefficients)
Model
1
SES and
demoeraohics
2 3 4 Final
Model
Socio- .11*
economic
Status
.01 .02 .11* .12*
Age -.31*** -.15**
1
t — i
L /t
*
*
-.10* -.12*
Gender .07 .05 .05 .06 .06
Latino -.11 -.11* -.12* -.13**
_ 13**
Armenian -.10
Social and
technological
environment
-.10 -.11
_ 14**
-.15**
Proportions of
family and
friends using
the Internet
.11* .11* .08
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
140
Table 17. (Continued)
Model
1 2 3 4 Final
Model
Receiving and
giving Internet-
related help
29*** jg***
Technological
environment
27*** 27*** 17***
Media
connections
Television
connections
.03 .02
Newspaper
connections
-.06 -.07
Interpersonal
connections
-.08 -.07
Goal scone and
intensity
Internet goal
scope and
intensity
4g***
R2 .09 .29 .30 .47 .45
BIC -1.17 -27.86 -22.73 -54.71 -66.04
N = 330
p < .05. ** p < .01. *** p < .001
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
141
v Bayesian Information Criterion (BIC) is an indicator useful for comparing different steps of
hierarchical regression models and for selecting the best model with the least number of independent
variables (Raftery, 1995). BIC is more conservative than the R-square increment significance test, and
it is especially sensitive to maintaining parsimoniousness. The formula for deriving BIC for multiple
regression is BIC=« In (1-R2 ) + p In n, where n is the sample size, R2 is the value of R2 for the model
of interest and p is the number of independent variables in the model. The BIC for logistic regression
is BIC= -% 2 + p In n, where x2 is the likelihood ratio test statistic for testing the null model with the
model of interest, p is the number of independent variables in the model, and n is the sample size. The
lower (i.e., the more negative) the BIC, the better the model captures the main features of the data
relative to other models (Biblarz, 2002; Raftery, 1995).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
142
CHAPTER 6. DISCUSSION AND IMPLICATIONS
In the final chapter, the author discusses (1) this dissertation’s theoretical and
empirical contributions, (2) the implications of the findings, (3) the findings in the
context of larger social inequality and communication technologies issues, (4)
limitations, and (5) policy implications of the research.
Theoretical Contributions
In terms of theoretical contributions, first, the now almost three decades-old
media system dependency theory was thoroughly reviewed, focusing on major
evolutions over its maturation. In the 1970s (Ball-Rokeach & DeFleur, 1976), media
system dependency theory focused upon the effects of macro-level media system
dependency relations on individuals’ cognitive, affective and behavioral dimensions.
In the 1980s (Ball-Rokeach, 1985), the focus of inquiry shifted to the origins of
individuals’ media system dependency relations. The origins identified were
structural dependency between media and other social systems, media system
activities, interpersonal discourse and the structural locations of individuals in
society. Also in the 1980s (Ball-Rokeach, 1985; Ball-Rokeach et al., 1984), a
typology of individuals’ media dependency goals, an important antecedent of
individuals’ media system dependencies, was developed: understanding (self and
social), orientation (action and interaction) and play (solitary and social). In the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
143
1990s (Ball-Rokeach, 1998), media system dependency theory was further
elaborated and clarified, particularly when it was explicitly contrasted with uses and
gratifications theory. Cross-level links between micro and macro levels were more
explicitly elaborated in the theoretical model, and micro-level agents’ roles in media
system dependency relations were given more emphasis (Ball-Rokeach, 1998). Also,
meso-level interpersonal networks were included in the theoretical model.
In addition to the review of theoretical pieces in the media system
dependency tradition, empirical studies that have either developed or applied the
media system dependency theory were reviewed and organized into four categories:
these categories focused, respectively, on (1) micro-level dependency relations, (2)
meso/macro-level dependency relations, (3) dependency relations in context of
ambiguous personal environments and (4) the effects of the larger social/political
environment upon dependency relations.
Third, limitations of media system dependency theory were identified. (1)
The term “dependency” has connotative meaning of inherent asymmetric
relationships between the two parties involved, often considered as weak individuals-
strong media relationships. (2) In media system dependency theory, the relationship
between media and individuals is conceptualized as asymmetric, the media
controlling gathering/creating, processing and disseminating resources, while only
dissemination is implicated in individuals’ media system dependency relations.
However, in the changing media environment, especially due to the popularization of
the Internet, individuals in this context can have control over creating and processing
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
144
information resources. Therefore, media system dependency relationship between
individuals and media becomes more of a variable rather than a constant. (3) Meso-
level agents are not sufficiently elaborated. (4) The personal environment has been
treated as an important factor affecting individuals’ media system dependency
relations, but the conceptualization is more abstract than grounded. Ways of
including the personal environment into empirical analyses have not been adequately
explored in the media system dependency studies.
The fourth theoretical contribution of this dissertation was the explication of
communication infrastructure theory, which evolved from media system dependency
theory to overcome its limitations. Communication infrastructure theory seeks to
understand how people are communicating with one another under the rapid
transformation of their social and personal environments. The communication
infrastructure is a network of individuals, organizations and media (storytelling
system) set in its communication environment (communication action context).
Communication infrastructure theory departs from media system dependency theory
in four major ways. (1) Communication infrastructure theory includes all the
communication modalities available in people’s everyday lives, while the mass
media system is the main concern in media system dependency theory. (2) The
concept of a storytelling system specifies macro-, meso- and micro-level of analyses
not only in terms of different social locations of the agents, but also of the kind of
stories that agents tell. (3) In communication infrastructure theory, individuals are
more active participants in the storytelling system, interacting with macro- and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
145
meso-level communication channels, while in media system dependency theory, the
macro-level media system was mainly conceptualized as constrains on individuals’
agentic activities. (4) With the concept of communication action context, the
communication infrastructure theory takes a more grounded approach in
conceptualizing personal and social environments. Although communication
infrastructure theory was introduced and explained in other studies (e.g., Ball-
Rokeach et al., 2001), its relationship to and differences with media system
dependency theory had not been articulated prior to this dissertation.
The fifth contribution of this dissertation was an elaboration and articulation
of the communication connectedness concept. Connectedness was developed in
terms of its main theoretical assumptions and sub-dimensions. The four dimensions
of connectedness specified are access to a communication medium, the intensity, the
scope and the centrality of the relationship.
Sixth, in addition to these theoretical framework concerns, the researcher
developed a framework called “double barriers in communication technology and
equality.” This framework conceptualizes the process by which social inequality not
only affects (1) individuals’ connections to information resources via communication
technologies, but also (2) the utilization of information resources attained via
connections to communication technologies in people’s social practices (See Chapter
1). That is, double barriers exist that make it difficult for the potential of a new
communication technology to radically alter pre-existing social inequality patterns.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
146
This framework helps conceptualize the relationship between communication
technology and social inequality at both micro-individual and macro-social levels.
Empirical Contributions
Working from the double barriers, media system dependency and
communication infrastructure frameworks, the author operationalized individuals’
connectedness relationships to the Internet into four dimensions. First, access
involved having or not having an Internet connection. Second, intensity was
measured by the time spent with a medium and perceived helpfulness of a medium in
attaining everyday goals. Third, scope was measured by the number of genre
activities (e.g., e-mail, playing games, chatting, maintaining personal websites,
listening to music, etc.) that a person participates in on the Internet. Centrality was
measured by how much a person would miss computers and the Internet if they were
to vanish.
An empirical model that hypothesized relationships between
multidimensional antecedent factors and Internet connectedness were derived based
on the ecological approach of media system dependency and communication
infrastructure theory. The inequality and communication technologies literature
reviewed in Chapter 3 also contributed to specifying variables to be included in the
model. First, the intensity and scope of individuals’ goals for connecting to the
Internet was identified as a factor that directly affects individuals’ connectedness to
the Internet. The second antecedent was the individual’s social environment,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
147
measured by the degree of Internet-related help received from or given to others and
proportion of friends and family members who use the Internet. Third, a
technological environment factor was created, measured by home computer
ownership, number of places where Internet access is available and years of Internet
use. Fourth, the number of media forms (mass media and interpersonal
communication channels) that people connect to for achieving their everyday goals
was selected as a factor that is likely to affect people’s Internet connectedness. Fifth,
the intensity of individuals’ connections to each of television, newspapers and
interpersonal communication channels was calculated to indicate their connections to
specific media forms. Finally, socioeconomic status (income and education) and
demographic factors (age, gender and ethnicity) were included as antecedents to
Internet connectedness.
Including the Internet connectedness, four dependent variables were
regressed upon theorized antecedent variables: media scope, specific media
connections, Internet access and Internet connectedness. For each dependent
variable, antecedent variables were included in a regression model, thus allowing for
partialing out each independent variable’s effect on the dependent variable. First, in
terms of media scope, higher socioeconomic status and Whites (compared to
Armenians) is associated with a broader media scope people connect to for achieving
their everyday goals of understanding, orientation and play. Second, in terms of
specific media connections, the higher socioeconomic group is more likely to choose
newspapers and the Internet for achieving goals, while they are less likely to choose
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
148
television. Older people are more likely to read newspapers but less likely to go on
the Internet or engage in interpersonal communications for their goals. Latinos are
less likely to go on the Internet than Whites, and Armenians are less likely to read
newspapers, go on the Internet and engage in interpersonal communications than
Whites, when other independent variables are held constant. There are important and
complex socioeconomic and ethnicity variations that persist despite present levels of
media diffusions.
Third, when Internet access was logistically regressed upon other variables,
higher socioeconomic status, younger age, being White (compared to Armenians)
and lower connections to television were associated with the higher likelihood of
having Internet access.
Fourth, in terms of the multidimensional Internet Connectedness Index,
higher socioeconomic status, younger age and being White (compared to both
Armenians and Latinos) were associated with higher Internet connectedness scores.
The social environment (higher interactions with others regarding giving and
receiving Internet-related help) and the technological environment (having a
computer at home, the number of places for going online, and length of experience
with the Internet) significantly influenced the Internet connectedness. Having higher
Internet goal scope and intensity was also found to have a significant impact on the
Internet connectedness.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
149
Implications of the Findings
Descriptive Findings
In light of the descriptive analyses reported at the beginning of Chapter 5, the
dissertation produced findings about people’s media connection patterns that may be
worth of note. First, when respondents were asked to identify important media
modalities for staying on top of what is happening in their community, television
came in first (55%), followed by newspapers (40%) and interpersonal
communication (25%). The Internet (14%) was not, at this point in time, a dominant
medium for providing community information. For the goal of getting information
on products to buy, about one-third of the respondents chose interpersonal
communication (34%), and another one-third chose television (32%). The Internet
was chosen by a quarter of the respondents (24%). The importance of “word-of-
mouth” in getting information about products to buy is prominent here. Finally, for
getting medical information, interpersonal communications (20%) and the Internet
(19%) were chosen by one-fifth of the respondents, followed by television (16%).
The Internet played a more significant role in terms of medical information than
television, unlike in other goals. The result is consistent with recent studies that
report increasing use of the Internet for obtaining health-related information and
services (Block & Mire, 1996; Dupuits, 2002; Finkelstein & Friedman, 2000;
Lieberman et al., 2003).
In terms of the descriptive findings about computers and the Internet, 67% of
the total respondents (n = 585) had Internet access (at least at one location), and 72%
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
150
had computers at home. More than half of Internet connectors (57%) went online at
home, one-third at work (35%), 14% at school, and 7% at public library. Home and
work are the major Internet access contexts. When asked about sources of help that
people can go to when having problems with the Internet, family (26%) and friends
(22%) were the most common sources, followed by technical support (11%), online
help (9%) and co-workers (7%). This shows the importance of the immediate social
network in shaping the ways in which people connect to the Internet (Howard et al.,
2001; Jung et al., 2002; Matei & Ball-Rokeach, 2001; Murdock et al., 1992;
Silverstone & Hirsch, 1992; Wellman et al., 2001).
The median number of goals for going online was six, out of nine possible,
and the median number of activities engaged in on the Internet was three, out of
eleven possible. That is, respondents connected to the Internet for six different goals,
and participated in three different activities on average. This suggests that the same
activity may fulfill more than one kind of goal that people have for using the Internet.
Among the Internet activities people participate in, e-mail was used by most Internet
connectors (94%) followed by surfing the web as an activity (78%). Over a half of
the respondents (58%) participated in newsgroups, Usenet and bulletin boards (BBS),
as well as listened to or downloaded music on the Internet (57%). About half
participated in chatting or instant messaging (52%), and one-third of them played
games on the Internet (37%). That is, the Internet is most strongly implicated as a
communication resource for the purpose of interacting with other people (e-mail,
newsgroups/Usenet/BBS, chatting), either simultaneously or not, and for
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
151
entertainment (playing games and listening to music) (Flanagin & Metzger, 2001).
About one-fifth of the respondents reported that they maintained a personal website.
Finally, mean time spent on the Internet a week among the Internet users was 12
hours, that is, about an hour and forty minutes a day.
Media Connections
Media Scope
When other variables are controlled, those who belong to higher income and
education groups are likely to have more media available to them to connect to when
trying to achieve their everyday goals. That is, they are also communication resource
rich in ways that matter. From the perspective of media system dependency theory,
the “structural location” of a person is an important factor that affects his or her
access to media and also to alternative information resources (Ball-Rokeach, 1985, p.
505). Income and educational levels, common indicators of people’s socioeconomic
status, influence the degree of diversity in communication channels available in
people’s everyday lives. On the other hand, those who are in lower socioeconomic
strata are likely to have a limited number of media that they can go to when trying to
achieve their goals, i.e., relatively poor regarding communication resources. For
example, television is usually associated with lower socioeconomic status (Graham,
1999; Matzger, 1983; Mullan, 1997; Dhavan V. Shah et al., 2001; Webster, 1986),
and those who rely highly on television seem to do so almost exclusively (Moy et al.,
1999; Van Eijck & Van Rees, 2000). These finding suggest that the diversity of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
152
communication channels employed by individuals is a variable that can be included
in future research.
The finding also suggests that the communication resource poor (i.e., lower
SES) are less likely to be able to effectively navigate their communities to learn of,
and to locate resources to aid in daily living. When compared to one or two media
channels, a wider scope of media first implies more diverse kinds of information
from different sources in different formats and, second, individuals having more
control over the kinds of information to access and incorporate for the specific goals
that they have. From the communication infrastructure perspective, a diverse, intense,
and integrated communication infrastructure facilitates more active and constructive
stories generated and exchanged by the members of a community (Ball-Rokeach et
al„ 2001).
Also shown to significantly influence media scope is ethnicity. Armenians
were found to have fewer media that they connect to for goal attainment when
compared to Whites and when socioeconomic status, age and gender were held
constant. That is, within the same communication infrastructure in Glendale,
Armenians use fewer media to achieve their goals than Whites. A study by Kim and
this researcher (Kim & Jung, 2002) explored the effect of “geo-ethnicity,” an
ethnicity bounded in a geographical, social and cultural environment. By proposing
the concept of geo-ethnicity, we examined the unique effect of an ethnic group living
in a certain area on their communication and social behaviors. In assessing people’s
neighborhood engagement, we found that in Glendale, utilizing the same dataset as
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
153
the one used in this dissertation, Whites have a higher degree of belonging to their
neighborhoods (measured by a belonging index introduced in Ball-Rokeach et al.,
2001) and higher community organizational membership than Armenians and
Latinos, indicating that Whites are more actively involved in the “storytelling
network” of their community relative to Latinos and Armenians (Kim & Jung, 2002).
The present finding that Armenians have fewer media that they connect to for
achieving their goals supports a similar argument. Whites, more than other ethnic
groups in Glendale, are in a better structural location to deploy different
communication media, and they are more likely to be actively involved in the
storytelling network in their community than Armenians and Latinos.
Specific Media Connections
In light of specific media connections, age is shown to be an important
variable that affects specific media connections. Television and newspapers are more
widely selected among the older age group, while the Internet and interpersonal
communication are more commonly chosen by the younger age group, as previous
research suggested (e.g., Lauf, 2001). This result touches upon the issue of a
changing media environment under the emergence of new communication
technologies. With the emergence of Internet and its various news sites, related
questions have arisen (Lauf, 2001; Peiser, 1999), such as What will the future of
newspapers be when current young people become middle age group?, and Are
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
154
relatively low newspaper connections among the younger age group an age or a
cohort effect?. The difference between age and cohort effects is that age effect
mostly concerns people’s current age, rather than the culture of a certain generation
carried on as they age. On the other hand, a cohort effect concerns the cultural and
behavioral characteristics of an age group. That is, when the age group gets older,
they carry the characteristics along. Therefore, if low newspaper connections among
young people are attributed to age itself, their newspaper connections will increase
when they become, for example, 40. On the other hand, if the cohort factor is in
effect, they will carry their low-newspaper/high-Intemet connections as they age.
The latter means more radical change in societal media connectedness patterns.
Internet Access
In terms of the variables affecting whether or not individuals have Internet
access, the results proved to be consistent with most past literature on the topic, i.e.,
higher socioeconomic status (Nie & Erbring, 2000; UCLA Center for
Communication Policy, 2001; US Department of Commerce, 2000, 2002), younger
age group (Adler, 1996; Cody et al., 1999; Lenhart, 2000; US Department of
Commerce, 2000, 2002; White & Weatherall, 2000) and ethnicity (being White)
(Babb, 1998; Bikson & Panis, 1999; US Department of Commerce, 1998,1999,
2000; Wilhelm, 2000) were associated with “haves” in Internet access.
Among the media connection variables that were included in the Internet
access regression model, having connections to television was found to have a
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
155
significantly negative effect on Internet access. That is, those who connected
strongly to television were less likely to have Internet access than those who relied
less on television (other variables in the model held constant). Flanagin and Metzger
(2001) clustered different communication media into three groups. The researchers
clustered television and Internet’s information retrieval and information giving
functions into the same group, sharing what they call similar “functional images” (p.
165). Flanagin and Metzger argued that as more and more people gain access to new
media such as the Internet, the newer media are likely to perform the roles of more
traditional media due to their capacities to improve and augment the capabilities of
existing media (p. 171). From this perspective, the negative association between
television and the Internet may be explained in terms of the Internet displacing
television (Coffey & Stipp, 1997; Kayany & Yelsma, 2000). However, more in-
depth studies are desirable to better interpret what it means to maintain high
connections to television in the new media environment or to move to the Internet.
Internet Connectedness
Internet connectedness was defined as a post-access measure of individuals’
multidimensional relationship with the Internet. Internet connectedness consists of
intensity, scope and centrality. Among those who already have Internet access,
disparities exist in terms of basic socioeconomic and demographic variables, such as
income, education, age and ethnicity in the ways in which people incorporate the
Internet into their everyday lives. This observation was strongly made in past
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
156
Internet connectedness research (Jung et al., 2001; Kim et al., 2002; Loges & Jung,
2001). Other studies have also detected post-access inequalities in socioeconomic
status (General Social Survey, 2000; Hindman, 2000; Howard et al., 2001; US
Department of Commerce, 2002), age (Lenhart, 2000; US Department of Commerce,
2002) and ethnicity (Howard et al., 2001; Spooner & Rainie, 2000; Wilhelm, 2000).
The technological environment was found to have a highly significant effect
on Internet connectedness. Among the three variables constituting the technological
environment factor, home computer ownership (e.g., Attewell & Battle, 1999; Perse
& Dunn, 1998) and years o f Internet use (e.g., Hargittai, 2002; Horrigan & Rainie,
2002) were consistently found to be important variables that influence people’s
Internet access and use. The number o f Internet access places (the third dimension of
the technological environment) has not been explored to same extent in previous
research. Relevant to the discussion of the effect of individuals’ structural locations
on their media connections, the kind of communication infrastructure that individuals
are located in and the capacities of individuals to make use of the technological
resources available to them are likely to affect the ways in which individuals
incorporate the Internet into their everyday lives. For example, those who only use
the Internet at work are likely to have a different pattern of connectedness when
compared to others who use it at home, work and a community center.
Not only was the technological environment considered in this study, but also
the social environment. A unique variable included in this study was the proportions
offamily and friends using the Internet and the degree o f interactions with others
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
157
regarding receiving and giving Internet-related help. Although several other studies
have emphasized the importance of social environment in affecting people’s Internet
connections (Atkin, 2001; Bradley, 2000; d'Haenens, 2001; Haddon, 2000; Kraut,
Rice et al., 1998; Murdock et al., 1992; Suss, 2001), few of them have empirically
tested the influence of the social environment on Internet connections (c.f., Jung et
al., 2002). In the final model of Internet connectedness (pp. 134-138), the degree of
interaction with others regarding Internet-related help was found to have a significant
effect on Internet connectedness, even when socioeconomic, demographic and other
variables in the model were controlled. This indicates that when other conditions are
equal, individuals who have social ties with whom they interact in resolving Internet-
related problems are more likely to engage in broader and more intense Internet
activities and to perceive the Internet to be more central to their everyday goal
attainment. This result extends the argument made by scholars emphasizing the
concept of social capital (Alkalimat & Williams, 2001; Attewell & Battle, 1999;
Borgida et al., 2002; Coleman, 1998; Pruijt, 2002; Putnam, 2000; Wellman et al.,
2001), in which people’s social network is considered a crucial resource, or the
capital, that is likely to provide individuals with meaningful benefits in leading their
social and career lives. That is, not only is the technological environment highly
important in helping people to appropriate a given opportunity, but also people’s
social environment. Internet connectedness is no exception. Both technological and
social environments as detailed in the present research need to be included in future
research designed to address communication technology and inequality issues.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
158
The largest variation in individuals’ Internet connectedness was explained by
Internet goal scope and intensity. The result has a direct theoretical origin in media
system dependency theory. According to the theory, goals are one of the most
important motivations for individuals to connect to certain communication media
(Ball-Rokeach, 1985). Therefore, the degree of intensity and the scope of goals that
people have in their everyday lives are likely to affect which media people choose
and which connectedness pattern they undergo in trying to obtain the goals via
certain media. In terms of Internet connectedness, if a person has a broader scope of
goals that he or she perceives the Internet to be useful for, the person is likely to
engage in broader scope of activities on the Internet to find resources to fulfill the
goals. Also, if a person perceives an Internet-related goal to be highly important in
his or her life, the person is likely to engage intensively in an Internet activity that
helps attain the goal. The proposition in the media system dependency theory that the
scope and intensity of goals is an important antecedent factor affecting individuals’
media dependency relations is strongly supported by the present findings on the
Internet.
Methodological Implications
The Internet connectedness index developed and used in this dissertation has
significant methodological implications for studies examining the Internet and
inequality. As more and more people gain access to the Internet, several studies
recognized the importance of examining post-access disparities in the ways in which
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
159
people use the Internet (Bonfadelli, 2002; Castells, 1999; Hargittai, 2002; Jung et al.,
2002; Jung et al., 2001; Lievrouw, 2000; Newhagen & Bucy, 2001; Patterson &
Wilson III, 2000; Van Dijk, 1997). However, not many studies developed an
empirical way to measure the disparities in people’s online activities. A majority of
studies used time measures, i.e., asking how much time people spend on the Internet
to indicate their Internet use (e.g., Kraut, Patterson et al., 1998; Nie & Erbring, 2000;
Nie & Hillygus, 2002; Robinson et al., 2002). However, as indicated earlier (Chapter
1), time spent on the Internet is not an appropriate measure for examining what
people do on the Internet and for which purposes (Jung et al., 2001).
Several studies developed ways to measure people’s Internet use beyond time
spent online (Anderson, Bikson, Law, & Mitchell, 1995; Bikson & Panis, 1999;
Hindman, 2000; Howard et al., 2001; Norris, 2001; Dhavan V. Shah et al., 2001;
Wilhelm, 2000). For example, Howard et al. (2001) developed a typology of Internet
use with two questions: How long have you had Internet access? and How frequently
do you log on from home? Four broad categories of Internet users were identified:
netizens, utilitarians, experimenters and newcomers. Hindman (2000) created a
seven-item index called “information technology index.” The items included in the
index are “have any type of personal computer in home,” “ever go online,” “ went
online yesterday,” “ever send or receive e-mail,” “sent or received any e-mail
yesterday,” “go online to get news” and “have ever purchased goods or services
online” (p. 554). Shah et al. (2001) categorized Internet use into four components
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
160
using factor analysis of 11 items asked in a survey: product consumption,
information exchange, financial management, and social recreation.
These measures, although useful in conceptualizing different Internet usage
patterns and in categorizing Internet users into several groups, are not very
informative when trying to diagnose inequalities in the breadth and depth of people’s
Internet use compared with other media and also over time. More importantly, these
measures were not developed from a theoretical framework. Rather, most were post
hoc measures, based on the data they had. When trying to examine how the Internet,
as a new medium, came into our society, and was incorporated into the existing
media environment where many traditional media existed, a measure that is capable
of applying to and having been applied to other media is crucial in detecting how the
Internet is different from other media, and how we should interpret the disparities
found in people’s Internet use. The Internet connectedness index developed in this
dissertation was based on the theoretical and empirical research of the media system
dependency measures on traditional mass media (Alman, 1993; Ball-Rokeach, 1985,
1998; Ball-Rokeach et al., 1990; Ball-Rokeach et al., 1984; Becker & Whitney,
1980; Colman, 1990; Grant et al., 1991; Hirschburg et al., 1986; Loges, 1992; Loges
& Ball-Rokeach, 1993) and their evolution into the concept of connectedness in
more recent research under the new media environment (Ball-Rokeach et al., 2000;
Ball-Rokeach et al., 2001; Gibbs, Ball-Rokeach, Jung, Kim, & Qiu, in press; Jung et
al., 2002; Jung et al., 2001; Kim et al., 2002; Loges & Jung, 2001). Therefore, the
connectedness measures can be used across different media when studying
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
161
individuals’ connections to one medium vis-a-vis other communication media. In
addition, due to the measure’s theoretical nature, the Internet connectedness index is
not only useful in detecting multidimensional disparities in people’s Internet use at
this moment, but it also would be useful in examining the long-term process of how
people incorporate the Internet into their everyday lives as their experiences with the
Internet develop.
Limitations
Limitations exist. First, the Internet Connectedness Index (ICI) is still in the
process of development and can be improved in future research. Although being able
to discern higher from lower Internet connectors, each dimension of the ICI (scope,
intensity and centrality) has not been analyzed separately. Analyzing three
dimensions separately will allow the researcher to examine the relationships between
social factors and each aspect of people’s Internet connectedness.
Second, the connectedness measure was only applied to the Internet. In the
future research, it may well be applied to other media, including traditional media
and new media. Applying the similar measure to different media will allow
researchers to compare and contrast what each media mean for individuals in the
context of others.
Third, the analyses in this dissertation focused mostly on disparities in
individuals’ connections to the Internet, one of two inequality dimensions discussed
in Chapter 1. Another dimension of inequality relating to the Internet is how the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
162
resources obtained on the Internet are utilized in people’s everyday lives. The ways
in which social factors affect how Internet resources are utilized in social, career and
political lives are another understudied research topic. This area, however, is more
challenging to examine empirically because how the Internet resources are actually
utilized in real life occasions should be tracked down. In addition, the consequence
of the Internet connections should be studied not only at a micro-level, but also at
meso- and macro-levels. The ways in which the Internet is applied to people’s goal
achievement in occupational and community settings, and also how the Internet is
being incorporated at organizational, inter-media, and community levels are left for
future research.
Policy Implications
Larger Issue: Communication Technology and Social Inequality
This dissertation’s first chapter argued that the advent of a communication
technology and discovery of new uses for the technology would not directly translate
to narrowing social inequalities. When the Internet diffusion rate was increasing
drastically, many people predicted what Internet access would bring to
disadvantaged people. The researcher argued that double barriers in connecting and
utilizing resources provided by communication technology make it difficult for a
communication technology per se to solve social inequality problems (Chapter 1).
The analyses in this dissertation focused upon the first barrier of connecting to
Internet resources. Individuals’ connections to resources for their goal attainment
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
163
were conceptualized into two steps: first gaining access to the Internet technology
and second, developing connectedness relations with the Internet. This dissertation’s
results strongly support the argument that even when the first step of gaining access
to the Internet is achieved, all Internet users do not develop their connectedness
patterns equally.
Moreover, unequal patterns of Internet connectedness are highly associated
with various social inequality indicators, including not only socioeconomic and
demographic factors, but also unequal social and technological environments
surrounding individuals’ daily lives. This result also aligns with general assumptions
made in the first chapter that communication technologies are embedded in a larger
social system where inequalities of people’s income, education, gender or ethnicity
exist. That is, when people start using the Internet, the ways in which they make use
of it are not an isolated phenomenon happening between individuals and a
technology on the cyberspace, but rather they are extensions of people’s offline
social, cultural, and occupational locations and environment. Individuals’
socioeconomic status, age, ethnicity, social environment, technological environment,
media connections, and goal intensity and scope are all reflected in multiple
dimensions of individuals’ Internet connectedness. That is, individuals who have
been advantaged in terms of these social indicators are again ahead of those who are
less advantaged, when their degrees of Internet connectedness are compared.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
164
Policy Implications
What does this result imply for policy interventions? The Internet’s potential
to provide various resources to users is not a solution to social disparities on its own.
Individual users should be able to know how to locate specific resources they need
for their goal attainment and to discern appropriate resources from those that are
either faulty or provide less useful information for their purposes. Possible
intervention comes in here. The Internet’s potential to help resolve social inequality
lies first in society’s effort to educate people with the skills to derive appropriate
resources from the Internet and, second, to provide open venues for those who have
learned the Internet skills to make use of them in their careers (Morino Institute,
2001).
The nature of disparity in connecting and utilizing Internet resources is
similar to that in post-secondary education. Post-secondary education (i.e., colleges
and universities) also has two barriers: entry and utilizing resources obtained during
the time spent in school. Economic capacity, capability to earn sufficient scores on
entrance exams and motivation are important antecedent variables for entering a
college. At the same time, people’s structural locations in society, especially their
social and cultural capitals, intervene in the process of utilizing the educational
experience and applying the resources to enhance their careers. Continuous efforts
are being made for colleges to serve as a means for narrowing inequalities, rather
than increasing the inequalities in our society. However, the matter is not simple.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
165
The Internet and inequality is not far from the situation that education has been
facing.
Policy makers, corporate managers and academics, who lead opinions and
can influence people’s expectation of the Internet, should be cautious in predicting
the Internet’s potential. Not much evidence suggests that the Internet is narrowing
persistent societal disparities. Particularly problematic is that as the Internet becomes
more and more central to certain groups of people (most likely high income, high
education, younger, white male groups), those who are not able to keep up with these
groups will have dual disadvantages — excluded from Internet resources and thus
excluded from the possibility of moving upward on the social ladder. In order for
policies and interventions to prevent the gap between haves and have-nots from
increasing and to make the Internet a technology that can give a wider opportunity to
those who have lacked alternative ways of overcoming their underprivileged social
positions, systematic educational programs should be actively developed that can
give people not only the technology, but also the skills and knowledge to make use
of it and to apply the resources and skills obtained from it. Several non-profit and
research organizations have emphasized the importance of developing policies and
educational programs to bridge the digital divide (e.g., Benton Foundation,
Metamorphosis Project, Morino Institute, The Children’s Partnership). For example,
since the mid-1990s, the Benton Foundation has taken the initiative to feature the
digital divide in public discourse by releasing reports (Goslee, 1998; Mazmanian,
1995); launching and maintaining the Digital Divide Network
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
166
(www.digitaldi videnetwork. or g) that offers a range of information, tools and
resources that help practitioners and others stay on top of digital divide
developments; and initiating and moderating digital divide e-mail listserv.
Morino Institute, a non-profit organization based in Washington, D.C., has
released many reports (Morino Institute, 1995, 1999, 2001) that focus on policy
suggestions on how the Internet can be used to benefit society as a whole. One of
their goals is to “prepare people to take advantage of the Internet's benefits with
particular emphasis on those being left behind: people in low-income communities
who already lack access to educational and economic opportunities”
(www.morino.org/under overview.asp). For example, they offer a 10 step action
agenda for individuals to first prepare themselves, then help change the institutions
in people’s communities and finally establish criteria for action (Morino Institute,
1995).
The Metamorphosis Project at the University of Southern California has
studied both theoretical and policy-oriented aspects of the Internet and its relation to
social disparities (Ball-Rokeach et al., 2000; Jung et al., 2001; Kim et al., 2002;
Loges & Jung, 2001; The Metamorphosis Project, 2003a, 2003b). The project
launched its first policy-oriented White Paper in 1999 on the visions and realities of
how new and old immigrants in the United States connect to the Internet and how
different ethnic groups form connectedness relationships with the Internet in
different ways (Ball-Rokeach et al., 2000). The Metamorphosis Project researchers
argued that the open architecture of Internet technology did not necessarily lower the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
167
ethnic and socioeconomic divides. Rather, the Internet connectedness patterns of the
respondents in the study differed by ethnic and other socio-cultural lines. The project
recently released two additional White Papers: first, “The Well-Connected
Community: A New Tool For Planning to Use the Internet to Build Community,”
which introduces a practical tool for planners and others to use to design strategies
and interventions that maximize the Internet’s impact on the local-level to build a
civil society (The Metamorphosis Project, 2003b), and second “Los Angeles-Area
Community Technology Centers: Successes and Challenges on the Road to
Community Building,” which reports the project’s research on Los Angeles area
community technology centers and provides policy recommendations for the
communication technology centers to not only provide Internet access, but also “the
skills, training, and exposure to technology that can offer career enhancement, life
improvement, and better connections to other family, friends, and neighbors” (The
Metamorphosis Project, 2003a, p. 3).
CTCNet (Community Technology Center Network, www.cetnet.org), a non
profit organization of more than one thousand community technology centers in the
United States, is actively engaged in holding annual conferences and producing
reports. The mission statement states that CTCNet is “a leading advocate of equitable
access to computers and related technologies; we will invite, initiate, and actively
encourage partnerships and collaborations with other individuals and organizations
that offer resources in support of its mission” (http://www.ctcnet.org/mission.html).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
168
The efforts of many organizations are likely to help people better incorporate
the Internet into their everyday lives. However, they may not achieve their missions
quickly. The Internet’s potential to influence social relations and structure should be
predicted and planned cautiously with a longer period of time in mind. We should be
reminded that the majority of us did not read online newspapers nor buy books on
the Internet ten years ago. The Internet is still a “new” technology. It would behove
us to avoid making hasty judgments about its potential and to give more careful
consideration to the ways in which the society can incorporate this technology to
narrow social inequalities as this new technology ages.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
169
BIBLIOGRAPHY
Adler, R. P. (1996). Older adults and computers: Report o f a national survey.
Alasuutari, P. (1999). Introduction: Three phases of reception studies. In P.
Alasuutari (Ed.), Rethinking the media audience: The new agenda (pp. 1-21).
Thousand Oaks, CA: Sage Publications.
Alkalimat, A., & Williams, K. (2001). Social capital and cyberpower in the African
American community: A case study of a community technology center in the
dual city. In L. Keeble & B. Loader (Eds.), Community infomatics:
Community development through the use o f information and communications
technologies (pp. 177-204). London: Routledge.
Allen, K., & Rainie, L. (2002). Parents online. Washington, DC: Pew Internet &
American Life Project.
Alman, K. C. (1993). The role o f television advertising in acculturation: A media
system dependency approach. Unpublished Ph.D. dissertation, Michigan
State University, Lansing, MI.
Amsden, A. H., & Clark, J. C. (1999). Software entrepreneurship among the urban
poor: Could Bill Gates have succeeded if he were black?... Or impoverished?
In D. A. Schon, B. Sanyal & W. J. Mitchell (Eds.), High technology and low-
income communities: Prospects for the positive use o f advanced information
technology (pp. 213-234). Cambridge, MA: MIT Press.
Anderson, R. H., & Bikson, T. K. (1998). Focus on generic skills for information
technology literacy. Santa Monica, CA: Rand Corporation.
Anderson, R. H., Bikson, T. K., Law, S. A., & Mitchell, B. M. (1995). Universal
access to e-mail-feasibility and societal implications. Santa Monica, CA:
Rand Corporation.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
170
Athreye, S. S. (1999). Relative underdevelopment as a barrier to technological
efficiency: A comparative study of ministeel plants in India and the UK.
Modern Asian Studies, 33(3), 733-758.
Atkin, D. J. (2001). Home ecology and children's television viewing in the new
media environment. In J. Bryant & J. A. Bryant (Eds.), Television and the
American family (2nd ed., pp. 49-74). Mahwah, NJ: Lawrence Erlbaum
Associates, Publishers.
Attewell, P., & Battle, J. (1999). Home computers and school performance. The
Information Society, 15, 1-10.
Babb, S. F. (1998). The Internet as a tool for creating economic opportunity for
individuals and families. Unpublished Ph.D. dissertation, University of
California, Los Angeles.
Ball-Rokeach, S. J. (1973). From pervasive ambiguity to a definition of the situation.
Sociometry, 36, 378-389.
Ball-Rokeach, S. J. (1985). The origins of individual media-system dependency: A
sociological framework. Communication Research, 12, 485-510.
Ball-Rokeach, S. J. (1998). A theory of media power and a theory of media use:
Different stories, questions, and ways of thinking. Mass Communication &
Society, 7(1/2), 5-40.
Ball-Rokeach, S. J. (2002). Media system dependency theory and communication
infrastructure theory. In J.-Y. Jung (Ed.). Los Angeles, CA: Excerpts from
personal communications.
Ball-Rokeach, S. J., & Cantor, M. G. (Eds.). (1986). Media, audience, and social
structure. Newbury, CA: Sage Publications.
Ball-Rokeach, S. J., & DeFleur, M. L. (1976). A dependency model of mass media
effects. Communication Research, 3, 3-21.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
171
Ball-Rokeach, S. J., Gibbs, J., Jung, J.-Y., Kim, Y.-C., & Qiu, J. L. (2000). The
globalization o f everyday life: Visions and reality (White Paper). Los
Angeles: Metamorphosis Project, Annenberg School for Communication,
University of Southern California.
Ball-Rokeach, S. J., Grant, A. E., & Horvath, A. T. (1993). On measuring
individuals' relationships with the media.Unpublished manuscript, Los
Angeles: University of Southern California.
Ball-Rokeach, S. J., Hale, M., Schaffer, A., Porras, L., Harris, P., & Drayton, M,
(1999). Changing the media production process: From aggressive to injury-
sensitive traffic crash stories. In D. P. Demers & K. Viswanath (Eds.), Mass
media, social control, and social change: A macrosocial perspective.
Malvern, IA: Iowa State University Press.
Ball-Rokeach, S. J., Kim, Y.-C., & Matei, S. (2001). Storytelling neighborhood:
Paths to belonging in diverse urban environments. Communication Research,
28(4), 392-428.
Ball-Rokeach, S. J., Power, J. G., Guthrie, K. K., & Waring, H. R. (1990). Value-
framing abortion in the United States: An application of media system
dependency theory. International Journal o f Public Opinion, 2, 249-273.
Ball-Rokeach, S. J., Rokeach, M., & Grube, J. W. (1984). The great American values
test: Influencing behavior and belief through television. New York: Free
Press.
Bamouw, E. (1990). Tube o f plenty: The evolution o f American television (2nd ed.).
Oxford: Oxford University Press.
Baym, N. K., Zhang, Y. B., & Lin, M.-C. (2001, May 25). Is the Internet really any
different? Social interactions across media. Paper presented at the
International Communication Association Annual Convention, Washington,
DC.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
172
Becker, L. B., Sobowale, I., & Casey Jr, W. (1979). Newspaper and television
dependencies: Their effects on evaluations of public officials. Journal o f
Broadcasting and Electronic Media, 23, 465-475.
Becker, L. B., & Whitney, D. C. (1980). Effects of media dependencies: Audience
assessment of government. Communication Research, 7, 95-120.
Beeghley, L. (1978). Social stratification in America: A critical analysis o f theory &
research. Santa Monica, CA: Goodyear Publishing Company.
Beentjes, J. W. J., Koolstra, C. M., Marseille, N., & Van der Voort, T. H. A. (2001).
Children's use of difference media: For how long and why? In S. Livingstone
& M. Bovill (Eds.), Children and their changing media environment: A
European comparative study (pp. 85-111). Mahwah, NJ: Lawrence Erlbaum
Associates, Publishers.
Bell, D. (1973). The coming o f post-industrial society: A venture in social
forecasting. New York: Basic Books.
Belson, W. A. (1961). The effects of television on the reading and the buying of
newspapers and magazines. Public Opinion Quarterly, 25(3), 366-381.
Belton, J. (1994). American cinema/American culture. New York: McGraw-Hill.
Bennett, W. L., & Entman, R. M. (Eds.). (2000). Mediated politics: Communication
in the future o f democracy (communication, society and politics). New York:
Cambridge University Press.
Bens, E. D., & Mazzoleni, G. (1998). The media in the age of digital
communication. In D. McQuail & K. Siune (Eds.), Media policy:
Convergence, concentration, & commerce (pp. 165-180). Thousand Oaks:
Sage Publications.
Besley, T., & Case, A. (1993). Modeling technology adoption in developing
countries. The American Economic Review, 83, 396-402.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
173
Biblarz, T. (2002). Model comparison and selection: The BIC sta/Af/c.Unpublished
manuscript, Los Angeles, CA.
Bikson, T. K., & Panis, C. W. A. (1999). Citizens, computers, and connectivity: A
review o f trends. Santa Monica, CA: Rand Corporation.
Bimie, S. A., & Horvath, P. (2002). Psychological predictorsof Internet social
communication. Journal o f Computer-Mediated Communication, 7(4),
Available at http://www.ascusc.org/icmc/vol7/issue4/horvath.html.
Block, E. F., & Mire, E. J. (1996). Trauma on the Internet: Early experience with a
World Wide Web server dedicated to trauma and critical care. The Journal o f
Trauma, 41(2), 265-270.
Blumer, J. G., & Katz, E. (1975). Uses o f mass communications. Beverly Hills, CA:
Sage Publications.
Boer, C. D., & Velthuijsen, A. S. (2001). Participation in conversations about the
news. International Journal o f Public Opinion, 13(2), 140-158.
Bogart, L. (1956). The age o f television: A study o f viewing habits and the impact o f
television on American life. New York: Frederick Ungar Publishing Co.
Bonfadelli, H. (2002). The Internet and knowledge gaps: A theoretical and empirical
investigation. European Journal o f Communication, 77(1), 65-84.
Borgida, E., Sullivan, J. L., Oxendine, A., Jackson, M. S., Riedel, E., & Gangl, A.
(2002). Civic culture meets the digital divide: The role of community
electronic networks. Journal o f Social Issues, 55(1), 125-141.
Bourdieu, P. (1984). Distinction: A social critique o f the judgment o f taste.
Cambridge, MA: Harvard University Press.
Bourdieu, P. (1986). The forms of capital. In J. C. Richardson (Ed.), Handbook o f
theory and research for the sociology o f education (pp. 241-258). Westport,
CT: Greenwood Publishing Group.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
174
Bouwman, H., & Van de Wijngaert, L. (2002). Content and context: An exploration
of the basic characteristics of information needs. New Media & Society, 7(3),
329-353.
Bradley, G. (2000). The information and communication society: How people will
live and work in the new millennium. Ergonomics, 43(1), 844-857.
Bronfenbrenner, U. (1990). Ecology o f human development: Experiments by nature
and design (Reprint ed.). Cambridge, MA: Harvard University Press.
Brown, D., & Bryant, J. (1989). An annotated statistical abstract of communications
media in the United States. In J. L. Salvaggio & J. Bryant (Eds.), Media use
in the information age: Emerging patterns o f adoption and consumer use (pp.
259-302). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.
Burt, R. S. (1992). Structural holes: The social structure o f competition. Cambridge,
MA: Harvard University Press.
Case, D. (1982). Stanford evaluation o f the Green Thumb Box experimental videotext
project for agricultural extension delivery in Shelby and Todd Counties,
Kentucky, Stanford, California. Palo Alto, CA: Institute for Communication
Research, Stanford University.
Castells, M. (1999). The information city is a dual city: Can it be reversed? In D. A.
Schon, B. Sanyal & W. J. Mitchell (Eds.), High technology and low-income
communities (pp. 27-41). Cambridge, MA: Massachusetts Institute of
Technology Press.
Chew, F., & Palmer, S. (1994). Interest, the knowledge gap and television
programming. Journal o f Broadcasting & Electronic Media, 38(3), 271-287.
Cody, M. J., Dunn, D., Hoppin, S., & Wendt, P. (1999). Silver surfers: Training and
evaluating Internet use among older adult learners. Communication
Education, 48, 269-286.
Coffey, S., & Stipp, H. (1997). The interactions between computer and television
usage. Journal o f Advertising Research, 37(2), 61-67.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
175
Coffin, T. E. (1954). Television's impact on society. The American Psychology, 10,
630-641.
Cohen, A. A., Levy, M. R., & Golden, K. (1988). Children's uses and gratifications
of home VCRs: Evolution or revolution. Communication Research, 15(6),
772-780.
Cohen, G., Salomon, I., & Nikjamp, P. (2002). Information-communications
technologies (ICT) and transport: Does knowledge underpin policy?
Telecommunications policy, 26(1/2), 31-52.
Coleman, J. (1998). Social capital in the creation of human capital. American
Journal o f Sociology, 94, 95-121.
Coleman, J., Katz, E., & Menzel, H. (1966). Medical innovation: A diffusion study.
Indianapolis, IN: Bobbs-Merrill.
Colman, W. R. (1990). Health moves to prime time: Evaluating the impact o f a
prime-time television Movie o f the Week on viewers' content-relevant health
beliefs. Unpublished Ph.D. dissertation, University of California, Los
Angeles, Los Angeles.
Community Technology Center Network. Available at http://www.cetnet.org.
Corrado, A., & Firestone, C. M. (Eds.). (1996). Elections in cyberspace: Toward a
new era in American politics. Washington, DC: Aspen Institute.
Crigler, A., Just, M., & Greene, S. B. (2000, August). Connecting with the Internet in
political campaigns: Experiments on race and user satisfaction. Paper
presented at the American Political Association Conference, Washington,
DC.
Czitrom, D. L. (1982). Media and the American mind: From Morse to McLuhan.
Chapel Hill, NC: University of North Carolina Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
176
DeFleur, M. L., & Ball-Rokeach, S. J. (1982). Theories o f mass communication (4th
ed.). New York: Longman.
DeFleur, M. L., & Ball-Rokeach, S. J. (1989). Theories o f mass communication (5th
ed.). New York: Longman.
Deutschmann, P. J., Ellingsworth, H., & McNelly, J. T. (1968). Communication and
social change in Latin America: Introducing new technology. New York:
Praeger.
d'Haenens, L. (2001). Old and new media: Access and ownership in the home. In S.
Livingstone & M. Bovill (Eds.), Children and their changing media
environment: a European comparative study (pp. 53-84). Mahwah, NJ:
Lawrence Erlbaum Associates.
Dillman, D. A. (1985). The social impacts of information technologies in rural North
America. Rural Sociology, 50, 1-26.
Donohue, G. A., Olien, C. N., & Tichenor, P. J. (1987). Media access and knowledge
gaps. Critical Studies in Mass Communication, 7(1), 87-92.
Donohue, G. A., Tichenor, P. J., & Olien, C. N. (1975). Mass media and the
knowledge gap: A hypothesis reconsidered. Communication Research, 2, 3-
23.
Douglas, S. J. (1999). Listening in: Radio and the American imagination, from Amos
T V 7 Andy and Edward R. Murrow to Wolfman Jack and Howard Stern. New
York: Time Books.
Duncan, O. D. (1961). A socioeconomic index for all occupations. In A. J. Reiss
(Ed.), Occupations and social status (pp. 109-138). New York: Free Press.
Dupuits. (2002). The effects of the Internet on pharmaceutical consumers and
providers. Disease Management & Health Outcomes, 10(11), 679-691.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
177
Durkheim, E. (1933/1964). The division o f labor in society (G. Simpson, Trans.).
New York: Free Press of Glencoe.
Dutton, W. H. (1999). Society on the line: Information politics in the digital age.
Oxford and New York: Oxford University Press.
Dutton, W. H., Rogers, E. M., & Jun, S.-H. (1987). Diffusion and social impacts of
personal computers. Communication Research, 14(2), 219-250.
Dutton, W. H., Sweet, P. L., & Rogers, E. M. (1989). Socioeconomic status and the
early diffusion of personal computing in the United States. Social Science
Computer Review, 7(3), 259-271.
Elliott, L. S. (1995). National identity and media system dependency in Belize.
Unpublished Ph.D. dissertation, University of Florida, Gainesville, FL.
Ettema, J. S., & Kline, F. G. (1977). Dificits, differences, and ceilings: Contingent
conditions for understanding the knowlege gap. Communication Research,
4(2), 179-202.
Eveland, W. P. J., & Scheufele, D. A. (2000). Connecting news media use with gaps
in knowledge and participation. Political Communication, 77(3), 215-237.
Feder, G., Just, R. E., & Zilberman, D. (1985). Adoption of agricultural innovations
in developing countries: A survey. Economic Development and Cultural
Change, 33, 255-298.
Finkelstein, J., & Friedman, R. H. (2000). Potential role of telecommunication
technologies in the management of chronic health conditions. Disease
Management & Health Outcomes, 5(2), 57-63.
Fischer, C. S. (1992). America calling: A social history o f the telephone to 1940.
Berkeley, CA: University of California Press.
Fisher, W. (1987). Human communication as narration: Toward a philosophy o f
reason, value, and action. Columbia, SC: University of South Carolina Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
178
Flanagin, A. J., & Metzger, M. J. (2001). Internet use in the contemporary media
environment. Human Communication Research, 27(1), 153-166.
Fong, E., Wellman, B., Kew, M., & Wilkes, R. (2001). Correlates o f the digital
divide: Individual, household and spatial variation (Report to Office of
LeamingTechnologies, Human Resources Development Canada). Toronto,
Canada.
Fry, D. L. (1981). Media dependence and community influentials: A study o f
accessibility, evaluation and cross-checking in one American community.
Unpublished Ph.D. dissertation, Ohio State University, Columbus, OH.
Fulk, J., Schmitz, J., & Steinfield, C. W. (1990). A social influence model of
technology use. In J. Fulk & C. W. Steinfield (Eds.), Organizations and
communication technology (pp. 71-94). Newbury Park, CA: Sage
Publications.
Gaziano, C. (1983). The knowledge gap: An analytic review of media effects.
Communication Research, 10(4), 447-486.
Gaziano, C. (1995, May 21, 1995). A twenty-five-year review o f knowledge gap
research. Paper presented at the American Association for Public Opinion
Research, Fort Lauderdale, FL.
Gaziano, C., & Gaziano, E. (1999). Social control, social change and the knowledge
gap hypothesis. In D. P. Demers & K. Viswanath (Eds.), Mass media, social
control, and social change: A macrosocial perspective (pp. 117-138).
Malvern, IA: Iowa State University Press.
General Social Survey. (2000). Estimated Internet usage differences by basic general
social survey demographics. Available at
www.webuse.umd.edu/handouts/tutorial/SDA-GSS Tutorial.pdf.
Gibbs, J., Ball-Rokeach, S. J., Jung, J.-Y., Kim, Y.-C., & Qiu, J. L. (in press). The
globalization of everyday life: Visions and reality. In M. Sturken, D. Thomas
& S. J. Ball-Rokeach (Eds.), Reinventing technology: cultural narratives o f
technological change. Philadelphia, PA: Temple University Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
179
Goslee, S. (1998). Losing ground bit by bit: Low income communities in the
information age. Washington, DC: Benton Foundation (www.benton.org).
Graham, M. (1999). Handbook on radio and television audience research. London:
BBC World Service Training Trust.
Grant, A. E., Guthrie, K. K., & Ball-Rokeach, S. J. (1991). Television shopping: A
media system dependency perspective. Communication Research, 18(6),
773-793.
Greenberg, B. (1964). Person-to-person communication in the diffusion of news
events. Journalism Quarterly, 41, 489-494.
Haddon, L. (2000). Social exclusion and information and communication
technologies. New Media & Society, 2(4), 387-406.
Halpem, P. (1994). Media dependency and political perceptions in the authoritarian
political system. Journal o f Communication, 44(A), 39-52.
Hargittai, E. (2002). Second-level digital divide: Differences in people's online skills.
First Monday, Peer-reviewed journal o f the Internet, 7(4),
http://firstmondav.org/issues/issue7 4/hargittai/index.html.
Hawkins, R. P., & Pingree, S. (1981). Uniform messages and habitual viewing:
Unnecessary assumptions in social reality effects. Human Communication
Research, 7,291-301.
Hawley, A. H. (1950). Human ecology: A theory o f community structure. New York:
The Ronald Press Company.
Hindman, D. B. (2000). The rural-urban digital divide. Journalism & Mass
Communication Quarterly, 77(3), 549-560.
Hirschburg, P. L., Dillman, D. A., & Ball-Rokeach, S. J. (1986). Media system
dependency theory: Responses to the eruption of Mount St. Helens. In S. J.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
180
Ball-Rokeach & M. G. Cantor (Eds.), Media, audience, and social structure
(pp. 117-126). Newbury Park, CA: Sage Publications.
Hobart, M. E., & Schiffinan, Z. S. (1998). Information ages: Literacy, numeracy,
and the computer revolution. Baltimore, MD: Johns Hopkins University
Press.
Hobday, M. (1994). Technological learning in Singapore: A test case of
leapfrogging. The Journal o f Development Studies, 30, 831-858.
Hoffman, D., & Novak, T. (1996). Marketing in hypermedia computer-mediated
environments: Conceptual foundations. Journal o f Marketing, 60, 50-68.
Hoffman, D., & Novak, T. (2000). The growing digital divide: Implications for an
open research agenda. In B. Kahin & E. Brynjolffson (Eds.), Understanding
the digital economy: Data, tools, and research. Cambridge, MA: MIT Press.
Holbrook, T. M. (2002). Presidential campaigns and the knowledge gap. Political
Communication, 19, 437-454.
Horrigan, J. B., & Rainie, L. (2002). Getting serious online. Washington, DC: Pew
Internet & American Life Project.
Howard, P. E. N., Rainie, L., & Jones, S. (2001). Days and nights on the Internet:
The impact of a diffusing technology. American Behavioral Scientist, 45(3),
383-404.
Johnsson-Smaragdi, U., d'Haenens, L., Krotz, F., & Hasebrink, U. (1998). Patterns
of old and new media use among young people in Flanders, Germany and
Sweden. European Journal o f Communication, 13(A), 479-501.
Joint Venture. (2002). Connecting today's youth with tomorrow's technology careers.
Silicon Valley, CA: Joint Venture.
Jung, J.-Y., Kim, Y.-C., Lin, W.-Y., & Cheong, P. H. (2002, July 15-19). The
influence o f social context on Internet connectedness o f adolescents in Seoul,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
181
Singapore and Taipei. Paper presented at the International Communication
Association, Seoul, Korea.
Jung, J.-Y., Qiu, J. L., & Kim, Y.-C. (2001). Internet connectedness and inequality:
Beyond the "divide". Communication Research, 25(4), 507-535.
Katz, E., Blumer, J. G., & Gurevitch, M. (1973-1974). Uses and gratifications
research. Public Opinion Quarterly, 57(4), 509-523.
Katz, E., Gurevitch, M., & Haas, H. (1973). On the use of mass media for important
things. American Sociological Review, 38, 164-181.
Katz, E., & Lazarsfeld, P. F. (1955). Personal influence: The part played by people
in the flow o f mass communications. Glencoe, IL: Free Press.
Kayany, J. M., & Yelsma, P. (2000). Displacement effects of online media in the
socio-technical context of households. Journal o f Broadcasting & Electronic
Media, 44(2), 215-229.
Kiesler, S., Zdaniuk, B., Lundmark, V., & Kraut, R. (2000). Troubles with the
Internet: The dynamics of help at home. Homan-Computer Interaction, 15,
323-351.
Kieve, J. (1973). The electric telegraph: A social and economic history. New York:
Barnes & Noble.
Kim, P. (2001). New media, old ideas: The organizing ideology of interactive TV.
Journal o f Communication Inquiry, 25(1), 72-88.
Kim, U. (1982). Cultural dependency and political instability: A case study on South
Korean mass media dependency upon the United States, 1970-1979.
Unpublished Ph.D. dissertation, University of Cincinnati, Cincinnati.
Kim, Y.-C., & Jung, J.-Y. (2001, May 24-28). Perceptions ofY2K Risks and
Individual Communication Connectedness. Paper presented at the
International Communication Association Convention, Washington, DC.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
182
Kim, Y.-C., & Jung, J.-Y. (2002, August 15-18). "Geo-ethnicity" and neighborhood
engagement: A communication infrastructure perspective. Paper presented at
the American Sociological Association Convention, Chicago, IL.
Kim, Y.-C., Jung, J.-Y., Ball-Rokeach, S. J., & Loges, W. E. (2002, July 15-19).
Ethnicity, place, and communication technology: Geo-ethnic effect on multi
dimensional Internet Connectedness in urban communities. Paper presented
at the International Communication Association Convention, Seoul, Korea.
Kraut, R., Kiesler, S., Boneva, B., Cummings, J., Helgeson, V., & Crawford, A.
(2002). Internet paradox revisited. Journal o f Social Issues, 58, 49-74.
Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukhopadhyay, T., & Scherlis,
W. (1998). Internet paradox: A social technology that reduces social
involvement and psychological well-being? American Psychologist, 55(9),
1017-1032.
Kraut, R., Rice, R. E., Cool, C., & Fish, R. S. (1998). Varieties of social influence:
The role of utility and norms in the success of a new communication
medium. Organization Science, 9(4), 437-453.
Kwak, N. (1999). Revisiting the knowledge gap hypothesis. Communication
Research, 26(4), 385-413.
Kyle Robison, K., & Crenshaw, E. M. (2002). Post-industrial transformations and
cyber-space: A cross-national analysis of internet development. Social
Science Research, 31(3), 334-363.
LaRose, R., Eastin, M. S., & Gregg, J. (2001). Reformulating the Internet paradox:
Social cognitive explanations of Internet use and depression. Journal o f
Online Behavior, 1(2), Available at
http://www.behavior.net/JOB/vlnl/paradox.html.
Lauf, E. (2001). Research note: The vanishing young reader: Sociodemographic
determinants of newspaper use as a source of political information in Europe,
1980-98. European Journal o f Communication, 16(2), 233-243.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
183
Lee, W., & Chan, B. (2001, 14-15 September). Internet as a facilitator o f social
relationship: A study o f young Singaporeans' online and offline friendship.
Paper presented at the Internet and Development in Asia, Internet Political
Economy Forum, Singapore.
Lee, W., & Kuo, E. C. Y. (2002). Internet and displacement effect: Children's media
use and activities in Singapore. Journal o f Computer-Mediated
Communication, 7(2), Available at
http://wAvw.ascusc.org/icmc/vol7/issue2/singapore.html.
Lenhart, A. (2000). Who's not online: 57% o f those without Internet access say they
do not plan to log on. Washington, DC: Pew Internet & American Life
Project.
Lieberman, M. A., Golant, M., Giese-Davis, J., Winzlenberg, A., Benjamin, H.,
Humphreys, K., et al. (2003). Electronic support groups for breast carcinoma.
Cancer, 97(4), 920-925.
Lievrouw, L. A. (2000). The information environment and universal service. The
Information Society, 16, 155-159.
Lin, C. A. (1998). Exploring personal computer adoption dynamics. Journal o f
Broadcasting and Electronic Media, 42, 95-112.
Lin, C. A., & Jeffres, L. W. (1998). Factors influencing the adoption of multimedia
cable technology. Journalism &Mass Communication Quaterly, 75(2), 341-
352.
Littlejohn, S. W. (2001). Theories o f human communication (7th ed.). Florence, KY:
Thomson Learning.
Livingstone, S. (1998). Mediated childhoods: A comparative approach to young
people's changing media environment in Europe. European Journal o f
Communication, 13(4), 435-456.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
184
Livingstone, S., & Bovill, M. (2001). Children and their changing media
environment: A European comparative study. Mahwah, NJ: Lawrence
Erlbaum Associates, Publishers.
Loges, W. E. (1992). Canaries in the coal mine: Perceptions o f threat and media
system dependency relations. Unpublished Ph.D. dissertation, University of
Southern California, Los Angeles.
Loges, W. E. (1994). Canaries in the coal mine: Perceptions of threat and media
system dependency relations. Communication Research, 21(1), 5-23.
Loges, W. E., & Ball-Rokeach, S. J. (1993). Dependency relations and newspaper
readership. Journalism Quarterly, 70, 602-614.
Loges, W. E., & Jung, J.-Y. (2001). Exploring the digital divide: Internet
connectedness and age. Communication Research, 28(4), 536-562.
Loury, G. C. (1997). A dynamic theory of racial income differences. In P. A.
Wallace & A. H. LaMond (Eds.), Women, minorities, and employment
discrimination (pp. 153-186). Lexington, MA: D. C. Heath.
Lubrano, A. (1997). The telegraph: How technology innovation caused social
change. New York: Garland Publications.
Maksabedian, J. (1980). The social process in innovation and technological change.
Revista Latinoamericana de Psicologia, 72(1), 109-117.
Malone, T. W., Yates, J., & Benjamin, R. I. (1987). Electronic markets and
electronic hierarchies. Communications o f the ACM, 30(6), 484-497.
Manross, G. G. (1987). Confirming media effects in voting behavior: A media
dependency perspective. Unpublished Ph.D. dissertation, University of
Southern California, Los Angeles.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
185
Marvin, C. (1988). When old technologies were new: Thinking about
communications in the late nineteenth century. Oxford: Oxford University
Press.
Matei, S., & Ball-Rokeach, S. J. (2001). Real and virtual social ties: Connections in
the everyday lives of seven ethnic neighborhoods. American Behavioral
Scientist, 45(3), 550-564.
Matei, S., Ball-Rokeach, S. J., Wilson, M. E., Gibbs, J., & Gutierrez Hoyt, E. (2001).
Metamorphosis: A field research methodology for studying communication
technology and community. Electronic Journal o f Communication, 11(2).
Matzger, G. D. (1983). Cable television audiences. Journal o f Advertising Research,
23(4), 41-47.
Mazmanian, D. A. (1995). Issues in telecommunications and democracy
(Communications Policy Working Paper). Washington, DC: Benton
Foundation (www.benton.orgh
McQuail, D. (1997). Audience Analysis. Thousand Oaks, CA: Sage Publications.
McQuail, D. (2000). Mcquail's mass communication theory (4th ed.). Thousand
Oaks: Sage Publications.
Merskin, D. L. (1993). Partner-seeking through newspaper personal advertisements:
A study o f media dependency and adoption o f innovation. Unpublished Ph.D.
dissertation, Syracuse University, Syracuse, NY.
Midgley, J. (1984). Diffusion and the development of social policy: Evidence from
the third world. Journal o f Social Policy, 13, 167-184.
Morino Institute. (1995). Unlocking the doors to opportunity. Washington, DC:
Morino Institute (www.morino.org).
Morino Institute. (1999). Statement o f the Morino Institute on "Bridging the digital
divide". Washington, DC: Morino Institute.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
186
Morino Institute. (2001). From access to outcomes: raising the aspirations for
technology initiatives in low-income communities. Washington, DC: Morino
Institute.
Morton, T. A., & Duck, J. M. (2000). Social identity and media dependency in the
gay community: The prediction of safe sex attitudes. Communication
Research, 27(4), 38-61.
Morton, T. A., & Duck, J. M. (2001). Communication and health beliefs: Mass and
interpersonal influences on perceptions of risk to self and others.
Communication Research, 25(5), 602-626.
Moy, P., Scheufele, D., & Holbert, R. (1999). Television use and social capital:
Testing Putnam's time displacement hypothesis. Mass Communication and
Society, 2, 27-45.
Mullan, B. (1997). Consuming television: Television and its audience. Cambridge,
MA: Blackwell.
Mulugetta, Y. M. (1987). Knowledge gap hypothesis and media preference.
Unpublished Ph.D. dissertation, University of Wisconsin, Madison.
Murdock, G., Hartmann, P., & Gray, P. (1992). Contextualizing home computing
resources and practices. In R. Silverstone & E. Hirsch (Eds.), Consuming
technologies: Media and information in domestic spaces (pp. 146-160).
London and New York: Routledge.
Negroponte, N. (1995). Being digital. London: Hodder & Stoughton.
Neu, C. R., Anderson, R. H., & Bikson, T. K. (1998). E-mail communication
between government and citizens: Security, policy issues, and next steps
(Issue paper). Santa Monica, CA: Rand Science and Technology.
Neuman, W. R. (1991). The future o f the mass audience. New York: Cambridge
University Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
187
Newhagen, J., & Bucy, E. (2001, November 16-17. 2001). Routes to media access:
Apprehending Internet Content. Paper presented at the IAMCR/ICA
Conference on Digital Divide, Austin, TX.
Nie, N. H., & Erbring, L. (2000). Internet and society: A preliminary report. Palo
Alto, CA: Stanford Institute for the Quantitative study of Society, Stanford
University.
Nie, N. H., & Hillygus, D. S. (2002). The impact of Internet use on sociability:
Time-diary findings. IT & Societey, 7(1), 1-20.
Nigg, J. M. (1982). Communication under conditions of uncertainty: Understanding
earthquake forecasting. Journal o f Communication, 32, 27-36.
Norris, P. (1996). Did television erode social capital? A reply to Putnam’ PS.
Political Science and Politics, XXIXif), 474-480.
Norris, P. (2001). Digital divide: Civic engagement, information poverty, and the
Internet worldwide. New York: Cambridge University Press.
Novak, T., & Hoffman, D. (1998, April 17). Information access: Bridging the racial
divide on the Internet. Science, 280, 390.
Nye, D. E. (1994). American technological sublime. Cambridge, MA: The MIT
Press.
Okolica, C., & Stewart, C. M. (1996). Factors influencing the use of voice messaging
technology: Voice mail implementation in a corporate setting. Central
Business Review, 75(1), 52-59.
Park, R. E. (1922). The immigrant press and its control. New York & London:
Harper & Brothers.
Park, R. E., Burgess, E. W., & McKenzie, R. D. (1925). The city. Chicago: The
University of Chicago Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
188
Patterson, R., & Wilson III, E. J. (2000). New IT and social inequality: Resetting the
research and policy agenda. The Information Society, 16, 77-86.
Peiser, W. (1999). The television generation's relation to the mass media in
Germany: Accounting for the impact of private television. Journal o f
Broadcasting & Electronic Media, 43(3), 364-385.
Perse, E. M., & Dunn, D. G. (1998). The utility of home computers and media use:
Implications of multimedia and connectivity. Journal o f Broadcasting and
Electronic Media, 42(4), 435-456.
Petrie, D., & Willis, J. (Eds.). (1995). Television and the household: Reports from
the BFI's audience tracking study. London: BFI Publisher.
Pew Internet and American Project. Available at http://www.pewintemet.org.
Power, J. G. (1995). Media dependency, bubonic plague, and the social constmction
of the Chinese other. Journal o f Communication Inquiry, 19(1), 89-111.
Price, V. (1992). Public opinion. Newbury Park, CA: Sage Publications.
Pmijt, H. (2002). Social capital and the equalizing potential of the Internet. Social
Science Computer Review, 20(2), 109-115.
Putnam, R. (1995). Bowling alone: America's declining social capital. Journal o f
Democracy, 6(1), 65-78.
Putnam, R. (2000). Bowling alone: The collapse and revival o f American community.
New York: Simon & Schuster.
Raftery, A. E. (1995). Bayesian model selection in social research. In P. V. Marsden
(Ed.), Sociological methodology (Vol. 25, pp. 111-163). Washington, DC:
American Sociological Association.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
189
Rainie, L., & Packel, D. (2001). More online, doing more: 16 million newcomers
gain Internet access in the last half o f2000 as women, minorities, and
families with modest incomes continue to surge online. Washington, DC: Pew
Internet and American Life Project.
Rand Corporation. Available at http://www.rand.org.
Rios, D. I., & Gaines Jr., S. O. (1998). Latino media use for cultural maintenance.
Journalism &Mass Communication Quarterly, 75(4), 746-761.
Robinson, J. P., Kestnbaum, M., & Neustadtl, A. (Eds.). (2002). The Internet and
other uses o f time. Oxford, UK and New York: Basil Blackwell.
Rodriguez, M. L. A. (1982). What people get from the news: How Filipinos
understand agrarian reform. Unpublished Ph.D. dissertation, The University
of Wisconsin, Madison, Madison.
Rogers, E. (1983). Diffusion o f innovations (3rd ed.). New York: Free Press.
Rogers, E. (1995). Diffusion o f innovations (4th ed.). New York: Free Press.
Schmitz, J., & Fulk, J. (1991). Organizational colleagues, information richness, and
electronic mail: A test of the social influence model of technology use.
Communication Research, 18, 487-523.
Shah, D. V., Kwak, N., & Holbert, R. L. (2001). "Connecting" and "disconnecting"
with civic life: patterns of Internet use and the production of social capital.
Political Communication, 18, 141-162.
Shah, D. V., McLeod, J. M., & Yoon, S. H. (2001). Communication, context, and
community: An exploration of print, broadcast, and Internet influences.
Communication Research, 25(4), 464-506.
Silverstone, R., & Hirsch, E. (Eds.). (1992). Consuming technologies: Media and
information in domestic spaces. London & New York: Routledge.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
190
Sklar, R. (1994). Movie-made America: A cultural history o f American movies. New
York: Vintage Books.
Skumanich, S. A., & Kintsfather, D. P. (1998). Individual media dependency
relations within television shopping programming: A causal model reviewed
and revised. Communication Research, 25(2), 200-220.
Smith, A., & Paterson, R. (Eds.). (1998). Television: An international history (2nd
ed.). Oxford: Oxford University Press.
Speight, K. (1999). Gaps in the worldwide information explosion: How the Internet
is affecting the worldwide knowledge gap. Telematics and Informatics, 16(3),
135-149.
Speter, K. M. (1996). A paradigm for change: The diffusion o f innovation theory and
the cellular industry. Unpublished Ph.D. dissertation, Pace University, New
York.
Spooner, T. (2001). Asian-Americans and the Internet: The young and the connected.
Washington, DC: Pew Internet & American Life Project.
Spooner, T., & Rainie, L. (2000). African-Americans and the Internet. Washington,
DC: Pew Internet and American Project.
Steinfield, C. W., Dutton, W. H., & Kovaric, P. (1989). A framework and agenda for
research on computing in the home. In J. L. Salvaggio & J. Bryant (Eds.),
Media use in the information age: Emerging patterns o f adoption and
consumer use (pp. 61-86). Hillsdale, NJ: Lawrence Erlbaum & Associates,
Publishers.
Stone, G. D. (2002). Both sides now: Fallacies in the genetic-modification wars,
implications for developing countries, and anthrological perspectives.
Current Anthropology, 43(A), 611-630.
Suoninen, A. (2001). The role of media in peer group relations. In S. Livingstone &
M. Bovill (Eds.), Children and their changing media environment: A
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
191
European comparative study (pp. 201-219). Mahwah, NJ: Lawrence Erlbaum
Associates, Publishers.
Suss, D. (2001). Computers and the Internet in school: Closing the knowledge gap?
In S. Livingstone & M. Bovill (Eds.), Children and their changing media
environment: A Eurpean comparative study (pp. 221-241). Mahwah, NJ:
Lawrence Erlbaum Associates.
Taylor, D. S. E. (1991). Application o f the uses and dependency model o f mass
communication to development communication in the western area o f Sierra
Leone. Unpublished Ph.D. dissertation, Kent State University, Kent, OH.
The Benton Foundation. Available at http://www.benton.org/Resources/home.html.
The Metamorphosis Project. (2003a). Los Angeles-area community technology
centers: Successes and challenges on the road to community building (White
Paper). Los Angeles, CA: Communication Technology and Community
Program, Annenberg School for Communication, University of Southern
California.
The Metamorphosis Project. (2003b). The well-connected community:
A new tool for planning to use the Internet to build community (White Paper). Los
Angeles, CA: Communication Technology and Community Program,
Annenberg School for Communication, University of Southern California.
Thierer, A. D. (2000). How free computers are filling the digital divide
(Backgrounder). Washington, DC: The Heritage Foundation.
Tichenor, P. J., Donohue, G., & Olien, C. (1980). Community conflict and the press.
Beverly Hills, CA: Sage Publications.
Tichenor, P. J., Donohue, G. A., & Olien, C. N. (1970). Mass media flow and
differential growth in knowledge. Public Opinion Quarterly, 34(2), 159-170.
Toffler, A. (1980). The third wave. New York: Morrow.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
192
Turner, R. H., & Paz, D. H. (1986). The mass media in earthquake warning. In S. J.
Ball-Rokeach & M. G. Cantor (Eds.), Media, audience, and social structure
(pp. 99-115). Newbury Park, CA: Sage Publications.
U.S. Census Bureau. (2000). United States Census 2000. Washington, D.C.: United
States Department of Commerce.
UCLA Center for Communication Policy. (2000). The UCLA Internet report:
Surveying the digital future. Los Angeles: University of California, Los
Angeles.
UCLA Center for Communication Policy. (2001). The UCLA Internet Report 2001:
Surveying the digital future. Los Angeles: University of California, Los
Angeles.
UCLA Center for Communication Policy. (2003). The UCLA Internet Report 2001:
Surveying the digital future, Year Three. Los Angeles: University of
California, Los Angeles.
US Department of Commerce. (1995). Falling through the net: A survey o f the "have
nots" in rural and urban America. Washington, DC: U.S. Department of
Commerce.
US Department of Commerce. (1998). Falling through the Net II: New data on the
digital divide. Washington, DC: US Department of Commerce.
US Department of Commerce. (1999). Falling through the net: Defining the digital
divide. Washington, DC: US Department of Commerce.
US Department of Commerce. (2000). Falling through the net: Toward digital
inclusion. Washington, DC: Department of Commerce.
US Department of Commerce. (2002). A nation online: How Americans are
expanding their use o f the Internet. Washington, DC: Department of
Commerce.
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
193
Van Dijk, J. A. G. M. (1997). Widening information gaps and policies o f prevention-
1 (Advice to the Information Society Forum of the European Commission).
Van Eijck, K., & Van Rees, K. (2000). Media orientation and media use: Television
viewing behavior of specific reader types from 1975 to 1995. C o m m u n ic a tio n
Research, 27(5), 574-616.
Van Slyke, C. (1997). The diffusion o f technology cluster innovations: The case o f
the Internet.Unpublished manuscript, Tampa, Florida.
Viswanath, K., & Finnegan Jr., J. R. (1996). The knowledge gap hypothesis:
Twenty-five years later. In B. R. Burleson (Ed.), Communication Yearbook
(Vol. 19, pp. 187-227). Thousand Oaks, CA: Sage Publications.
Waring, H. R. (1996). Media system dependency and identity: The development o f
America's gay and lesbian alternative media and the transformation o f
homosexuality. Unpublished Ph.D. dissertation, University of Southern
California, Los Angeles.
Webster, J. G. (1986). Audience behavior in the new media environment. Journal o f
Communication, 26(3), 77-91.
Wellman, B., Hasse, A. W., Witte, J., & Hampton, K. (2001). Does the Internet
increase, decrease, or supplement social capital?: Social networks,
participation, and community commitment. American Behavioral Scientist,
45(3), 436-455.
West, P. C. (1983). Collective adoption of natural resource practices in developing
nations. Rural Sociology, 48, 44-59.
White, J., & Weatherall, A. (2000). A grounded theory analysis of older adults and
information technology. Educational Gerontology, 26, 371-386.
Wilhelm, A. (2000). Democracy in the digital age. New York: Routledge.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
194
Williams, F., Rice, R. E., & Rogers, E. M. (1986). Research methods and the new
media. New York: Free Press.
Williams, F., Strover, S., & Grant, A. E. (1994). Social aspects of new media
technologies. In J. Bryant & D. Zillman (Eds.), Media effects: Advances in
theory and research (pp. 463-482). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Wolff, E. N. (2001, February 12). The rich get richer: And why the poor don't. The
American Prospect, 12, Available at
http://www.prospect.Org/print/V 12/13/wolff-e.html.
Xu, B. (2000). Multinational enterprises, technology diffusion, and host country
productivity growth. Journal o f Development Economics, 62(2), 477-493.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Asset Metadata

Creator Jung, Joo-Young Janice (author)

Core Title Internet connectedness and its social origins: An ecological approach to communication media and social inequality

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Doctor of Philosophy

Degree Program Communication

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag information science,mass communications,OAI-PMH Harvest,sociology, general

Language English

Advisor Ball-Rokeach, Sandra (committee chair), Biblarz, Timothy (committee member), McLaughlin, Margaret (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-367213

Unique identifier UC11339368

Identifier 3103911.pdf (filename),usctheses-c16-367213 (legacy record id)

Legacy Identifier 3103911.pdf

Dmrecord 367213

Document Type Dissertation

Rights Jung, Joo-Young Janice

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA