VALUE PREDISPOSITIONS, COMMUNICATION, AND ATTITUDES TOWARD
NANOTECHNOLOGY: THE INTERPLAY OF PUBLIC AND EXPERTS
by
Shirley S. Ho
A dissertation submitted in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy
(Mass Communications)
at the
UNIVERSITY OF WISCONSIN-MADISON
2008
© Copyright by Shirley S. Ho 2008
All Rights Reserved
i
ACKNOWLEDGMENTS
As my graduate student life in Madison is approaching an end with the completion of
this dissertation, I would like to acknowledge all of the people who have created a memorable
experience for me in my four-and-a-half years of grad school, even if no words can fully express
my sincere gratitude.
First and foremost, I would like to extend my heartfelt gratitude to my Ph.D. advisor, Dr.
Dietram A. Scheufele, for his support and encouragement throughout my doctoral program.
Admittedly a high-maintenance student, I would like to thank him for the extraordinary
amount of advising effort and time that he has set aside for me. I am eternally grateful to his
patience and dedication in unflaggingly answering all my persistent questions and inquiries.
His invaluable guidance and scholarship has, no doubt, inspired and expedited my intellectual
growth. The priceless experience of being his teaching assistant for J658 has also prepared me to
become a more confident teacher and researcher. He has helped to cultivate talents that I never
knew I had. Needless to say, I have been very fortunate to have learnt from and worked with
such an amazingly talented, sincere, and nurturing scholar.
The same gratitude goes to Dr. Dominique Brossard, who has played a pivotal role in
my graduate career. Being one of the warmest and most caring professors I have known, she
has always said the most encouraging words to me whenever I faced academic frustration. I
would like to thank her for being such a wonderful teacher, co-worker, committee member, and
friend. I am forever indebted to her for introducing me to the area of science and risk
communication, and her insightful and stimulating comments for many of my research projects.
The rest of my committee members have also given me invaluable guidance over the
years. I would like to thank Dr. Douglas M. McLeod, who was also my master’s program
ii
advisor, for introducing me to communication research and providing expert advice for my
research. I would like to thank Dr. Albert C. Gunther who had given me thought-provoking
theoretical guidance both in my dissertation and my other research projects. I would also like to
thank Dr. David Kaplan, my minor advisor in educational psychology, for introducing me to
the world of structural equation modeling, and my many initially agonizing weekends with
MPlus.
In addition to my committee members, I would like to thank Dr. Sharon Dunwoody and
Dr. Dhavan V. Shah, for the numerous enjoyable and mentally stimulating courses that I have
taken with them. My many thanks also go to the following MSRG folks whom I shared
studying experiences and/ or worked closely with in other research projects: Andrew R. Binder,
Rosalyna Wijaya, Tsung-Jen Shih, Eunkung Kim, Michael F. Dahlstrom, Kajsa Dalrymple,
Anthony Dudo. I especially thank Rosalyna who has been an awesome friend and a great
housemate whom I shared my daily laughter, frustrations, and joyous moments with. I would
like to express my special thanks to Andy for commenting, editing and proofreading this
dissertation, and for being such a motivating research buddy and an excellent LISREL resource
in the last couple of years. I would also like to thank Anthony for offering his help with the
“VantagePoint” news analysis in this dissertation. I would like to thank the following friends
who kindly provided support and help at different stages of my memorable years in Madison:
Marcus Ong, Hsaowei Yap, Kyurim Kyoung, and Erika Gunadi. I will definitely miss the crazy
Halloween parties, the snow, and all of the interesting characters in charming small town
Madison.
Several mentors in my home country, Singapore, deserve thanks for their indirect
contributions to this dissertation. I would like to thank Dr. Xiaoming Hao, Dr. Peng Hwa Ang,
and Dr. Benjamin H. Detenber at the Wee Kim Wee School of Communication and Information,
iii
Nanyang Technological University (NTU), Singapore, for offering me valuable professional
advice for the past ten years. In addition, I would like to express my gratitude for NTU for
awarding me a generous four-year fellowship, and the School of Journalism and Mass
Communication at UW-Madison, for offering me a one-year financial support during the course
of my M.A. and Ph.D. studies. I was able to devote full concentration to my research and
studies, thanks to the financial support from these institutions.
This dissertation would not be completed without the encouragement from my family.
Special thanks go to my parents James Ho and Lily Wang, and my sisters, Susan Ho, Jean Ho,
and Heather Ho, for all these years of support.
Last but certainly not least, I would like to thank and dedicate this dissertation to my
husband and best friend, Chun Hong Low. He has shown extraordinary patience by sustaining
an arduous long-distance relationship with me. I sincerely thank him for the love, support, and
encouragement that he has unconditionally and selflessly given me. I am definitely looking
forward to building a good life together and spending the many happy days ahead with him in
Singapore.
Thank you very much everyone!
iv
ABSTRACT
VALUE PREDISPOSITIONS, COMMUNICATION, AND ATTITUDES TOWARD
NANOTECHNOLOGY: THE INTERPLAY OF PUBLIC AND EXPERTS
Shirley S. Ho
Under the supervision of Professor Dietram A. Scheufele
at the University of Wisconsin-Madison
Almost three decades of research in public understanding of science have provided
inconclusive evidence about the “scientific literacy model” (Miller, Pardo, & Niwa, 1997),
triggering critics to advocate the “cognitive miser model” (Fiske & Taylor, 1991) as an
alternative to explain how the general public make decisions about such emerging science as
nanotechnology. Is the scientific literacy model obsolete? What are the specific mechanisms
through which the public develop their opinions about the risks and benefits of nanotechnology
and their level of support for funding of the emerging technology? How do the public make
sense of the ample amount of scientific information in the mass media? These unresolved issues
call for a development of a more theory-based and complex causal model examination of the
mechanisms behind public decision-making of emerging technology. In addition, are there
differences between expert and public opinion about nanotechnology? Is expert opinion more
objective than public opinion when it comes to decision-making about the emerging
technology? Will the mass media have a differential impact on expert and laypersons’ attitudes
toward nanotechnology? This dissertation endeavors to answer these questions in a series of
two studies.
v
Using a nationally representative survey data of 1,015 adults conducted in the United
States in 2007, Study 1 draws on the “differential gains model” (Scheufele, 2001, 2002) and the
“cognitive mediation model” (Eveland, 2001, 2002) as theoretical frameworks to examine the
moderating and mediating mechanisms through which cognitive and heuristic factors influence
public perceived risks-versus-benefits of nanotechnology and their level of support for federal
funding of nanotechnology. Findings from the regression analyses indicate that the public
primarily use value predispositions (i.e., religious beliefs, deference to scientific authority, and
trust in scientists) and positive news frames from the mass media as heuristic cues to make
judgments about risks-versus-benefits and support for federal funding of nanotechnology.
Conversely, factual scientific knowledge is demonstrated to play a significant, but minor role in
influencing perceived risks-versus-benefits. In addition, reflective integration in the form of
elaborative processing had a significant negative influence on perceived risks-versus-benefits.
In line with the differential gains model, the results of Study 1 indicate that the influence
of science media use on the attitudinal outcomes (i.e., perceived risks-versus-benefits and
support for federal funding of nanotechnology) were moderated by elaborative processing. The
significant interactions suggest that people tended to rely on new scientific information
gathered from the mass media to form attitudes toward nanotechnology. The effect from the
media was heightened when people paid attention to science news in the media and reflected
upon the messages they received. By applying the cognitive mediation model as the second
theoretical framework, the structural equation model in Study 1 also reveals an informational
pathway and a heuristic pathway through which the mass media directly and indirectly exert
its influence on public attitudes toward nanotechnology. Taken together, these findings bridged
the disconnection between the differential gains model and the cognitive mediation model by
underscoring the simultaneous moderating and mediating roles of reflective integration when it
vi
comes to understanding how the mass media differentially influence individuals’ attitudes
toward emerging technologies. Notably, the findings also suggest that the scientific literacy
model and the cognitive miser model are two parallel, complementary processes that
individuals use to form opinions about nanotechnology.
Next, using a mail survey of 363 nanotechnology scientists and engineers conducted
between May and June 2007 in the U.S., Study 2 compares public and experts’ attitudes toward
nanotechnology and addresses the pertinent question of whether experts are indeed more
objective in their judgment of nanotechnology than do the public. The regression analyses
provide partial support for the hypotheses regarding the impact of scientific status (i.e., experts
versus lay public) on perceived risks-versus-benefits of nanotechnology. First, compared with
the experts, the results demonstrate that the public judged nanotechnology as having more risks
and lesser benefits, after controlling for all appropriate exogenous factors. Second, the findings
show that experts, equipped with their professional training and experience, used relatively less
heuristic cues such as religious guidance, to make risks-versus-benefits judgment of
nanotechnology than did the public.
On the other hand, the findings provide strong support for the hypotheses on the effect
of scientific status on support for federal funding of nanotechnology, in which the experts
indicated greater support for federal funding of the emerging technology than did the public.
Moreover, the results demonstrate that the experts draw on significantly less heuristic cues in
the form of value predispositions and science media frames to make decision about funding
support for nanotechnology than did the public. In sum, these findings suggest that the experts
are in a position to independently assess risks and benefits, and indicate that experts and the
public use different considerations to make judgments about risks and benefits, and level of
vii
funding support for the emerging technology. The theoretical and practical implications of the
results from Study 1 and Study 2 were discussed.
viii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS
ABSTRACT
TABLE OF CONTENTS
Page
i
iv
viii
LIST OF FIGURES
xii
LIST OF TABLES
xiv
CHAPTER 1 INTRODUCTION
1
1.1. Unresolved Issues in Understanding How Public Form Opinions about
Nanotechnology
1.1.1. Ongoing Debate between the “Scientific Literacy Model” and the
“Cognitive Miser Model”
1.1.2. The Role of Communication Processes: Bridging the “Differential
Gains Model” and the “Cognitive Mediation Model”
3
1.2. Unresolved Issues in Public versus Expert Opinion of Nanotechnology
1.3. Objectives, Originality, and Significance
1.3.1. Study One – Effects of Value Predispositions, Mass Media, and
Cognitive Processing on Public Attitudes toward Nanotechnology:
Testing Moderating and Mediating Mechanisms
1.3.2. Study Two – Experts versus Public Attitudes toward Nanotechnology
8
9
9
11
1.4. Organization of the Chapters
12
CHAPTER 2 NANOTECHNOLOGY, MASS MEDIA, AND PUBLIC OPINION
2.1. Introduction to Nanotechnology
2.1.1. Nanotechnology and Its Societal Implications
2.1.2. Nanotechnology and Its Implications for Economy and
Governmental Policy
2.1.3. Why Public Opinion Matters
2.1.4. Mass Media and Public Opinion
2.2. Public Opinion Trends
2.3. Media Coverage of Nanotechnology over the Years
3
5
16
16
17
18
21
22
25
28
ix
CHAPTER 3 EFFECTS OF VALUE PREDISPOSITIONS, MASS MEDIA, AND
COGNITIVE PROCESSING ON PUBLIC ATTITUDES TOWARD
NANOTECHNOLOGY: TESTING MODERATING AND MEDIATING
MECHANISMS (STUDY 1)
32
3.1. Outcome Variables
3.1.1. Perceived Risks-versus-Benefits (Attitudinal Outcome Variable 1)
3.1.2. Support for Federal Funding of Nanotechnology (Attitudinal
Outcome Variable 2)
33
33
35
3.2. Value Predispositions as Heuristic Cues in Opinion Formation
3.2.1. The Role of Religious Beliefs
3.2.2. Deference to Scientific Authority as Heuristic Shortcut
36
36
38
3.3. Science Media Use and Opinion Formation
3.4. Cognitive Processing: Reflective Integration and Learning from the News
Media
3.4.1. Effect of Elaborative Processing on Cognitive and Attitudinal
Outcomes
3.4.2. Effect of Interpersonal Discussion on Cognitive and Attitudinal
Outcomes
39
43
3.5. The “Differential Gains Model” – Moderating Role of Reflective Integration
on Cognitive and Attitudinal Outcomes
3.6. The “Cognitive Mediation Model” – Mediating Role of Reflective
Integration on Cognitive and Attitudinal Outcomes
3.7. Effect of Factual Scientific Knowledge on Attitudinal Outcomes
3.8. The Role of Trust in Scientists
3.8.1. Mediating Role of Trust in Scientists
49
3.9. Effects of Perceived Risks-versus-Benefits on Support for Federal
Funding of Nanotechnology
59
46
47
52
54
56
58
CHAPTER 4 METHODS AND RESULTS (STUDY 1)
61
4.1. Methods
4.1.1. Data and Sampling
4.1.2. Measures
4.1.2.1. Attitudinal Outcome Variables
4.1.2.2. Independent Variables
4.1.2.3. Control Variables
4.1.3. Analytic Strategies
4.1.3.1. Missing Values Treatment
4.1.3.2. Ordinary Regression Analysis
4.1.3.3. Structural Equation Modeling
61
61
62
62
64
66
67
67
69
71
x
4.2. Results
4.2.1. Direct and Moderating Relationships
4.2.2. Direct and Mediating Relationships
CHAPTER 5 DISCUSSION (STUDY 1)
76
76
80
87
5.1. Explanations for Findings on Direct Effects
5.2. Explanations for Findings on Moderating Mechanisms
5.3. Explanations for Findings on Mediating Mechanisms
5.4. Implications
88
92
94
96
CHAPTER 6 EXPERTS VERSUS PUBLIC ATTITUDES TOWARD
NANOTECHNOLOGY (STUDY 2)
100
6.1. Differences in Expert and Public Judgments of Risk
6.2. Expert and Public Differences in Levels of Support for Federal Funding of
Nanotechnology
6.3. Factors that Influence Perceived Risks-versus-Benefits and Support for
Federal Funding of Nanotechnology
CHAPTER 7 METHODS AND RESULTS (STUDY 2)
101
106
107
110
7.1. Methods
7.1.1. Measures
7.1.1.1. Outcome Variables
7.1.1.2. Independent Variables
7.1.1.3. Control Variables
7.1.2. Analytical Approach
110
111
112
115
117
117
7.2. Results
7.2.1. Experts versus Public: Factors Predicting Perceived Risks-versusBenefits of Nanotechnology
7.2.2. Experts versus Public: Factors Predicting Support for Federal Funding
of Nanotechnology
118
119
CHAPTER 8 DISCUSSION (STUDY 2)
8.1. Explanations for Experts and Public Differences in Perceived Risks-versusBenefits
8.2. Explanations for Experts and Public Differences in Support for Federal
Funding of Nanotechnology
8.3. Implications
121
125
126
129
130
xi
CHAPTER 9 OVERALL DISCUSSION AND CONCLUSION
132
9.1. Summary
9.2. Major Theoretical, Conceptual, and Practical Contributions
9.3. Limitations and Directions for Future Research
9.3.1. Study One
9.3.2. Study Two
132
134
140
140
142
9.4. Conclusion
144
BIBLIOGRAPHY
146
FIGURES
166
TABLES
192
APPENDIX
220
Appendix A
Appendix B
220
221
xii
LIST OF FIGURES
FIGURE 2.1
Media Issue-Attention Cycle
FIGURE 2.2
Public attitude towards Nanotechnology Acceptance
FIGURE 2.3
Public Perceived Nanotechnology Risks
FIGURE 2.4
Public Perceived Nanotechnology Benefits
FIGURE 2.5
Public Self-report Level of Awareness about Nanotechnology
FIGURE 2.6
Public Self-report Level of being Informed about Nanotechnology
FIGURE 2.7
Public Level of General Scientific Knowledge
FIGURE 2.8
Public Level of Knowledge about Nanotechnology
FIGURE 2.9
Public Amount of Attention Paid to Newspaper Content
FIGURE 2.10
Public Amount of Attention Paid to Television Content
FIGURE 2.11
Public Amount of Attention Paid to Online News Content
FIGURE 2.12
Media Coverage of Nanotechnology across 21 Newspapers
FIGURE 2.13
The New York Times and the Washington Post Coverage of Nanotechnology
FIGURE 2.14
Emergence of Nanotechnology as an Issue across High, Medium, and Low
Circulation Newspapers
FIGURE 2.15
Percentage of Risks-Related Nanotechnology Articles across the 21
Newspapers between January 1999 and August 2008
FIGURE 4.1
Science Media Use, Elaborative Processing, and Perceived Risks-versusBenefits of Nanotechnology (scale ranges only partially displayed on Y-axis)
FIGURE 4.2
Science Media Use, Elaborative Processing, and Public Support for Federal
Funding of Nanotechnology (scale ranges only partially displayed on Y-axis)
FIGURE 4.3
Structural Equation Model Predicting Public Support for Federal Funding of
Nanotechnology: Relationships among Endogenous Variables
FIGURE 6.1
Experts versus Public Support for Federal Funding of Nanotechnology
FIGURE 6.2
Experts versus Public Perceived Benefits of Nanotechnology
FIGURE 6.3
Experts versus Public Perceived Risks of Nanotechnology
FIGURE 7.1
Levels of Religious Beliefs, Scientific Status, and Perceived Risks-versusBenefits of Nanotechnology (scale ranges only partially displayed on Y-axis)
xiii
FIGURE 7.2
Deference to Scientific Authority, Scientific Status, and Support for Federal
Funding of Nanotechnology (scale ranges only partially displayed on Y-axis)
FIGURE 7.3
Science Media Use, Scientific Status, and Support for Federal Funding of
Nanotechnology (scale ranges only partially displayed on Y-axis)
FIGURE 7.4
Trust in Scientists, Scientific Status, and Support for Federal Funding of
Nanotechnology (scale ranges only partially displayed on Y-axis)
FIGURE 7.5
Perceived Risks-versus-Benefits, Scientific Status, and Support for Federal
Funding of Nanotechnology (scale ranges only partially displayed on Y-axis)
xiv
LIST OF TABLES
Table 2.1
Comparison of 2004 and 2007 Public Opinion: Descriptive Statistics of Similar
Question Items
Table 4.1
Descriptive Statistics of Question Items in the 2007 Public Opinion Survey
Table 4.2
Bivariate Correlations among the Variables in the 2007 Public Opinion Survey
Table 4.3
Ordinary Regression Model Predicting Public Perceived Risks-versus-Benefits of
Nanotechnology (Attitudinal Outcome 1) (standardized regression coefficients)
Table 4.4
Ordinary Regression Model Predicting Public Support for Federal Funding of
Nanotechnology (Attitudinal Outcome 2) (standardized regression coefficients)
Table 4.5
Ordinary Regression Model Predicting Public Level of Factual Scientific
Knowledge (Cognitive Outcome) (standardized regression coefficients)
Table 4.6
Influence of Exogenous Variables on Other Variables (Structural Equation
Model)
Table 4.7
Relationships among Endogenous Variables (Structural Equation Model)
Table 4.8
Summary of Findings in the Ordinary Regression Models and Structural
Equation Model
Table 7.1
Descriptive Statistics of Question Items in the 2007 Experts Survey
Table 7.2
Bivariate Correlations among the Variables in the 2007 Experts Survey
Table 7.3
Ordinary Regression Model with Perceived Risks-versus-Benefits of
Nanotechnology as Outcome Variable for the Public Sample (standardized
regression coefficients)
Table 7.4
Ordinary Regression Model with Perceived Risks-versus-Benefits of
Nanotechnology as Outcome Variable for the Expert Sample (standardized
regression coefficients)
Table 7.5
Ordinary Regression Model with Perceived Risks-versus-Benefits of
Nanotechnology as Outcome Variable with the Public and Expert Samples
Combined (standardized regression coefficients)
Table 7.6
Ordinary Regression Model with Support for Federal Funding of
Nanotechnology as Outcome Variable for the Public Sample (standardized
regression coefficients)
Table 7.7
Ordinary Regression Model with Support for Federal Funding of
Nanotechnology as Outcome Variable for the Expert Sample (standardized
regression coefficients)
Table 7.8
Ordinary Regression Model with Support for Federal Funding of
Nanotechnology as Outcome Variable with the Public and Expert Samples
Combined (standardized regression coefficients)
1
CHAPTER 1
INTRODUCTION
According to the 2006 State of the Union Speech by President George W. Bush,
nanotechnology is among the emerging technologies for which funding will be doubled over
the next ten years in the United States. Nanotechnology, the science of researching materials at
the atomic level, has been emphasized by the federal government because it will be a key
technology of the 21st century, with the potential not only to drive our next industrial revolution,
but also to reform and revolutionize the economy and other areas of our lives (National Science
and technology Council, 2000). With wide applications cutting across important sectors such as
medicine and healthcare, environment, and national defense, nanotechnology promises to
overcome many of the challenges that the world faces today (National Science and technology
Council, 2000). U.S. officials and industry leaders have projected that the annual global revenue
of nanotech-based products will reach $3.1 trillion by 2015 (Lux Research, 2008).
Despite the promise of nanotechnology, there are fears that the novel technology could
lead to various health and environmental problems, and other negative social, moral, and
ethical consequences (PCAST, 2005). For example, concerns have recently been raised that the
potential health risks of nano-particles could be similar to that of asbestos (Bergstein, 2008;
Chang, 2008). Moreover, critics also fear that nanotechnology could be used for human
enhancement, which may run counter to religious beliefs and raise ethical concerns (e.g.,
Bainbridge, 2003; Sententia, 2004).
Even though there are currently over 800 commercial products using nanotechnology in
the marketplace (Project on Emerging Technologies, 2008), the American public is, for the most
part, unfamiliar with the potential risks and benefits of this emerging technology (Scheufele &
2
Lewenstein, 2005). In a democratic society, public opinion is one of the key deciding factors of
future governmental policies and development of emerging technologies. More importantly,
public opinion about nanotechnology is likely to shape future funding-related policies and
affect the competitiveness of the U.S. in the international arena (Roco & Bainbridge, 2003).
Although the U.S. is currently leading the “nano race” in terms of public and private funding
and in terms of the number of patents filed, this technological supremacy may be threatened if
public attitudes toward nanotechnology were to become overwhelmingly negative.
Interestingly, public opinion surveys have shown that public attitudes toward
nanotechnology are currently leaning towards the optimistic side, with the public perceiving
greater benefits over risks, despite the fact that majority of the public are unaware of the
emerging technology (Peter D. Hart Research Associates, 2007) and approach the topic with few
preconceived ideas or firmly held attitudes about risks and benefits (Scheufele & Lewenstein,
2005). In many ways, nanotechnology is the opposite of what Iyengar and Kinder (1987) called
“chronically accessible,” that is, an issue that has been discussed heavily in public discourse and
that is so prominent in the audiences’ mind that it is often difficult to observe changes that are
due to communication effects. Given the complexity of the topic, coupled with low levels of
public awareness of the technology, nanotechnology is therefore an ideal issue to explore the
processes by which audiences gather information and try to make sense of this information
through interpersonal or intrapersonal channels.
Specifically, what are the specific mechanisms through which the public develops
opinions about the risks and benefits of nanotechnology and support for funding of the
emerging technology? By situating the queries within the debate between the “scientific literacy
model” (Miller et al., 1997) and the “cognitive miser model” (Fiske & Taylor, 1991), and by
drawing from two major theories in the field of communication – the “differential gains model”
3
(Scheufele, 2001, 2002) and the “cognitive mediation model” (Eveland, 2001, 2002) – this
dissertation will develop a more complete and sophisticated model of public understanding of
science.
In addition, how do attitudes of the public compare with the scientific experts in the
field of nanotechnology? Are public and expert attitudes in line or are they diverging? Do we
have to rally the scientific community and industry to have dialogues with the public? By
systematically comparing public opinion with experts’ opinion toward nanotechnology, this
dissertation will examine many of these issues in-depth.
1.1. Unresolved Issues in Understanding How Public Form Opinions about Nanotechnology
1.1.1. Ongoing Debate between the “Scientific Literacy Model” and the “Cognitive Miser
Model”
A major unresolved issue in understanding how the public forms opinions about science
in general and nanotechnology in particular concerns the tension between the “scientific literacy
model” (Miller et al., 1997) and the “cognitive miser model” (Fiske & Taylor, 1991). Studies of
public opinion processes of controversial science and technology have been going on for many
years, and there has been a longstanding debate between these two approaches. Even though
Brossard and her colleagues have established an extensive line of research on scientific literacy
and have resolved many of the supposed discrepancies (e.g., Brossard & Nisbet, 2007; Ho,
Brossard, & Scheufele, 2008), there remain unanswered questions that will be examined in this
dissertation.
The “scientific literacy model,” also known as the “knowledge deficit model” (Miller,
1998, 2004; Miller & Kimmel, 2001; Miller et al., 1997), assumes that a more scientifically
informed public would be more supportive of scientific research and would be able to make
4
more rational and knowledgeable judgments about scientific issues (Miller, 1998, 2004). The
model further assumes that the public is willing to seek out relevant information to develop
informed opinions about scientific issues. On the flip side, the low-information public, who are
ignorant of science (hence the term “knowledge deficit model”) are unsupportive of scientific
progress (Miller, 1998, 2004). In essence, the model assumes that a scientifically literate citizenry
is one that can effectively participate in public debates about science and hold the government
accountable over the direction of science policies.
Even though scientific knowledge has been demonstrated to have direct positive
influence on public attitudes toward science and scientific issues (Brossard, Lewenstein, &
Bonney, 2005; Nisbet et al., 2002; Sturgis, Cooper, & Fife-Schaw, 2005), and to have contingent
effects on public attitudes toward science and technology (e.g., Brossard, Scheufele, Kim, &
Lewenstein, in press; Ho et al., 2008; Sturgis & Allum, 2004), results of a large number of other
empirical studies have, on the contrary, demonstrated that scientific knowledge has a small to
modest role in shaping public opinions toward emerging science and technology (e.g., Allum,
Sturgis, Tabourazi, & Brunton-Smith, 2008; Brossard & Nisbet, 2007; Brossard & Shanahan,
2003; Lee, Scheufele, & Lewenstein, 2005; Priest, 2001; Scheufele & Lewenstein, 2005). As a
result, scholars have proposed the “cognitive miser model” as an alternative to the scientific
literacy model for explaining how the public forms opinions about science and technology. This
model posits that people are cognitive misers who will use a minimal amount of energy and
effort to make quick decisions in their daily lives; to be efficient, people often use cognitive
shortcuts or easily accessible heuristic cues, such as value predispositions, to make judgments
(Fiske & Taylor, 1991). In fact, recent studies have revealed that value predispositions such as
religious beliefs and trust in scientists explain a significant amount of variance in public
5
attitudes toward controversial scientific issues (e.g., Brossard & Nisbet, 2007; Brossard et al., in
press; Ho et al., 2008).
Instead of regarding them as two separate processes that work independently, this
dissertation argues that the scientific literacy model and the cognitive miser model are
complementary, parallel processes that work in tandem. In other words, individuals can
simultaneously use both cognition (e.g., factual scientific knowledge and sophisticated
knowledge about the scientific world) and heuristic cues (e.g., religious beliefs, deference to
scientific authority, and news frames) to form judgments about emerging technologies. In fact,
Ho, Brossard, and Scheufele (2008) have demonstrated that cognition and heuristic cues can
interact and that the impact of information on attitudes toward controversial scientific issue can
depend on which heuristics a person uses. Therefore, this dissertation will take an integrative
approach by considering both cognitive and heuristic factors when understanding opinion
formation about nanotechnology.
1.1.2. The Role of Communication Processes: Bridging the “Differential Gains Model”
and the “Cognitive Mediation Model”
For most Americans, the primary sources of information about science and technology
are television, the Internet, and newspapers (Pew Internet & American Life Project, 2006). Both
the content and tone of science in the mass media play a crucial role in shaping public attitude
toward science and technology (Ho et al., 2008; Nisbet, Brossard, & Kroepsch, 2003; Nisbet &
Lewenstein, 2002). This leads us to another unresolved issue: How does the public make sense
of the scientific information in the mass media? Previous studies that examined media effects on
public attitudes toward nanotechnology have looked at the amount of attention audiences pay to
science on the mass media without considering the fact that the public are active audiences
6
capable of using different types and levels of cognitive processing strategies to reflect upon and
absorb the scientific information that they gathered from the mass media.
This dissertation will draw on the notion of “reflective integration” (Kosicki & McLeod,
1990), as well as two major communication theoretical models namely the “differential gains
model” (Scheufele, 2001, 2002) and the “cognitive mediation model” (Eveland, 2001, 2002).
Applying these models will build a more complete and sophisticated model for understanding
the potential moderating and mediating mechanisms behind how the public form opinions
about nanotechnology. To be concise, “reflective integration” refers to an informationprocessing strategy in which people use to think about, try to understand, discuss with others,
and make sense of the news content from the mass media (Kosicki & McLeod, 1990).
Originally rooted in political communication, the “differential gains model” (Scheufele,
2001, 2002) extends the idea of reflective integration and proposes that political reflective
integration involves an “intrapersonal dimension” and an “interpersonal dimension” (Scheufele,
2001, p. 24). At the core of this model is a moderating process in which the impact of media use
on an individual’s political knowledge and other attitudinal and behavioral outcomes will be
highest if the individual (a) ponders the mediated information and integrates it into his/her
preexisting knowledge (“intrapersonal reflection”) and (b) talks about the mediated information
with others after exposure to media content (“interpersonal reflection”) (Scheufele, 2001, 2002).
The term “differential gains” therefore implies that individuals will benefit differentially from
mass media, depending on their levels of intrapersonal or interpersonal reflection. Statistically,
intrapersonal reflection is processing that is equivalent to the interaction of news elaboration
and media use; and interpersonal reflection is processing that correspondence to the interaction
of interpersonal discussion and media use. The differential gains model has been empirically
supported by numerous studies, especially in the contexts of political knowledge and civic
7
participation (e.g., Hardy & Scheufele, 2005; Nisbet, Nisbet, Scheufele, & Shanahan, 2004;
Scheufele, 2002).
On the other hand, Eveland’s (2001, 2002) “cognitive mediation model” posits that
cognitive information-processing strategy (i.e., news attention and elaboration) will mediate the
influence of mass media use on public learning from the news media. Statistically, this refers to
the process that the positive relationship between news media use will substantially decrease or
become non-significant when controlling for communication processing variables (i.e., news
elaboration and interpersonal discussion). Empirically, the basic premises of the cognitive
mediation model have garnered strong support from numerous studies in the area of political
communication (e.g., Beaudoin & Thorson, 2004; Eveland & Thomson, 2006; Maurer &
Reinemann, 2006; Shah, Cho, Eveland, & Kwak, 2005).
Scholars have generated separate lines of research based on these two theoretical models
(e.g., Eveland & Thomson, 2006; Hardy & Scheufele, 2005; Maurer & Reinemann, 2006; Nisbet &
Scheufele, 2004) without acknowledging the nexus that could be drawn between them.
Extending these two theoretical models to the realm of science communication, this dissertation
argues that there is good reason to propose that reflective integration could moderate and mediate
the influence of mass media use on public cognitions and attitudes toward nanotechnology. By
testing the competing hypotheses of the differential gains model and the cognitive mediation
model in a single study, this dissertation will attempt to bridge the disconnection between these
two theoretical models and develop a more complete and refined model that could explain and
predict how the public forms opinions about nanotechnology.
8
1.2. Unresolved Issues in Public versus Expert Opinion of Nanotechnology
Another major area of research in science communication has been the extensive line of
comparative studies that look at the differences between the lay public and expert attitudes
toward science and technology, especially for perceived risks. Scholars in risk communication
research have generally found that experts view risks differently from members of the lay
public, and that expert judgments are closer to reality than those of the public (e.g., Cole &
Withey, 1981; Sandman, Weinstein, & Klotz, 1987; Slovic, 1987). Put another way, expert
judgments of risk are often viewed as objective and can be measured and quantified
scientifically, whereas public attitude toward risk are often deemed as subjective and qualitative.
Assessing attitudinal gaps between the lay public and the scientific experts has important
practical implications. In particular, attitudinal gaps will be a key indicator of whether huge
amount of resources should be devoted to conference sessions and forums to create
opportunities for the scientific experts and elites to have dialogues with the lay public in order
to close the attitudinal gap.
Even though previous studies have examined these attitudinal gaps across various
scientific issues (e.g., Kraus, Malmfors, & Slovic, 1992; Lazo, Kinnell, & Fisher, 2000; McDaniels,
Axelrod, Cavanagh, & Slovic, 1997), no researchers have yet used multivariate statistical
analyses to look systematically at experts-laypersons attitudinal gap (if any) for nanotechnology
in the United States. Some of these unanswered questions are therefore worthwhile to explore
in this dissertation: Are there differences between expert and public opinion about
nanotechnology? Is expert opinion more objective than public opinion when it comes to
decision-making about the emerging technology? Perhaps another more interesting question
that arises is: Will the mass media and other heuristic factors have differential influence on
attitudes toward nanotechnology depending on the scientific status of the two groups (i.e., the
9
experts versus the laypersons)? These are unresolved issues that this dissertation will
investigate.
1.3. Objectives, Originality, and Significance
Above and beyond providing separate descriptions of media coverage of
nanotechnology, public attitudes toward nanotechnology, and expert opinions of the emerging
technology, the major contribution of this dissertation is its investigation of the intersection
between the mass media, public opinion, and expert opinion. In other words, this dissertation
not only compares expert and public opinion of nanotechnology but also examines the
simultaneous influence of the mass media on public opinion and expert opinion of the
emerging technology.
Given the ongoing debate between the scientific literacy model and the cognitive miser
model, and the inconclusive findings on the impact of scientific knowledge on public opinion, it
is crucial to understand whether and how the public forms opinions about nanotechnology. An
important way to achieve such an understanding is to explicate the mechanisms through which
both cognitive factors and heuristic factors influence public attitudes toward nanotechnology.
This dissertation uses two studies (hereafter “Study 1” and “Study 2”) to achieve its two major
objectives.
1.3.1 Study One - Effects of Value Predispositions, Mass Media, and Cognitive
Processing on Public Attitudes toward Nanotechnology: Testing Moderating and
Mediating Mechanisms
The first objective of this dissertation is to explore the moderating and mediating
mechanisms through which cognitive and heuristic factors influence public perceived risksversus-benefits of nanotechnology and levels of support for federal funding of nanotechnology,
10
which is accomplished by Study 1. By applying the differential gains model and the cognitive
mediation model as theoretical frameworks, Study 1 considers reflective integration, in the form
of elaborative processing and interpersonal discussion, as a moderator that strengthens or
weakens the influence of mass media on public level of factual scientific knowledge and their
attitudes toward the emerging technology. Study 1 also pays special attention to the role of
reflective integration in mediating science media use and the important cognitive and
attitudinal outcomes related to nanotechnology.
As far as the research area of science communication is concerned, this dissertation is the
first original study in the field to examine these mediating mechanisms for nanotechnology.
Previous studies have only assessed the moderating mechanisms behind the formation of public
attitudes toward emerging technologies (e.g., Brossard et al., in press; Lee et al., 2005). As such,
the explication of such mediating and moderating mechanisms in this dissertation should
contribute to developing a model that explains and predicts how public form opinions about
nanotechnology. Such a model could also offer both theoretical and practical implications for
science communication scholars, practitioners, and policymakers alike. In addition, the findings
will help communication scholars perform a more theoretically informed evaluation of how
public judgments are formed. As for practical implications, the findings will help
communication practitioners design more effective strategies to provide the most accurate and
up-to-date information about nanotechnology to the public.
Moreover, this dissertation is also the first study in the field of communication to apply
and test the competing hypotheses between the differential gains model and the cognitive
mediation model to understand how cognitive processing in the form of reflective integration
influence public decision-making about scientific issue. By so doing, this study not only helps to
develop a communication theory-centered approach to understanding public attitudes toward
11
science and technologies but also contributes theoretically to the differential gains model and
the cognitive mediation model by bringing the communication models out of the traditional
political context to a scientific context. Furthermore, by extending both models to opinion
formation about nanotechnology, Study 1 undoubtedly expanded the applicability of both
theories.
1.3.2 Study Two – Experts versus Public Attitudes toward Nanotechnology
Based on the results and recommendations in Study 1, the second objective of this
dissertation is to compare public and experts’ attitudes toward nanotechnology and to address
the important question of whether experts are indeed more objective in their judgment of
nanotechnology than do the public. Building on a recent descriptive study conducted by
Scheufele et al. (2007) that compared perceived risks and benefits of nanotechnology between
nanoscientists and the general public in the U.S., Study 2 focuses on three general questions: (a)
it examines the extent to which experts and the public differ in terms of both their perceived
risks-versus-benefits of nanotechnology and their levels of support for federal funding of the
emerging technology; (b) it examines the extent to which the significant heuristic factors found
in Study 1 would influence both experts and public attitudes nanotechnology; and finally, (c) it
determines if the experts use the same or different set of considerations (i.e., the heuristic factors
identified in Study 1) to make judgments of nanotechnology in comparison with the public.
An exhaustive literature search in the area of science communication shows that this
dissertation is the first study to use multivariate statistical analyses to examine differences in the
public and expert attitudes toward nanotechnology. Study 2 is also much more rigorous
methodologically than most previous studies in terms of sampling procedure (e.g., experts are
more representative of the nano-scientists working in the U.S.) and sample size (e.g., a sample
size of 363 is larger than a lot of previous studies). In addition, Study 2 accounts for other
12
exogenous variables such as age and gender, and provides more advanced statistics by going
beyond descriptive statistics, which is a methodological improvement from previous studies.
Given the importance of the mass media as an influence on opinions about emerging
technologies, this dissertation is also the first study to present findings regarding the
intersection of mass media, experts’ opinion, and public opinion. The comparison of the public
and experts’ attitudes toward nanotechnology should produce worthwhile theoretical and
practical implications. Theoretically, the comparisons between the public and experts should
contribute to existing risk communication literature. In practice, the findings from Study 2
should inform practitioners on whether dialogues between the experts and the lay public are
necessary so as to bridge the potential risks and other attitudinal gaps between these two
groups.
1.4. Organization of the Chapters
This dissertation consists of eight chapters. The general rubric of the dissertation is as
follows: Chapter 2 provides an introduction to nanotechnology, explains its importance to the
society, and looks at the intersection of nanotechnology, mass media, and public opinion. In this
chapter, I also justify the need for social scientific input in understanding public perceived
benefits and risks related to nanotechnology, and public acceptance of the emerging technology.
Using identical (or near-identical) questions from two nationally representative public opinion
surveys conducted in 2004 and 2007, I describe trends and changes in public opinion about
nanotechnology between these two years. Some of the responses compared include public
perceived risks and benefits of nanotechnology, public support for federal funding of
nanotechnology, and public attention to science in the mass media in both years. In addition,
using the software “VantagePoint” and the Lexis-Nexis database, a comprehensive search for
13
newspaper articles was conducted, followed by a simple content analysis in Chapter 2 to
determine the amount and tone of media coverage of nanotechnology over a period of thirty
years.
Study 1 is covered from Chapters 3 through 5. First, Chapter 3 reviews the extant
literature, including the underlying moderating and mediating mechanisms behind public
perceived risks-versus-benefits and public support for federal funding of nanotechnology. In
this chapter, I will provide the background, rationale, and arguments for the formal hypotheses
postulated in Study 1. In particular, I will provide a concept explication of the two attitudinal
outcome variables of interest in Study 1 – perceived risks-versus-benefits and support for
federal funding of nanotechnology. Furthermore, this chapter introduces the theory of framing
(Entman, 1991; Goffman, 1974; Iyengar, 1991; Scheufele, 1999, 2000), the cognitive mediation
model (Eveland, 2001), and the differential gains model (Scheufele, 2002) as theoretical
frameworks to argue for an informational pathway and a heuristic pathway through which the
public could use to make decisions about nanotechnology. I will also introduce the concept of
cognitive processing in the form of reflective integration into the existing literature.
Chapter 4 tests the hypotheses postulated in Study 1, and describes the methods and
results of Study 1. The methods section in this chapter provides details on the data and
sampling procedures employed, the measures used, and the analytical approaches for testing
the hypotheses proposed in Study 1. Specifically, the methods section describes the 2007
nationally representative survey data on public opinion about nanotechnology (N = 1,015) and
explains the index construction procedure and the internal consistency for each of the new
indices created. Following this, I will elaborate on and provide the rationale for the application
of the multivariate techniques in Study 1. In particular, descriptive analyses, including bivariate
correlations, and advanced inferential multivariate techniques, including ordinary regression
14
analysis, will be used to examine the potential direct and moderating relationships, and
structural equation modeling analysis will be used to examine the potential direct and
mediating relationships among the variables concerned. This chapter also discusses the nature
and treatment of missing values in Study 1. This will be followed by the results section, in
which the findings for the hypotheses posited in Study 1 will be reported.
Chapter 5 discusses the results of the findings in Study 1. I include a quick recap of the
major findings in Study 1, and then move on to devote three sub-sections to explain the direct
effects, the moderating mechanisms, and the mediating mechanisms. This chapter also
dedicates one section to outline the implications of the findings for theory and practice.
Chapters 6 and 7 cover Study 2, which examines the interplay between the experts and
the public in terms of their attitudes toward nanotechnology. In Chapter 6, I will review extant
literature comparing risk judgments between the lay public and the experts, critically assess
those previous studies, and explain how Study 2 will fill in the gaps found in those previous
studies.
In chapter 7, I will provide a methods section describing in detail the data collection
procedure of the 2007 survey of the U.S. nanotechnology scientists that will be used for
comparison with the public in this study. Study 2 uses identical or similar measures from both
the expert and the public samples for comparison and reports the internal consistency for new
indices that will be constructed. This chapter also describes the analytical approach, in which
Study 2 will be using ordinary regression analyses. Finally, I will describe the results of Study 2.
The results section in this chapter reports the ordinary regression model of factors predicting
perceived risks-versus-benefits of nanotechnology and factors predicting support for federal
funding of nanotechnology within the public-only sample, the experts-only sample, and the
combined sample of public and experts.
15
Chapter 8 provides an in-depth discussion on the findings and the implications of Study
2 for theory and practice. Specifically, this chapter explains the major attitudinal differences and
similarities found between the two groups.
Finally, Chapter 9 provides closure to the dissertation by reviewing, summarizing, and
synthesizing the results of both Study 1 and Study 2. This chapter also describes the major
theoretical, conceptual, and practical contributions of this dissertation to science communication
and policymaking. Moreover, this chapter outlines the major shortcomings of Studies 1 and 2
and recommend possible ways to overcome them in future research. Lastly, this chapter
provides an overall conclusion for the dissertation.
16
CHAPTER 2
NANOTECHNOLOGY, MASS MEDIA, AND PUBLIC OPINION
As mentioned earlier in Chapter 1, nanotechnology has been hailed as the next key
technology of the 21st century and an important stimulus for our next industrial revolution
(National Science and Technology Council, 2000). In fact, U.S. policymakers and scientists alike
have confidently predicted that nanotechnology possesses the potential to radically transform
and revolutionize the economy and other domains of our lives (National Science and
Technology Council, 2000). On the other hand, there are fears that the emerging technology
could also lead to various health and environmental problems, and other negative social, moral,
and ethical consequences (PCAST, 2005).
To gain a better understanding of nanotechnology, Chapter 2 will provide an
introduction to nanotechnology and its implications for society, economy, and governmental
policy. This chapter will also look at the intersection of nanotechnology, mass media, and public
opinion. In this chapter, I will justify the need for social scientific input in understanding public
perceived benefits and risks related to nanotechnology, and public acceptance of the emerging
technology. Following this, this chapter will describe public opinion trends and media coverage
of nanotechnology over the years.
2.1. Introduction to Nanotechnology
“Nano” is derived from the Greek word for dwarf, and if we were to use it as a prefix
for any unit such as a second or an ounce, it refers to a billionth of that unit. Officially, the
National Nanotechnology Initiative (National Science and Technology Council, 2007) defined
nanotechnology as follows:
17
…the understanding and control of matter at dimensions between approximately
1 and 100 nanometers, where unique phenomena enable novel applications.
Encompassing nanoscale science, engineering, and technology, nanotechnology
involves imaging, measuring, modeling, and manipulating matter at this length
scale. A nanometer is one-billionth of a meter. A sheet of paper is about 100,000
nanometers thick; a single gold atom is about a third of a nanometer in diameter.
Dimensions between approximately 1 and 100 nanometers are known as the
nanoscale. Unusual physical, chemical, and biological properties can emerge in
materials at the nanoscale. These properties may differ in important ways from
the properties of bulk materials and single atoms or molecules.” (p. 5)
2.1.1. Nanotechnology and Its Societal Implications
The impact of nanotechnology on the society is expected to be enormous due to its wide
applications and its potential benefits. Currently, nearly more than 800 commercial products
that contain nano-engineered particles are known to be available in the marketplace, and the
application of nanotechnology is continuing on an upward trend, according to the Woodrow
Wilson International Center for Scholars (2007). For example, nanoscale materials are used in
some cosmetics to improve their effectiveness, and nanoscale titanium oxide or zinc oxide is
used in sunscreens to effectively reflect ultraviolent rays to prevent sunburns. Scratch- and
glare-resistant coatings manufactured using nanoscale materials are being applied to eyeglasses,
windows, and car mirrors. In addition, batteries using nanotechnology are being made to
deliver greater power and speed with lesser heat to improve overall efficiency, and nanoscale
dry powder are being made to neutralize gas and liquid toxins in chemical spills.
Looking forward, nanotechnology may help overcome some of the world’s biggest
challenges across many fields such as medicine and healthcare, electronics, aeronautics,
agriculture, energy, homeland security and national defense, environment, information
18
technology, and transportation. Some of these potential benefits include: clean, secure
affordable energy; stronger, lighter, more durable materials; low-cost filters to provide clean
drinking water; lighting that uses a fraction of the energy; sensors to detect and identify
harmful chemical or biological agents; and techniques to clean up hazardous chemicals in the
environment. In the medical world, this technology could potentially improve detection and
treatment of diseases such as cancer and HIV/AIDS with fewer side effects. Some of the
materials being created, such as cages of carbon atoms, also known as buckyballs, show
promise as tools for environmental cleanup. Others, such as carbon nanotubes, are expected to
revolutionize the electronics industry. In the United States alone, corporate giants such as
General Electric, Motorola, DuPont, and Lucent are now pursuing nanotechnology. Boeing is
working on applications to cut the weight of rockets, aircraft, and satellites. Kodak has focused
on uses in flat-screen displays, inkjet paper, and medical imaging. 3M sees potential
applications in optical films, self-cleaning glass, and dentistry. Therefore, nanotechnology has
the prospect to affect nearly every industry and every facet of our daily lives.
2.1.2. Nanotechnology and Its Implications to Economy and Governmental Policy
In 2007 alone, $147 billion worth of nanotech-enabled products were produced in the
market, and the annual global revenue of nanotech-based products is expected to reach $3.1
trillion by 2015 (Lux Research, 2008). As mentioned in the introduction, nanotechnology is
highlighted as among the emerging technologies in which research funding will be doubled
over the next decade, according to President Bush’s 2006 State of the Union speech. Earlier on,
under the priority of the Bush administration, the National Nanotechnology Initiative (NNI)
was established in the fiscal year 2001 to coordinate federal research and development in
nanotechnology. To give this effort a boost, President Bush signed the 21st Century
Nanotechnology Research and Development Act in 2003 to authorize funding for
19
nanotechnology research and development for four years beginning in 2005, and put into law
programs and activities supported by the NNI.
According to the 2007 NNI Strategic Plan, the vision of the NNI is a future in which the
ability to understand and control matter at the nanoscale will lead to a revolution in technology
and industry that benefits society. To this end, there are four main goals to be fulfilled: (1) to
advance a world-class nanotechnology research and development program; (2) to foster the
transfer of new technologies into products for commercial and public benefit; (3) to develop and
sustain educational resources, a skilled workforce, and the supporting infrastructure and tools
to advance nanotechnology; and (4) to support responsible development of nanotechnology.
(The 2007 NNI Strategic Plan updates and replaces the first version that was published in
December 2004.) The NNI is essentially a multiagency, multidisciplinary program consisting of
25 participating agencies, such as the National Science Foundation, the Food and Drug
Administration, and the National Institutes of Health, as of December 2007. The establishment
of the NNI is a clear indication of the federal government’s effort to maintain the U.S. as a longterm global leader of nanotechnology in the world stage.
Moreover, according to figures reported by the European Commission (2005), U.S.
federal funding of nanotechnology research and development has reached 910 million euros,
with state funding amounting to 333.3 million euros and private funding reaching 1700 million
euros. As a result, the worldwide division of the overall expenditure (including public and
private) in 2004 was 37 percent for the U.S., 24 percent for Europe, 28 percent for Japan, and 11
percent for the rest of the world. The U.S. is clearly leading the world in the research and
development expenditure. With the large budget, it is not surprising that more than 4,800
patents have been identified under the nanotechnology classification created by the U.S. Patent
and Trademark Office (National Science and Technology Council, 2007). Besides this, several
20
nanotechnology-specific journals, such as Nature Nanotechnology and Journal of Nanoparticle
Research, have been launched by professional academic societies in recent years.
Despite this optimism and efforts, government officials have also expressed concerns
about the uncertainty over the health and environmental impacts of nanotechnology, and its
potential social, moral, and ethical implications (PCAST, 2005). For instance, a recent study
published in Nature Nanotechnology by Poland et al. (2008) found that carbon nanotubes may
pose health risks similar to asbestos, which is a cause of lung disease. Besides this, there are
fears that nanotechnology may lead to more pollution and environmental contamination as
toxic nano-particles may penetrate into the human bodies. Moreover, pundits and social
activists alike have raised a number of ethical, moral, and legal issues related to nanotechnology.
Some critics are concerned about the possibility that tiny new surveillance nano-devices may
lead to the loss of personal privacy, the likelihood that terrorists may misuse nanotechnology
against the U.S., and the chance that nanotechnology may lead to the uncontrollable spread of
very tiny self-replicating robots (PCAST, 2005). Other ethical concerns include the idea that
nanotechnology may be used to alter human bodies and to eradicate all human diseases and
aging, which may run counter to commonly-held moral or even, religious beliefs (PCAST, 2005).
The concern about using nanotechnology for human enhancement is, in part, a result of
the Nano-Bio-Info-Cogno (NBIC) program that is run by the National Science and Technology
Council. The NBIC program is the synergistic combination of four major provinces of science
and technology which are progressing rapidly: (a) nanoscience and nanotechnology; (b)
biotechnology and biomedicine, including genetic engineering; (c) information technology,
including advanced computing and communications; and (d) cognitive science, including
cognitive neuroscience (Roco & Bainbridge, 2002). The confluence of the four technologies offers
the promise of improving human lives in many ways. In addition, the U.S. Food and Drug
21
Administration officially defined “nanotechnology” as part of the NBIC technologies that
highlight the unity of nature at the nanoscale, and the intelligible processes of evolution that
have constructed life and intelligence, from the nanoscale, without divine intervention
(Bainbridge, 2003; Sententia, 2004). However, researchers have pointed out that using
nanotechnology for human performance enhancement will inevitably come with criticisms
(Sandel, 2004; Sententia, 2004). The debate will evolve into a discussion of whether genetically
engineering some of the behavioral traits in human is ethical (Stock, 2002). Therefore,
developers of NBIC technologies face a multitude of obstacles around the basic right to
augment human beings including political, disciplinary, and religious sectarianism (Caplan,
2003; Sententia, 2004).
2.1.3. Why Public Opinion Matters
However, with the increased in nanotechnology-related commercial products available
in the marketplace, the American public is still largely unfamiliar with the potential risks and
benefits of this emerging technology. As a democratic society, public opinion is one of the key
determinants of future governmental policies and development of the emerging technology in
the U.S. More importantly, public opinion about nanotechnology is likely to shape future
funding-related policies and affect the competitiveness of the U.S. in the world. As it stands
now, public attitudes toward nanotechnology is leaning towards the optimistic side, despite the
fact that majority of the public are unaware of the emerging technology (Scheufele &
Lewenstein, 2005). Nevertheless, there are concerns that public attitudes toward
nanotechnology may develop in the opposite direction, such that public opinion becomes
overwhelmingly negative (e.g., Sententia, 2004). For example, a unique challenge facing the
NBIC converging technologies is the association of the negative qualities, both real and
perceived, of one of the technologies to the others. There are speculations that negative attitudes
22
toward genetically modified organisms, human cloning, and stem cell research could be
transferred from the biotechnology field to nanotechnology, even though the emerging field
and its products may be unrelated to the earlier biotechnology-related work. Unless the
government and industry were to work on building public confidence in nanotechnology, the
public may reach for the “No-Nano” label in future, potentially resulting in huge economic and
social losses.
2.1.4. Mass Media and Public Opinion
For a majority of the American public, the mass media – especially television, the
Internet, and newspapers – are the primary source of information about science and technology
(Pew Internet & American Life Project, 2006). A recent study conducted by the Pew Internet &
American Life Project (2006) shows that 41 percent and 20 percent of Americans retrieved most
of their science news and information from television and the Internet respectively. This is
followed by newspapers and magazines together, which are cited by 14 percent of the public as
their main sources of news and information about science. Both the content and tone of science
in the mass media have been demonstrated to play a crucial role in shaping public attitudes
toward science and technology (Ho et al, 2008; Nisbet et al., 2003; Nisbet & Lewenstein, 2002).
Put differently, media coverage about a particular issue could shape public opinion about that
issue, influencing not only the amount of attention (e.g., Ho et al., 2007), but also what and how
the public think about the issue (e.g., McCombs & Shaw, 1972; Scheufele, 1999). Currently,
media coverage of nanotechnology has been overwhelmingly positive. However, empirical
research has shown that media attention to science and technology usually goes through a cycle
(McComas & Shanahan, 1999; Nisbet et al., 2003). Downs (1972) theorized that public attention
to environmental issues characteristically passes through five stages: (a) a pre-problem stage
leads to (b) a period of alarmed discovery associated with specific problems or hazards. Then (c)
23
the public realizes the cost of making significant progress, and this stage is followed by (d) a
“gradual decline of intense public interest” (p. 39-40). This decline in attention leads to the (e)
post-problem phase, in which media attention toward the issue settles down (although at a
higher level than the initial part of the cycle). (Refer to illustration on Figure 2.1, derived from
Nisbet & Huge, 2006.) Empirically, McComas and Shanahan (1999) have shown that the
frequency of newspaper coverage reflected the cyclical attention to global warming, and such
patterns have also been shown in the context of the stem cell controversy (Nisbet et al., 2003)
and plant biotechnology (Nisbet & Huge, 2006). Nanotechnology may go through such a cycle
as well. Policymakers, scientists, and communication practitioners should take this into account
as they plan communication strategies, so that nanotechnology could avoid the unfortunate fate
of genetically modified technology.
As Roco and Bainbridge (2003) aptly pointed out, “Negative public attitudes toward
nanotechnology could impede research and development, leaving the benefits of
nanotechnology unrealized and its economic potential untapped, or worse, leaving the
development of nanotechnology to countries and researchers who are not constrained by
regulations and ethical norms held by most scientists worldwide” (p. 6-7). With these
considerations in mind, scholars argued that public attitudes toward nanotechnology should be
examined in-depth at an early stage of the technology development (Renn & Roco, 2006; Roco,
2003). In particular, social scientific research is needed to examine public understanding and
attitudes toward nanotechnology and audience response to various media products. The future
of nanotechnology will be determined in large part by the public assessment of its potential
risks and benefits. With the help of opinion polls and nationally representative surveys, we can
then determine where the public stands on nanotechnology-related issues, track the changing
24
attitudes toward nanotechnology over time, and compare public understanding of
technological issues across different cultures.
Suffice it to say, the “public” is not a single homogenous group. Instead, there are many
different groups of the “publics” with a wide range of views, shaped by a variety of internal
factors, such as demographic variables, socioeconomic status, political ideology, and religious
beliefs, and a host of external factors such as the mass media and interpersonal discussions.
Each public has its own unique expectations, interests, hopes, and fears about science in general
and nanotechnology in particular. As such, it is pertinent to understand the underlying values
and belief systems that contribute to the attitudes of these publics in order to bring about
effective communication about nanotechnology. In other words, it is necessary to understand
the mechanisms behind how the public makes judgments about the emerging technology. This
will allow us to improve the ability to identify and address the unique and common concerns of
each group, and to enable more informed public participation in future dialogues with all other
stakeholders about nanotechnology.
Given the important roles of public opinion and mass media in the future research and
development of nanotechnology, it is first necessary to offer a longitudinal assessment and
examine shifts (if any) in public opinion about nanotechnology. Second, it is also necessary to
track changes in the volume and tone of media coverage of nanotechnology over the years.
Descriptive analyses of public opinion and media coverage trends will be provided in the
following sections. In essence, the following sections will enable us to answer the following
questions: Where are public attitudes toward nanotechnology moving since 2004? How have
media coverage of nanotechnology developed over the years and where will nanotechnology
coverage go next?
25
2.2. Public Opinion Trends
To examine public opinion trends, descriptive analyses of identical or near-identical
items in both the 2004 and 2007 nationally representative surveys are presented here1. In other
words, those questions that are similarly worded were selected for comparison. The means and
standard deviations of the public responses to each of the items in both the 2004 and 2007
survey data sets are shown in Table 2.1.
Overall, there were only slight variations in terms of public attitude towards
nanotechnology, despite the quadrupling of public and private funding in nanotechnology
research, the increased in media coverage of nanotechnology, and $40 million per year in
funding for research on the societal implications of nanotechnology over the course of the last
three years. The differences, if any, were not large. This could be due, in part, to the short timespan in between the surveys and perhaps, also because nanotechnology is still at an early stage
of the issue-attention cycle in the media. With more data tapping the same opinion collected in
the future, we may be able to see greater fluctuations over time.
Specifically, when it comes to nanotechnology acceptance (see Figure 2.2), there was a
slight increase in support for federal funding of nanotechnology between 2004 and 2007. On a
ten-point scale, ranging from 1 “do not agree at all” to 10 “agree very much,” the value
increased from an average of 5.36 in 2004 to an average of 5.90 in 2007. It is important to note
that even though we did see an increase, public support for federal funding of nanotechnology
had not been too overwhelming.
Generally, the public perceives greater benefits than risks related to nanotechnology. As
shown in Figure 2.3, we see small fluctuations in public attitudes toward nanotechnology risks
Details about the 2004 public opinion sample are provided in Appendix A. Information about the 2007
public opinion sample are provided in Chapter 4.
1
26
between 2004 and 2007. When asked if nanotechnology will lead to a loss of privacy on a tenpoint scale ranging from 1 “do not agree at all” to 10 “agree very much,” the average public
response dropped from 6.43 point in 2004 to 5.87 point in 2007. Conversely, when asked if
nanotechnology will lead to a loss of jobs, the average public response increased from 4.99 point
in 2004 to 5.16 point in 2007. The proportion of public agreeing that nanotechnology will lead to
an arms race (4.99 point in 2004; 5.16 point in 2007) and produce self-replicating robots (3.29
point in 2004; 3.17 point in 2007) remained quite constant in both years. Overall, perceived risks
had been very low. With respect to perceived benefits (see Figure 2.4), there were again, only
small differences in both years. The public who agreed that nanotechnology will lead to better
treatment of diseases, bring about a cleaner environment, improve human activities, and
improve national security fluctuates from 6.41 point to 7.28 point in both years (1 “do not agree
at all”; 10 “agree very much”).
However, on similar ten-point scales, respondents’ self-reported awareness (3.69 point
in 2004; 3.99 point in 2007) and levels of being informed about nanotechnology (3.18 point in
2004; 3.39 point in 2007) remained low in both years (see Figures 2.5 and 2.6). Put another way,
the public are for the most part unaware and under-informed about nanotechnology. Likewise,
public knowledge about science in general and nanotechnology remained constant between
2004 and 2007 although we can spot an upward trend (see Figures 2.7 and 2.8). The public were
asked to respond to three knowledge questions related to science in general and to respond to
five knowledge questions related to nanotechnology. When it comes to knowledge about
science in general, the public scored an average of 1.77 points in 2004 and an average of 1.98
points in 2007 (out of a total score of 3 points). Likewise, with respect to knowledge about
nanotechnology, the public scored an average of 3.90 points in 2004 and an average of 4.07 (out
of a total score of 6 points).
27
Interestingly, public attention toward new areas of research in science and technology in
the mass media changed between 2004 and 2007. Figure 2.9 presents the average amount of
attention that the public paid to such stories in newspapers, which was fairly low in both 2004
and 2007 (slightly less than 5 points out of 10). With respect to television attention (see Figure
2.10), there is an increase in the amount of attention paid to science-related television programs
between 2004 and 2007. These include attention paid to news stories related to science and
technology, news stories about specific scientific developments such as nanotechnology, science
fiction dramas, and science documentaries. Noticeably, the average amount of attention paid to
science fiction dramas increased from 2.74 point in 2004 to 5.67 point in 2007. As demonstrated
in Figure 2.11, there was also marked increase in attention paid to science-related content
among Internet users. In particular, the average amount of attention paid to online content
related to science and technology increased from 2.38 point in 2004 to 4.36 point in 2007. The
average amount of attention paid to online content related to specific scientific developments
such as nanotechnology also increased from 2.09 point in 2004 to 4.12 point in 2007. The increase
in attention to online content is in line with the Pew Research Center’s recent finding that the
public is turning to the Internet as a resource for news and information about science (Pew
Internet and American Life Project, 2006). Despite the increase in attention, the level of attention
still remains somewhat low.
On the whole, public opinion about nanotechnology, measured in terms of acceptance,
perceived risks and benefits, knowledge, self-reported awareness and being informed, and
attention to science and technology content in the media, has remained somewhat stable
between 2004 and 2007. This could be due to the short interval between 2004 and 2007. It could
also be due to the fact that nanotechnology is still at an early stage of the issue-attention cycle in
28
the media. More fluctuations in public opinion will likely develop as the technology moves
through various stages of the issue-attention cycle.
2.3. Media Coverage of Nanotechnology over the Years
This section provides a descriptive analysis of the amount and tone of media coverage of
nanotechnology from January 1969 to August 2008. This analysis was carried out in a series of
comprehensive steps. First, to determine the amount of media coverage of nanotechnology, a
comprehensive search term was used to gather the nanotechnology-related articles (see
Appendix B for details). Three groups of U.S. newspapers were selected for the analysis: (1)
High circulation newspapers (papers over 500,000 circulation) - USA Today, New York Times,
Washington Post, Houston Chronicle, Boston Globe, Minneapolis Star Tribune, The Atlanta JournalConstitution; (2) Medium circulation newspapers (papers between 100,000 and 500,000
circulation) - Pittsburgh Post-Gazette (Pennsylvania), Plain Dealer (Cleveland, Ohio), Milwaukee
Journal Sentinel (Wisconsin), The Seattle Times, St. Louis Post-Dispatch (Missouri), St. Petersburg
Times (Florida), and Sacramento Bee (California); and (3) Small circulation newspapers (papers
under 100,000 circulation) - The Augusta Chronicle (Georgia), The Santa Fe New Mexican (New
Mexico), Bangor Daily News (Maine), Lewiston Morning Tribune (Idaho), The Herald (Rock Hill, S.
C.), Star-News, Wyoming Tribune-Eagle (Cheyenne). Altogether, these 21 papers were selected to
represent newspapers in the respective circulation categories. The newspapers in each of the
circulation categories were further stratified across different newspaper chains to ensure high
geographical representativeness.
Next, the search results of the news articles from Lexis-Nexis were imported into the
software program “VantagePoint” for cleaning and analyses. “VantagePoint” is a desktop textmining tool that enables researchers to navigate through large volumes of search results from
29
databases such as Lexis-Nexis and analyze the structured text to find patterns and relationships
(Search Technology, 2008). Importing refers to getting the raw data from the Lexis-Nexis search
results into VantagePoint and mining the raw data to get more data from it. The data were then
transformed into a consistent set, combining things that the researcher wants to analyze as a
group. This is followed by analysis of the final cleaned data.2
Figure 2.12 shows the media coverage of nanotechnology across all 21 newspapers.
It is clear that media coverage of nanotechnology has been scant. It rose steadily from the late
1990s onwards and peaked in 2003. Media coverage has been pretty high ever since, covering at
least 100 news articles on nanotechnology per year.
Figure 2.13 shows the coverage of nanotechnology in the New York Times and the
Washington Post. As these are elite newspapers that could set the agenda on what the public
would think about, it therefore followed a similar trend as that in Figure 2.10.
In addition, Figure 2.14 displays the media coverage of nanotechnology across the high
circulation, medium circulation, and small circulation newspapers. As shown across all three
lines in the figure, nanotechnology has evolved from a purely elite issue to a local issue over the
years. Nanotechnology was first covered in the high circulation newspapers in the late 1970s
and early 80s. The issue made its foray into the medium circulation newspapers in the late 80s.
Finally, news about nanotechnology only started to appear in the low circulation newspapers in
the late 90s. This suggests a lagged effect in which nanotechnology is gaining prominence over
time, turning from an elite issue to a local issue.
The data on the amount and tone of media coverage of nanotechnology using VantagePoint and LexisNexis were originally collected with the assistance of Anthony Dudo, under the Center for
Nanotechnology in Society at Arizona State University. I would like to acknowledge his help for the data
collection.
2
30
These stages in nanotechnology coverage should not be surprising. In the late 1970s,
major elite newspapers that set the national agenda began to offer science sections to readers. In
1978, for example, the New York Times introduced the weekly “Science Times” section that was
published on Tuesdays. The Science Times proved popular with readers, boosting circulation on
Tuesdays (Diamond, 1994; Gelb, 2003; Wilford, 2003). The introduction of the science sections in
major elite newspapers in the late 70s/early 80s may explain why nanotechnology first
appeared in the high circulation newspapers during that time. Many science-related articles of
this type are syndicated to medium and low circulation newspapers from elite newspapers such
as the New York Times (Clark & Illman, 2006). As such, when scientific issues gained prominence
in the elite newspapers in the 80s and the newspaper editors started to assign designated
journalists to cover science beats, these practices had a trickle-down effect on medium and low
circulation newspapers in the U.S.
While these results show that nanotechnology gained more prominence in newspaper
coverage, it says little about how this emerging technology was covered. To gauge the tone of
coverage over the years, a full Lexis-Nexis search for risk-related articles across all the 21 high,
medium, and low circulation U.S. newspapers was conducted. The search term “risk*”3 was
added to the search term used for the amount of coverage of nanotechnology (see Appendix B)
to assemble risk-related nanotechnology articles (in the full text) between 1 January 1999 and 31
August 2008. Figure 2.15 shows the percentage of risks-related nanotechnology articles in the
newspapers between 1999 and August 2008. Even though the number of news articles about
nanotechnology had been climbing steadily, the percentage of risk-related articles remained
rather low. With the exception of a slight peak in 2006, the proportion of risks-related articles in
This cursory assessment of risks-related articles was far from being comprehensive. Future studies
should expand on the search terms (e.g., “danger,” “threat,” etc.) to provide a more comprehensive
examination of the risks-related nanotechnology news articles in Lexis-Nexis.
3
31
relation to all the articles was below 20 percent between 1999 and August 2008. This suggests
that media coverage of nanotechnology has been overwhelmingly positive or at least has not
paid attention to risks, highlighting the benefits of the emerging technology over its risks.
Overall, this descriptive analysis shows the number of news articles on nanotechnology
has been negligible between 1969 and early 90s, but coverage rose steadily from the late 90s
onwards. Not only that, nanotechnology has evolved from a purely elite issue to a local issue, as
exemplified by the trend of coverage in the media. Finally, when it comes to the tone of media
coverage about nanotechnology, it is patent that coverage has been overwhelmingly positive, as
the proportion of risks-related news stories about nanotechnology has consistently remained
less than 30 percent of the total nanotechnology coverage over the past 10 years. Despite this,
this descriptive analysis presents a rudimentary snapshot about media coverage. A more
comprehensive content analysis with the aid of at least three human coders should be able to
provide a more detailed tone of nanotechnology coverage in the mass media.
32
CHAPTER 3
EFFECTS OF VALUE PREDISPOSITIONS, MASS MEDIA, AND COGNITIVE PROCESSING
ON PUBLIC ATTITUDES TOWARD NANOTECHNOLOGY: TESTING MODERATING AND
MEDIATING MECHANISMS
(STUDY 1)
As elucidated in Chapter 2, understanding American public opinion towards
nanotechnology is pertinent to the advancement of the emerging science in the United States.
What are the key factors that influence public attitudes toward nanotechnology? How do the
public reach democratic decisions about funding support for nanotechnology? In response to
these questions, Chapter 3 therefore undertakes the task of examining the mechanisms
underlying these important attitudinal outcomes: (1) public perceived risks-versus-benefits of
nanotechnology and (2) public attitude towards federal funding of the emerging technology, by
situating the queries within the current debate between the “scientific literacy model” (Miller et
al., 1997) and the “cognitive miser model” (Fiske & Taylor, 1991). Specifically, this chapter
examines the degree to which key factors, including value predispositions, science media use,
and cognitive information processing (in the form of “reflective integration”) influence public
attitudes toward nanotechnology.
More importantly, the second aim of this study is to bridge two theoretical
communication models: the “differential gains model” (Scheufele, 2001) and the “cognitive
mediation model” (Eveland, 2001, 2002; Eveland, Shah, & Kwak, 2003). Based on these
theoretical models, I will systematically examine the extent to which reflective integration both
moderates and mediates the impact of science media use on the two attitudinal outcome
variables of interest (i.e., perceived risks-versus-benefits of nanotechnology and support for
33
federal funding of nanotechnology). Thus, instead of a simple direct media effects model, this
study posits that the relationships between science media use, reflective integration, and the
attitudinal outcomes of nanotechnology are far more complex than previously assumed. My
ultimate purpose is to build a more complete model of public opinion formation about
nanotechnology.
I will begin by providing a detailed concept explication of the attitudinal outcome
variables (i.e., perceived risks-versus-benefits of nanotechnology and attitude towards federal funding of
nanotechnology) and the independent variables of interests (i.e., value predispositions, science
media use, reflective integration, and factual scientific knowledge), followed by the theoretical
arguments for their potential direct, indirect, and interactive relationships with the outcome
variables. In addition, the hypotheses with regard to the relationships among the independent
and outcome variables will also be introduced in this chapter.
3.1. Outcome Variables
3.1.1. Perceived Risks-versus-Benefits (Attitudinal Outcome Variable 1)
For the purpose of this study, perceived risks-versus-benefits (instead of perceived risks
in isolation) will be examined as the outcome variable. Given the fact that the “real” risks are
not apparent for nanotechnology at the current stage of its development, and public opinion
and media coverage of this emerging technology is overwhelmingly positive, simply examining
perceived risks without consideration for the perceived benefits of the technology would
preclude us from gaining a full understanding of public attitudes.
Public opinion surveys conducted in the U.S. have shown that the public differ
significantly among themselves when it comes to perceived benefits of nanotechnology but not
perceived risks. As evidenced by a study conducted by Scheufele and Lewenstein (2005),
34
members of the public who are aware of nanotechnology perceived significantly greater
benefits than those who are unaware of the emerging technology; conversely, differences in
perceived risks between those who are aware and those who are unaware are negligible. This
provides the rationale for examining perceived risks and benefits in tandem.
According to another public opinion survey conducted by Currall, King, Lane, Madera,
and Turner (2006), the public do not consider the risks or benefits of nanotechnology
independently. For instance, the effect of public perceived benefits on the use of
nanotechnology applications was more pronounced when risks were lower than when risks
were high. In addition, when the benefits were low, consumers were more concerned about
risks than when the benefits were high. Currall et al. concluded that risks and benefits are both
enmeshed in a complex decision-making calculus, and argued for a more balanced approach,
where potential benefits and risks are addressed together.
Moreover, media coverage of nanotechnology in the U.S. has remained overwhelmingly
positive over the last ten years. A content analysis conducted by Gaskell, Ten Eyck, Jackson, and
Veltri (2004, 2005) found that even though risk coverage of nanotechnology in the New York
Times has increased from 1999 to 2003, the overall proportion of risk-related news has been very
small compared to the proportion of benefit-related coverage.
In another content analysis of U.S. newspapers and wire services published between
January 2000 and December 2004, Friedman and Egolf (2005) showed that health and
environmental risks related to nanotechnology did not dominate news coverage. In fact, they
found that most of the articles were balanced, describing risks with both positive and negative
information, and they concluded that the mild concern about risks in the news media clearly
does not counterbalance all the positive stories about the benefits and promises of
nanotechnology found by other scholars (e.g., Stephens, 2005). Friedman and Egolf (2005)
35
pointed out that “from this analysis, it does not appear that these U.S. or U.K. newspapers and
wire services published articles from 2000 to 2004 that would negatively influence public
opinion about nanotechnology” (p. 10).
Moreover, as previously discussed in Chapter 2, a simple Lexis-Nexis search revealed
that despite the steady increase in the number of news articles about nanotechnology between
1999 and August 2008, the proportion of risk-related articles remained very low. This offers
further justification for assessing perceived risks-versus-benefits instead of just variances in
perceived risks among the public.
Since previous studies on public reactions to nanotechnology have followed a similar
distinction (e.g., Lee et al., 2005), the attitudinal outcome variable for this study will be
constructed by subtracting perceived benefits of nanotechnology from perceived risks of
nanotechnology, with higher scores indicating greater perceived risks (see “Methods” section
for a detailed description on the construction of this outcome variable).
3.1.2. Support for Federal Funding of Nanotechnology (Attitudinal Outcome Variable 2)
The second attitudinal outcome variable of interest in this chapter is public support for
federal funding of nanotechnology research. The decision to focus on this aspect of public
acceptance of nanotechnology is twofold. First off, public support has become increasingly
important in order to sustain federal funding initiatives and maintain general support for
science and technology in the political area (Roco & Bainbridge, 2003). As such, the importance
of federal funding initiatives is itself a worthwhile outcome variable that should be examine indepth. Second, support for federal funding of nanotechnology is an issue that is close at heart to
the public and is one with practical implication that directly involves the taxpayers’ money.
Therefore, it may be a straightforward task for the public to form opinion about support for
federal funding of nanotechnology per se.
36
3.2. Value Predispositions as Heuristic Cues in Opinion Formation
Based on findings from recent empirical studies that examined the influence of value
predispositions on science and technology (e.g., Brossard & Nisbet, 2007; Brossard et al., in
press; Ho et al., 2008), I identify religious beliefs, deference to scientific authority, and trust in
scientists as values that could potentially influence both key attitudinal outcome variables. In
addition, I distinguished trait-like value predispositions (i.e., religious beliefs and deference to
scientific authority) from state-like dispositions (i.e., trust in scientists) in this dissertation. I will
explain this distinction later on in Section 3.8.
3.2.1. The Role of Religious Beliefs
Since the majority of the Americans are unfamiliar with nanotechnology, people are
likely to use religious guidance as a heuristic cue to form judgments about the emerging
technology. Recent research has shown that religious guidance is one of the major factors
driving public resistance to science generally and to other emerging technologies specifically
(Brossard et al., in press; Gaskell, Einsiedel et al., 2005; Ho et al., 2008; Nisbet, 2005; Parrott, Silk,
Krieger, Harris, & Condit, 2004).
This is hardly surprising given the historical intransigence and normative
inconsistencies between religion and science (Brooke, 1998; Miller et al., 1997). The conflict
between religion and science is as old as the age of Enlightenment (Nelkin, 1979; Toulmin &
Goodfield, 1965). Valenti (2002) aptly summarized the tension: “Religion requires faith—belief
without question. Science demands we take nothing on faith, reject any anecdotal evidence.
How might these seemingly opposed disciplines collaborate to improve public understanding
of science and impact pending policy making without undermining spiritual well being?” (p.
58)
37
One explanation for this tension has to do with the perception that science tampers with
nature or is akin to playing God (Sjoberg, 2004; Sjoberg & Winroth, 1986), putting it at odds
with religious beliefs. For instance, Gaskell et al. (2000) showed that people perceive genetically
modified technology as interfering with nature and natural processes, and therefore see the
technology as risky and possibly immoral. Furthermore, those respondents who held strong
religious beliefs in Gaskell et al.’s data were more likely to show strong opposition to scientific
research that involved human beings than those who held weaker religious beliefs.
Nanotechnology is not spared from the potential friction between religion and science.
The U.S. Food and Drug Administration officially defined “nanotechnology” as part of the
Nano-Bio-Info-Cogno (NBIC) technologies that highlight the unity of nature at the nanoscale,
and the intelligible processes of evolution that have constructed life and intelligence, from the
nanoscale, without divine intervention (Bainbridge, 2003; Sententia, 2004). Sententia (2004)
pointed out that developers of NBIC technologies face a multitude of obstacles, including
political, disciplinary, and religious sectarianism. Bainbridge (2003) argued that this allinclusive approach to nanotechnology may go against people’s religious beliefs, affecting their
perceived risks-versus-benefits and reducing their support for nanotechnology in future.
Using a representative U.S. public opinion survey conducted in 2004, Brossard et al. (in
press) found a direct and negative relationship between strength of religious beliefs and
support for funding of nanotechnology, and they concluded that people use religiosity as a
attitudinal filter when it comes to forming opinions about the new technology. Religious people
may lump nanotechnology, biotechnology, and stem cell research together and perceive them
as means to enhance human qualities. In short, some people may believe that researchers are
“playing God” when they create materials that do not occur in nature, especially where
38
nanotechnology and biotechnology intertwine. Based on these considerations, the following
hypotheses are put forth:
Hypothesis 1a: Strength of religious beliefs will positively predict public perceived risksversus-benefits of nanotechnology.
Hypothesis 1b: Strength of religious beliefs will negatively predict public support for
federal funding of nanotechnology.
3.2.2. Deference to Scientific Authority as Heuristic Shortcut
Deference to scientific authority is another value predisposition that can affect attitudes
toward science and technology (Brossard & Nisbet, 2007; Ho et al., 2008). As defined by
Brossard and Nisbet (2007), deference to scientific authority is “a long-term socialized trait that
guides citizens’ responses to a range of technical controversies” (p. 10). They demonstrated that
the more individuals defer to scientific authority, the more likely they were to hold positive
views on controversial scientific issues such as agricultural biotechnology. Likewise, Ho,
Brossard, and Scheufele (2008) showed that deference to scientific authority is positively
associated with public support for human embryonic stem cell research.
The American educational system has instilled a strong sense of respect for scientists
and scientific institutions among the citizens, and this has fostered a culture of deference to
scientific authority in the U.S. These have been reflected in education that involved teaching
people to view scientific research as solitary activities that are kept away from external social
and political pressures (Bimber & Guston, 1995), and to perceive science as a pure and unbiased
pursuit that increases our knowledge about the world (Irwin, 2001).
Deference to scientific authority evolves over a long period of time from an inherent
unequal power relationship between the lay public and the highly knowledgeable scientists and
the relevant scientific institutions. Extending this notion to the current study, deference to
39
scientific authority may be a value predisposition that is in direct conflict with religious beliefs
in influencing individuals’ views about nanotechnology. Hence, the following hypotheses are
posited:
Hypothesis 2a: Deference to scientific authority will negatively predict public perceived
risks-versus-benefits of nanotechnology.
Hypothesis 2b: Deference to scientific authority will positively predict public support for
federal funding of nanotechnology.
3.3. Science Media Use and Opinion Formation
For most Americans, television, the Internet, and newspapers continue to be the primary
sources of information about science and technology (Pew Internet & American Life Project,
2006). Both the content and valence of science in the mass media have been demonstrated to
perform a crucial role in shaping public attitude toward science and technology (Ho et al., 2008;
Nisbet et al., 2003; Nisbet & Lewenstein, 2002). When it comes to media content, I have shown
in Chapter 2 that news coverage of nanotechnology has emphasized its positive prospects. And,
as mentioned earlier numerous content analyses have demonstrated that news coverage of
nanotechnology highlighted more benefits than risks in the U.S. media (Friedman & Egolf, 2005;
Gaskell et al., 2004; Stephens, 2005). In a content analysis of the New York Times from 2000 to
2003, Gaskell et al. (2004) found an overwhelming coverage of benefits than risks for
nanotechnology, and concluded that “media coverage is more slanted towards a supportive
culture of science and technology in the U.S.” (p. 496)
Likewise, by examining nanotechnology coverage in major U.S. and non-U.S.
newspapers published from 1988 through 2004, Stephens (2005) found that the proportion of
articles in which benefits outweighing risks (versus risks outweighing benefits) is three to one.
40
In particular, Friedman and Egolf (2005) found that even when health and environmental risks
were covered in U.S. newspapers, most of the articles published from 2000 to 2004 were
balanced and described risks with both positive and negative information. The researchers
concluded that news coverage in the U.S. would not negatively influence public opinion about
nanotechnology.
Besides the direct information from the mass media on the development of
nanotechnology, some communication researchers have argued that the nature or tone of media
coverage of nanotechnology can serve as a simple decision rule in influencing the risks and
benefits considerations among the public (Nisbet & Scheufele, 2007; Scheufele & Lewenstein,
2005). Evolving from cross-disciplinary research in economics (Kahneman & Tversky, 1979),
psychology (Kahneman & Tversky, 1984), and sociology (Goffman, 1974), framing is defined by
communication scholars as selecting “…some aspects of a perceived reality and make them
more salient in a communicating text, in such a way as to promote a particular problem
definition, causal interpretation, moral evaluation, and/or treatment recommendation”
(Entman, 1993, p. 52).
In terms of news coverage, how audiences interpret information is a function of how a
particular story is presented by journalists (Scheufele, 1999). In other words, framing effects
occur when audiences interpret two news stories differently based simply on the stories’ modes
of presentation (i.e., variations based on visual or terminological tools), even if the exact same
information is presented. Media frames provide audiences with cognitive shortcuts for
efficiently processing new information, especially for issues that audience members are
unfamiliar with.
In addition, experimental studies have demonstrated that framing of nanotechnology
has an effect on how audience perceived risks and benefits of the technology (e.g., Cobb, 2005;
41
Schutz & Wiedemann, 2008). For example, Cobb (2005) found that stories highlighting multiple
benefits resulted in a larger proportion of respondents agreeing with the statement that
“benefits will outweigh the risks.” Particularly in the area of emerging technologies, where most
citizens have little or no direct experience, media coverage of these technologies provides a key
heuristic to the audience (Ho et al., 2008; Nisbet et al., 2003; Nisbet & Lewenstein, 2002).
In essence, then, the mass media play a dual function in science communication. On one
hand, the media are information providers that offer a source of informal learning about
emerging science, such as nanotechnology, for most Americans. On the other hand, media
frames such as the positive tone of coverage about nanotechnology offer audience the heuristic
cues to make quick decisions about the technology (Scheufele & Lewenstein, 2005). In other
words, individuals gain a hybrid of information and news frames from the mass media.
In fact, extant research has shown that science media use had a negative and strong
effect on perceived risks-versus-benefits of various technologies (Durant, Evans, & Thomas,
1992; Lee et al., 2005; Nelkin & Lindee, 1995). Moreover, mass media has been consistently
shown to influence risk judgment as the outcome variable in numerous risk communication
contexts (Coleman, 1993; Fleming, Thorson, & Zhang, 2006; Morton & Duck, 2001; Slater &
Rasinski, 2005; Tyler & Cook, 1984). This finding is not surprising since a majority of the public
exposure to nanotechnology comes from media outlets, such as television, movies, and books
(Castellini et al., 2007), and since public level of attention to a specific issue has been shown to
follow variation in media attention (Ho, Brossard, & Scheufele, 2007).
Given the overall positive content and valence of the news media on nanotechnology
over the past few years, this study contends that science media use will negatively influence
individuals’ perceived risks-versus-benefits of nanotechnology, and concomitantly, drive
42
individuals’ support for federal funding of nanotechnology. The following hypotheses are
postulated:
Hypothesis 3a: Science media use will negatively predict public perceived risks-versusbenefits of nanotechnology.
Hypothesis 3b: Science media use will positively predict public support for federal
funding of nanotechnology.
By “science media use,” I am referring to public attention to science content in the news
media. It is important to note that attention is different from mere exposure, as the former
denotes the audience’s conscious effort to direct concentration to content in the news media.
Research that have conceptualized mass media use as audience attention tend to find larger
media effects than those that use audience exposure to the mass media as a construct (e.g.,
Chaffee & Schleuder, 1986). For example, using data from a two-year longitudinal study of
adolescents and their parents, Chaffee and Schleuder (1986) demonstrated that survey measures
of attention to newspaper and television news were reliable and stable over time. Not only that,
the researchers showed that attention to news media contributed more to public knowledge
gain than mere exposure to news in the mass media.
Therefore, science news media use hereafter refers to the amount of attention that the
public pays to science news in the media. Empirical studies – especially in the area of political
communication – have shown that television news, print news, and online news all contribute
to political information gain in the public, albeit to varying extents (e.g., Chaffee & Frank, 1996;
Chaffee & Kanihan, 1997). As such, science news media use as a construct is treated as
comprising of television news, print news, and online news in this dissertation.
43
3.4. Cognitive Processing: Reflective Integration and Learning from the News Media
Media effects may be much more complex than it appears. Extensive research has
demonstrated that audience are not passive consumers of media information; instead, driven by
their sense of needs and motivations, audiences are active consumers, capable of selecting,
processing, and integrating information that they gather from the mass media (Blumler & Katz,
1974). Extending this idea to the current study, audience cognitive involvement and processing
of media information may play a pertinent role in the amount of new information about
nanotechnology gained from the mass media, and ultimately influence public attitudes toward
the emerging technology. Cognitive processing can influence the public level of cognition and
attitudes.
The central argument of this chapter is that cognitive processing will affect public
attitudes toward nanotechnology. Kosicki and McLeod (1990) defined cognitive informationprocessing strategies as “tactics that individuals use to try to cope with the amount and kind of
mass media information that they encounter in their everyday lives” (p. 73). Most people are
limited information processors who use media to help them attain certain goals. Given the
limited amount of time, people tend to adopt certain processing tactics to sift out and make
sense of messages that are useful to them.
Kosicki and McLeod (1990) first identified three common strategies that individuals use
to deal with the flow of information from the mass media. These strategies include: selective
scanning, active processing, and reflective integration. Selective scanning refers to individuals’ effort
to use mass media only to seek information relevant to them and to tune out content that is
irrelevant or not of interest to them. Due to the limited amount of time and resources that
individuals have, they usually employ selective scanning to cope with the large volume of
mediated information available to them. Active processing refers to individuals’ attempts to
44
make sense of news stories based on their own needs, by going beyond the exact information
given by the media and by seeking out additional sources. Individuals usually engage in active
processing when they assume that mass mediated information in general is incomplete, biased,
or in other ways tinted by the intentions of the communicator. Finally, reflective integration refers
to individuals’ efforts to think about the information that they gather from the mass media, as
well as their attempts to talk to others about what they have learnt from the mass media in
order to fully comprehend the newly acquired information.
Four key assumptions underlie the three cognitive processing strategies. According to
Kosicki and McLeod (1990), the assumptions are as follows:
(1) People actively use media and over time find various strategies of coping
with information very useful. These strategies are necessary because people are
limited in their ability to process information and have limited time to devote to
media;
(2) These mental strategies are manifested in everyday habits or behaviors.
People continue to use these strategies over time because they find them helpful.
However, people may use differing amounts of them, in different patterns;
(3) People are able to monitor their thought processes and strategies about
thinking, as well as verbalize about them; and
(4) These mental operations relate to interaction with media generally, and so cut
across the use of both newspapers and television. Although there may be
considerable cyclical fluctuations, patterns of information processing strategies
are sufficiently stable for a given media genre (e.g., political news) so as to merit
their consideration as molar, recurrent strategies. (p. 74)
45
Put differently, this perspective assumes that people are capable of processing
information and that they develop stable strategies to cope with and make sense of the large
amount of messages provided by the mass media (Fredin, Kosicki, & Becker, 1996; McLeod,
Kosicki, & Pan, 1991). The processing strategies are consistent across individuals and are
reflected in their daily behavior when they deal with mediated information (Fredin et al., 1996;
Kosicki & McLeod, 1990; McLeod et al., 1991). Therefore, these information-processing
strategies are under individuals’ volitional control.
As reflective integration has been fully explicated and extensively studied in numerous
communication contexts (e.g., Eveland, 2001, 2002; Eveland et al., 2003; Scheufele, 2001, 2002),
this dissertation will leverage on this area of research by specifically examining the impact of
reflective integration on public level of scientific knowledge and attitudes toward
nanotechnology. Originally, Kosicki and McLeod (1990) defined reflective integration as:
“…the postexposure salience of information such that it occupies the mind and is
the subject of interpersonal communication. The key, however, is the
incorporation of new information into the person’s existing cognitive framework
for understanding the subject.” (p. 75-76, emphasis added)
Simply put, the concept of reflective integration includes pondering and thinking about
a specific issue covered in the mass media and talking about it with others. As such, Kosicki and
McLeods’ (1990) initial operationalization of reflective integration included items tapping the
degree to which individuals pondered and thought about news stories, as well as the frequency
with which they engaged in interpersonal discussions to see what others thought.
46
Over the next decade, Eveland (2001, 2002) re-examined the concept of reflective
integration in a number of data sets, assessing the internal consistency and stability of the
original items proposed by Kosicki and McLeod (1990), as well as the construct validity of the
scales formed from these items. Based on these reliability and validity tests, he modified and
expanded reflective integration into two separate dimensions: (1) “elaborative processing” or
news elaboration (Eveland, 2001, 2002; Eveland et al., 2003), which refers to the process by
which people ponder, try to understand, elaborate on, and make sense of the news content; and
(2) “interpersonal discussion,” which refers to talking about mass-mediated messages about a
certain issue with others so as to process news content more carefully by connecting it with
preexisting knowledge (e.g., Eveland & Thomson, 2006). Consequently, the operational
definitions were expanded to include two distinct sets of items that were used to measure
elaborative processing and interpersonal discussion respectively.
It is important to highlight that the notion of reflective integration, as well as the
differential gains model (Scheufele, 2001, 2002) and the cognitive mediation model (Eveland,
2001, 2002) that I will elaborate on later, are intertwined with the production of scientific
knowledge. As such, in the following sections involving reflective integration, a more finegrained assessment of its influence is warranted. Specifically, I will systematically examine its
direct, indirect, and additive influence on both the cognitive outcome (i.e., factual scientific
knowledge) and the attitudinal outcomes (i.e., perceived risks-versus-benefits of
nanotechnology and support for federal funding of nanotechnology).
3.4.1. Effect of Elaborative Processing on Cognitive and Attitudinal Outcomes
Elaboration, as a measure of cognitive involvement, is a behavioral style that people use
to associate new ideas and information with what is already known, look for similarities with
past experiences, and find ways to apply the information (Eveland, 2002). As Eveland argues,
47
“[b]y engaging in these elaborations on the content provided in the newscast, the viewer will
increase the strength of memory store as well as the ability to recall the story through more
numerous mental pathways” (Eveland, 2001, p. 573). While news elaboration has been widely
applied in political communication, this study argues that it is also applicable to science topics
and scientific knowledge. Any new information incorporated into a pre-existing knowledge
structure through the process of reflective integration will increase the level of scientific
knowledge and will be easily accessible for formulating judgments about the risks and benefits
of nanotechnology and nanotechnology acceptance.
In addition to scientific knowledge, this similar relationship can also be extended to
scientific attitudes. Since information about nanotechnology has been overwhelmingly covered
in positive terms in the media, it is expected that more benefits than risks will be perceived and
greater support for federal funding of nanotechnology will be reported, if people employ
elaborative processing. The following relationships for both the cognitive and attitudinal
outcomes are hypothesized:
Hypothesis 4a: Elaborative processing will positively predict factual scientific knowledge.
Hypothesis 4b: Elaborative processing will negatively predict public perceived risksversus-benefits of nanotechnology.
Hypothesis 4c: Elaborative processing will positively predict public support for federal
funding of nanotechnology.
3.4.2. Effect of Interpersonal Discussion on Cognitive and Attitudinal Outcomes
As mentioned earlier, the second dimension of reflective integration is interpersonal
discussion (Scheufele, 2001), in which people make sense of the information they gather from
the mass media by talking to other people about these issues, discussing the pros and cons, and
weighing alternatives to reach a conclusion. The potential impact of interpersonal
48
communication on knowledge, as well as judgments and attitudes, is not new. In the area of
political communication, scholars have shown that talking about certain issues with other
citizens will enable people to understand these issues in all their complexity, connecting them
with other preexisting knowledge, and consequently make informed judgments (e.g., Scheufele,
2002). Scheufele (2002) argued that discussion might be to a large degree about sharing
experiences and applying mass-mediated information to the real world, and people who
process news content more carefully by talking it over with others are also more likely to extract
relevant pieces of knowledge. Furthermore, research has shown that people who have
conversations with others are more likely to understand news better (Robinson & Levy, 1986).
Researchers have also shown that interpersonal discussions in which disagreement occurs tend
to stimulate the greatest amount of cognitive activity (Levine & Russo, 1995) and provide
individuals with opportunities to learn about one another and about reasons for their
conflicting opinions (Gamson, 1992).
Applying this conception to the current study, people who engage in scientific
discussion with others would also be more likely to retrieve relevant pieces of science-related
information that they have gathered from the mass media. Johnson (1993) pointed out that
discussions with family, friends, neighbors, and co-workers are likely to reinforce mass media
effects. Since the media has on the most part portrayed nanotechnology and science in general
favorably, interpersonal discussion about science and nanotechnology should reinforce this
perspective.
In the area of risk communication, numerous studies have shown that interpersonal
communication heightens perceived risks, especially at the personal level (Coleman, 1993;
Dunwoody & Neuwirth, 1991; Griffin & Dunwoody, 2000; Morton & Duck, 2001). However,
Dunwoody and Neuwirth (1991) also noted that interpersonal discussion will only increase
49
perceived risks provided that the risk about a particular issue is made salient in the mass media.
For nanotechnology that is still at the early stage of the issue attention cycle, the risk aspect of
the issue is not prominent in the mass media. In that respect, interpersonal communication will
only reinforce media messages, in which individuals who discuss science more frequently will
also perceive lesser risks and more benefits about nanotechnology than will those who discuss
science less frequently. Therefore, drawing from existing risk communication and political
communication literature, this study argues that interpersonal communication will be positively
associated with individuals’ level of scientific knowledge, negatively associated with
individuals’ perceived risks-versus-benefits of nanotechnology, and positively associated with
individuals’ support for federal funding of nanotechnology.
Based on these considerations, the following hypotheses on cognitive and attitudinal
outcomes are put forth:
Hypothesis 5a: Scientific discussion will positively predict factual scientific knowledge.
Hypothesis 5b: Scientific discussion will negatively predict public perceived risks-versusbenefits of nanotechnology.
Hypothesis 5c: Scientific discussion will positively predict public support for federal
funding of nanotechnology.
3.5. The “Differential Gains Model” – Moderating Role of Reflective Integration on
Cognitive and Attitudinal Outcomes
Based on the fundamental arguments of the differential gains model (Scheufele, 2001),
this chapter posits that reflective integration, including elaborative processing and scientific
discussion, will moderate the effect of attention to science news on public level of scientific
knowledge and attitudes toward nanotechnology, above and beyond the main effects of mass
50
media use and reflective integration. The term “intrapersonal reflection” refers to the statistical
interaction between science media use and elaborative processing on cognitive and attitudinal
outcomes, and the term “interpersonal reflection” refers to the interaction between science
media use and talking about scientific issues with others on cognitive and attitudinal outcomes.
News attention, or the tendency to focus mentally on specific content during exposure to
news, is a necessary but not sufficient condition for elaboration on that same content. The
impact of mass-mediated information on individuals’ understanding of the scientific world and
ultimately their judgments about specific scientific issues should be highest if they pay attention
to relevant information in the mass media and – at the same time – ponder mass-mediated
information and try to integrate it into their existing cognitive frameworks. Likewise, a similar
maximal effect of media information on individuals’ knowledge and attitudes toward science
and technology should occur if they pay attention to relevant scientific news stories and talk
about it with other people, learn about other ways of thinking about the issue, and, ultimately,
develop a better understanding of the problem and possible ways of solving it. Elaborative
processing, scientific discussion, or science media use alone should produce weaker effects. If a
person neither uses media nor engages in any form of elaborative processing or scientific
discussion, the effects should be minimal. In fact, Brossard and Kim (2007) have shown that
pondering media messages and talking to others during and after media exposure play a
significant additive role in predicting scientific knowledge. Therefore, the following
relationships between intrapersonal and interpersonal reflection on the cognitive outcome are
postulated:
Hypothesis 6a: Elaborative processing will moderate the effect of science media use on
factual scientific knowledge.
51
Hypothesis 6b: Interpersonal discussion about science will moderate the effect of science
media use on factual scientific knowledge.
In addition, this dissertation argues that reflective integration can promote a deeper
understanding of the scientific world and provide a stronger cognitive base and a more
sophisticated knowledge structure for opinion formation about scientific issues than mere
factual, textbook-style scientific knowledge. By sophisticated knowledge, I am referring to the
ability of individuals to associate, integrate, and relate various news issues or topics, which will
also include the knowledge of how concepts within a specific domain are interrelated.
It therefore makes sense to deduce that people who pay attention to news and actively
engage in reflective integration should possess greater cognitive sophistication about the
scientific world. In turn, they should perceive greater benefits than risks for nanotechnology
and offer greater support for funding of the emerging technology. The following hypotheses on
the attitudinal outcomes are postulated:
Hypothesis 7a: Elaborative processing will moderate the influence of science media use
on public perceived risks-versus-benefits of nanotechnology.
Hypothesis 7b: Elaborative processing will moderate the influence of science media use
on public support for federal funding of nanotechnology.
Hypothesis 8a: Scientific discussion will moderate the influence of science media use on
public perceived risks-versus-benefits of nanotechnology.
Hypothesis 8b: Scientific discussion will moderate the influence of science media use on
public support for federal funding of nanotechnology.
52
3.6. The “Cognitive Mediation Model” – Mediating Role of Reflective Integration on
Cognitive and Attitudinal Outcomes
In addition to the additive effects of science media use and reflective integration on
cognitive and attitudinal outcomes, this chapter contends that reflective integration has a
mediating role on scientific knowledge, and consequently, on public perceived risks-versusbenefits and support for federal funding of nanotechnology. This section draws on the cognitive
mediation model (Eveland, 2001, 2002) to provide theoretical support for this argument.
Briefly, Eveland’s (2001, 2002) cognitive mediation model of learning from the news
posits three theoretical claims. First, the model proposes that motivations for gratifications
seeking should lead individuals who expose themselves to the news media to engage in
information-processing strategies that will enable them to achieve their learning goals.
Specifically, news attention and elaboration of news content are the two types of information
processing strategies that individuals would use. Second, the model assumes that people must
first pay attention to the news content before they could engage in elaborative processing. In
other words, news attention should be considered antecedent to elaborative processing of news
in the model. Finally, greater attention to the news and elaborative processing of news content
should both lead to higher levels of learning from the news.
In its simplest form, the cognitive mediation model contains six key linkages:
gratifications to news attention, gratifications to elaboration, attention to elaboration,
gratifications to knowledge (expected to be non-significant after all controls are applied),
attention to knowledge, and elaboration to knowledge. Numerous studies have shown
empirical support for the cognitive mediation model (e.g., Eveland, 2001, 2002; Eveland et al.,
2003).
53
Besides cognitive activity in the form of thinking, activity in terms of talking is also
critical to increasing knowledge (Kosicki & McLeod, 1990; Scheufele, 2001, 2002), and
particularly scientific knowledge. Coming from a constructivist perspective of teaching science
and cooperative science learning, scholars have argued that both intrapersonal discussion (i.e.,
elaborative processing) and interpersonal discussions are key drivers of increasing knowledge.
Within a classroom context, students should engage in discussions with others so that cognitive
conflict could be addressed and inadequate reasoning could be altered (Belenky, Clinchy,
Goldberger, & Tarule, 1986; Driver, 1995; von Glasersfeld, 1995). For instance, a few studies in
the field of science education, have found that group discussion facilitates science learning.
Driver, Asoko, Leach, Mortimer, & Scott (1994) have also found that students who actively
engage in classroom discussion were more likely to increase their scientific knowledge. In a
recent study conducted by Brossard and Kim (2007), the researchers demonstrated that both
cognitive processing in the form of pondering media messages and interpersonal discussion
mediate the influence of science news use on scientific knowledge related to the issue of stem
cell research.
Extending the second part of the cognitive mediation model on the notion of news
attention as a precursor to elaborative processing in their impact on public learning from the
news, I therefore hypothesize the following for the cognitive outcome:
Hypothesis 9a: Elaborative processing will mediate the effect of science media use on
factual scientific knowledge.
Hypothesis 9b: Scientific discussion will mediate the effect of science media use on factual
scientific knowledge.
54
In addition to the above-mentioned indirect effects, this chapter also argues that
reflective integration will mediate the effect of science media use on public attitudes toward
nanotechnology. This is based on the idea that both news elaboration and interpersonal
discussion will engender a more sophisticated knowledge structure, going beyond mere factual
scientific knowledge. With this consideration in mind, the following hypotheses for the
attitudinal outcomes are postulated:
Hypothesis 10a: Elaborative processing will mediate the effect of science media use on
public perceived risks-versus-benefits.
Hypothesis 10b: Elaborative processing will mediate the effect of science media use on
public support for federal funding of nanotechnology.
Hypothesis 11a: Scientific discussion will mediate the effect of science media use on
public perceived risks-versus-benefits.
Hypothesis 11b: Scientific discussion will mediate the effect of science media use on
public support for federal funding of nanotechnology.
3.7. Effect of Factual Scientific Knowledge on Attitudinal Outcomes
As mentioned earlier, proponents of the scientific literacy model argue that people have
a knowledge deficit about science; only through increased in scientific knowledge can people
come to appreciate the workings of science and form positive opinions about new technologies
(Miller, 1998, 2004). However, evidence for the knowledge-attitude link has been scarce, as
many studies have found small or almost negligible relationship between scientific knowledge
and attitudes toward science (Brossard & Nisbet, 2007; Nisbet, 2005; Scheufele & Lewenstein,
2005).
55
As a result, the scientific literacy model has received criticisms (e.g., Allum et al., 2008;
Priest, 2001). However, the effect of scientific knowledge on public attitudes should not be
completely discounted. First off, researchers may not have fully captured the multidimensionality of scientific knowledge, with respect to both its conceptualization and
operationalization. Most indicators in previous studies, at best, measured public level of
textbook-style, factual scientific knowledge. For instance, Miller (1998) stated that “…scientific
knowledge encompasses the comprehension of a vocabulary of basic scientific constructs
sufficient to read competing views in a newspaper or magazine…understanding of the process
or nature of scientific inquiry…[and] some levels of understanding of the impact of science and
technology on individuals and on society” (p. 204).
The first dimension of Miller’s concept of knowledge is measured using a number of
true-false questions about scientific constructs considered to be the standard of ideal knowledge
(e.g., “The earth goes around the sun once per year” and “Antibiotics kill bacteria but not
viruses”). The second dimension is measured through assessments of the respondents’
understanding of experimental logic and probability, in addition to open-ended questions on
“what it means to study something scientifically” (Miller, 2004). This conceptualization and
operationalization of scientific knowledge are unsatisfactory as there are other possible
dimensions of knowledge, especially more sophisticated knowledge structures that are left
untapped. Therefore, it comes with little surprise that most empirical studies employing these
indicators found small to negligible effect of scientific knowledge on public attitudes toward
science and technology.
Put differently, studies have failed to explicate the different types of knowledge, in
which the problem may not lie in scientific knowledge itself. Scientific knowledge extends
beyond the simple learning of “facts” that can be straightforwardly defined and measured
56
(Irwin & Wynne, 1996). Therefore, there is no reason to assume in consequence that scientific
knowledge does not have an additional and independent effect, for reasons that are thus far not
clearly understood. Based on these considerations, the following hypotheses are postulated:
Hypothesis 12a: Factual scientific knowledge will negatively predict public perceived
risks-versus-benefits of nanotechnology.
Hypothesis 12b: Factual scientific knowledge will positively predict public support for
federal funding of nanotechnology.
3.8. The Role of Trust in Scientists
In addition to trait-like value predispositions, it is likely that the cognitive misers will
utilize short-term, state-like disposition such as trust, as a shortcut to form judgments about
nanotechnology. Trust refers to citizens’ willingness to rely on the endorsements of experts,
such as scientists and regulators, as well as institutions such as the federal government, to
manage risks associated with emerging technologies (Earle & Cvetkovich, 1995; Giddens, 1991;
Luhmann, 1979; Sztompka, 1999). Giddens (1991) pointed out that trust in a variety of abstract
systems is a necessary part of everyday life, and the characteristics of abstract systems imply
constant interaction with “absent others” – people we have never met but whose actions
directly affect our lives. Irwin and Wynne (1996) demonstrated that people were much more
concerned with whom to trust than with the scientific aspects of an issue itself. According to
risk communication scholars, trust acts as an uncertainty reduction mechanism, driving down
citizens’ concerns over the unforeseen risks and costs of emerging science and technologies
(Freudenburg, 1992, 1993; Slovic, 1999), thereby enabling citizens to act or form judgments
about emerging science and technology without understanding the risks involved.
57
In fact, numerous studies found trust in relevant actors to promote support for emerging
science such as biotechnology (Brossard & Nisbet, 2007; Brossard & Shanahan, 2003; Priest,
2001; Priest, Bonfadelli, & Rusanen, 2003; Sinclair & Irani, 2005), gene technology (Siegrist, 2000),
stem cell research (Ho et al., 2008), and nanotechnology (Lee et al., 2005). In particular, Priest et
al. (2003) found trust to be more important than scientific knowledge in predicting levels of
support for biotechnology in the U.S. and Europe. Citizens often substitute trust for knowledge
when forming attitudes about new technologies (Luhmann, 1979). Trust as a tool in decisionmaking is efficient when individuals have limited knowledge and personal experience, and
when they have little chance to anticipate the future consequences of a particular technology
(Olofsson, Ohman, & Rashid, 2006). This is highly applicable to the emerging nanotechnology
field with which most people are unfamiliar.
Although they seem alike, this dissertation argues that trust in scientists and deference
to scientific authority are two fundamentally different concepts. While deference to scientific
authority is a trait-like value predisposition that is general and applicable to a wide range of
scientific controversies, trust is a state-like quality that is specific to a particular science or
technology. For example, an individual may have high respect for scientific authority, but they
may not have high trust for scientists in the business industry who conduct research in
nanotechnology. Several science communication scholars have echoed this distinction. For
example, in the area of agricultural biotechnology, Brossard and Nisbet (2007) defined
deference to scientific authority as a long-term first-order orientation, whereas trust was treated
as a short-term second-order value disposition. The researchers found that as a long-term value
predisposition, deference to scientific authority tended to promote trust in scientists and
scientific institutions, and at the same time, both of these factors each had independent main
effects on public attitudes toward agricultural biotechnology.
58
Taking it as a distinct concept, I argue that trust in scientists will affect public perceived
risks and benefits, and drive public acceptance of nanotechnology. For instance, Siegrist (2000)
demonstrated that trust in companies and their scientists who perform gene manipulations is
related to public perceived benefits and risks, which ultimately influence public acceptance of
gene technology. Priest (2001) also showed that judgments about the levels of risk associated
with new technologies, such as bioengineered foods, are to a significant degree a function of
judgments about the trustworthiness of scientists and their employers. With respect to
nanotechnology, Lee et al. (2005) found that an increase in the level of trust in scientists
increases the level of support for this new technology. However, Lee et al. did not examine the
influence of trust in scientists on support for federal funding of nanotechnology in their study.
To address this research gap, the following hypotheses are posited:
Hypothesis 13a: Trust in scientists will negatively predict public perceived risks-versusbenefits of nanotechnology.
Hypothesis 13b: Trust in scientists will positively predict public support for federal
funding of nanotechnology.
3.8.1. Mediating Role of Trust in Scientists
As a state-like disposition, trust in scientists could mediate the influence of science
media use and elaborative processing on public attitudes toward nanotechnology. In terms of
how media influence public dispositions toward science, research has shown that newspaper
reading, science magazine reading, and science television viewing foster positive attitudes
toward science (e.g., Nisbet et al., 2002). Nisbet et al. (2002) demonstrated that general
newspaper and television use promoted belief in the promise of science and was negatively
related to reservations about science.
59
The positive news frames provided in the mass media about nanotechnology could
promote public trust in scientists; trust could, in turn, influence public attitudes toward the
emerging technology. Moreover, the favorable scientific information in the media about
nanotechnology could also promote trust in scientists. Therefore, the following hypotheses are
posited:
Hypothesis 14a: Trust in scientists will mediate the influence of science media use on
public perceived risks-versus-benefits of nanotechnology.
Hypothesis 14b: Trust in scientists will mediate the influence of science media use on
public support for federal funding of nanotechnology.
Hypothesis 15a: Trust in scientists will mediate the influence of elaborative processing on
public perceived risks-versus-benefits of nanotechnology.
Hypothesis 15b: Trust in scientists will mediate the influence of elaborative processing on
public support for federal funding of nanotechnology.
3.9. Effects of Perceived Risks-versus-Benefits on Support for Federal Funding of
Nanotechnology
In addition, this chapter contends that public perceived risks-versus-benefits will have
an impact on their decision-making about support and funding for nanotechnology. Coming
from the psychometric approach, Slovic (1987) defines risk perception as “the judgments people
make when they are asked to characterize and evaluate hazardous activities and technologies”
(p. 280). As discussed earlier, these public assessments of risks are subjective and emphasize
“qualitative” factors more than “quantitative” considerations (e.g., magnitude of harm and
acceptability are more important than probability). Public assessments of risk frequently do not
align with scientific assessments of risks (McComas, 2006). Scientific experts often view non-
60
experts as overestimating low-probability, large hazard risks, and having inconsistent attitudes
about various risks.
Research have shown that the public tend to perceive hazards or issues as risky if they
are not within their control (Starr, 1969), seem “dreadful” and “novel” (Fischhoff, Slovic,
Lichtenstein, Read, & Combs, 1978), and interfere with nature (Sjoberg, 2002). Perceived risks
have been found to be a function of other factors, such as gender, level of education, and
personal values (Covello & Sandman, 2001; Dunwoody & Neuwirth, 1991; Slovic, 1999), and the
mass media (as shown in Chapter 2). Regardless of these factors, the more individuals perceive
a hazard or a technology as risky, the less likely they are to accept it.
Numerous studies have found that perceived risks and benefits are associated with
levels of acceptance of technology (Frewer, Howard, & Shepherd, 1998; Siegrist, 2000; Siegrist,
Cvetkovich, & Roth, 2000; Sjoberg, 2002, 2004). For example, Siegrist (2000) demonstrated that
while perceived benefits was positively associated with acceptance of gene technology,
perceived risks was negatively associated with support for the technology. Sjoberg (2004)
opined that outright rejection of an emerging technology is often a function of perceived high
risks in the technology per se. Hence, I posit that perceived risks and benefits related to
nanotechnology are likely to affect support for funding of its research.
Hypothesis 16: Perceived risks-versus-benefits will negatively predict public support for
federal funding of nanotechnology.
61
CHAPTER 4
METHODS AND RESULTS (STUDY 1)
Chapter 4 will describe the methods and results of Study 1. Specifically, the methods
section will provide details on the data and sampling procedure employed, measures used, and
the analytical approaches for testing the hypotheses postulated in Study 1. I will elaborate on
and provide the rationale for the application of the multivariate techniques in this dissertation.
This will be followed by the results section, in which the findings for the hypotheses posited in
Study 1 will be reported.
4.1. Methods
4.1.1. Data and Sampling
Data for Study 1 came from a nationally representative random-digit-dial telephone
survey of 1,015 U.S. adult respondents age 18 and older conducted by the University of
Wisconsin Survey Center.4 The fieldwork was conducted from May to July 2007 with an
average length of 21.47 minutes per interview, and the approximate margin of error was +/- 3
percent. In order to minimize systematic non-response, significant amount of time and effort
were invested in call-backs and refusal conversions. The overall response rate for this survey
was 30.6 percent, using standard AAPOR response rate calculations (formula 3) that include
both refusals and unreachable but eligible telephone numbers.
The data were originally collected by Professor Dietram A. Scheufele, under grants support from the
National Science Foundation (SES-0531194) and the University of Wisconsin-Madison Graduate School
(135GL82). I would like to acknowledge his generosity in making these data available for my dissertation.
4
62
4.1.2. Measures
Respondents’ perceived risks-versus-benefits of nanotechnology and support for federal
funding of nanotechnology are the attitudinal outcome variables of interest in Study 1.
Independent variables include religious beliefs, deference to scientific authority, trust in
scientists, science media use, elaborative processing, science discussion, and factual scientific
knowledge. Finally, three demographic variables serve as control variables: age, gender, and
socioeconomic status. The variables measured with multiple items were constructed as the
averaged indices of the respective items. The index constructions were based on the results of
Cronbach’s alpha reliability for continuous variables and the KR-20 index for dichotomous
variables after a close examination of the descriptive statistics of each item. For variables
measured with only two items, the index constructions were based on the values of Pearson’s
correlation. Table 4.1 shows the descriptive statistics and the exact question wording from the
original survey questionnaire for each item considered in Study 1.
4.1.2.1. Attitudinal Outcome Variables
Support for federal funding of nanotechnology. The first attitudinal outcome variable was
measured using one item on a ten-point scale (1 = “do not agree at all,” 10 = “agree very
much”): “Overall, I support federal funding for nanotechnology” (M = 5.90, SD = 2.85).
Perceived risks-versus-benefits of nanotechnology. The construction of the second attitudinal
outcome variable took several steps. First, perceived risks of nanotechnology was created by
constructing an additive index of seven ten-point items (1 = “do not agree at all,” 10 = “agree
very much”): (a) “Nanotech may lead to the loss of personal privacy because of tiny new
surveillance devices,” (b) “Nanotech may lead to an arms race between the U.S. and other
countries,” (c) “Nanotech may lead to new human health problems,” (d) “Nanotech may be
used by terrorists against the U.S.,” (e) “Because of nanotech we may lose more U.S. jobs.,” (f)
63
“Nanotech may lead to the uncontrollable spread of very tiny self-replicating robots,” and (g)
“Nanotech may lead to more pollution and environmental contamination” (M = 33.75, SD =
12.27). Cronbach’s alpha reliability coefficient indicated high level of internal consistency (X
= .82). Thus, the seven items were summed to create an index of perceived risks of
nanotechnology, with a low score indicating low level of perceived risks and a high score
indicating high level of perceived risks.
Next, perceived benefits of nanotechnology was created by constructing an additive index of
seven ten-point items (1 = “do not agree at all,” 10 = “agree very much”): (a) “Nanotech may
lead to new and better ways to treat and detect human diseases,” (b) “Nanotech may lead to
new and better ways to clean up the environment,” (c) “Nanotech may give scientists the ability
to improve human physical and mental abilities,” (d) “Nanotech may help us develop increased
national security and defensive capabilities,” (e) “Nanotech may lead to technologies that will
help solve our energy problems,” (f) “Nanotech may revolutionize the computer industry,” and
(g) “Nanotech may lead to a new economic boom” (M = 47.50, SD = 14.48). Cronbach’s alpha
also indicated strong internal consistency among the items (X = .91). Therefore, the seven items
were summed up in which higher scores indicate higher level of perceived benefits of
nanotechnology.
Finally, perceived risks-versus-benefits of nanotechnology was constructed by subtracting the
summed perceived benefits of nanotechnology from the summed perceived risks of
nanotechnology (M = -13.70, SD = 15.32), with higher scores indicating greater perceived risks
relative to perceived benefits. This measure of perceived risks-versus-benefits of
nanotechnology has been used by previous studies in the context of nanotechnology (e.g., Lee et
al., 2005), and the measure has at least two advantages over other similar measures used in
previous studies (e.g., “To what extent do you believe that the risks of scientific research
64
outweigh its benefits?”). First, this measure provides an objective assessment of people’s
perceived risks-versus-benefits by not forcing respondents to make subjective calculations
about the relative importance of risks and benefits for their own attitudes on the topic.
Moreover, this measure is unlikely to be susceptible to order effects and hence is a more reliable
indicator of the construct.
4.1.2.2. Independent Variables
Religious beliefs. Respondents were asked to indicate on a ten-point scale (1 = “no
guidance at all,” 10 = “a great deal of guidance”), how much guidance does religion provide in
their everyday life (M = 6.00, SD = 3.01).
Deference to scientific authority. This variable was measured using two items on a tenpoint scale (1 = “do not agree at all,” 10 = “agree very much”): (a) “Scientists know best what is
good for the public,” and (b) “Scientists should do what they think is best, even if they have to
persuade people that it is right.” The items were averaged to create a composite scale (M = 4.30,
SD = 2.02, r = .39, p < .001).
Science media use. Respondents were asked to indicate how much attention they pay to
the following items when they read newspapers, watch television, and read online content on a
ten-point scale (0 = “no attention at all,” 10 = “very close attention”): (a) “Stories related to
science and technology,” (b) “Stories about scientific studies in new areas of research such as
nanotechnology,” and (c) “Stories about the social or ethical implications of emerging
technologies.” This corresponded to nine separate items. Cronbach’s alpha reliability coefficient
indicated high internal consistency among these nine items (X = .89). Therefore, these items
were averaged to create a composite index, with higher score indicating greater amount of
attention (M = 4.73, SD = 2.12).
65
Elaborative processing was tapped using two items measured on a ten-point scale (1 = “do
not agree at all,” 10 = “agree very much”): (a) “After I encounter news about a scientific
development, I am likely to stop and think about it,” and (b) “If I need to act on science
information, the more viewpoints the media give me the better.” The two items were averaged
to form an index, with higher scores indicating greater amount of information processing (M =
7.15, SD = 2.11, r = .42, p<.001).
Science discussion. Science discussion was measured by asking respondents to indicate on
a ten-point scale (1 = “never,” 10 = “all the time”) how often they talked with family, friends, or
co-workers about: (a) “Stories related to science and technology, (b) “Stories about scientific
studies in new areas of research such as nanotechnology,” and (c) “Stories about the social or
ethical implications of emerging technologies.” Based on a high Cronbach’s alpha reliability
coefficient (X = .90), the three items were averaged to create a composite index, with higher
scores indicating higher level of science discussion (M = 4.40, SD = 2.18).
Factual scientific knowledge was an additive index of five dichotomous items asking
respondents whether (a) “Lasers work by focusing sound waves,” (b) “Antibiotics kill viruses as
well as bacteria,” (c) “Electrons are smaller than atoms,” (d) “Ordinary tomatoes do not contain
genes, while genetically modified tomatoes do,” and (e) “More than half of human genes are
identical to those of a chimpanzee.” For each item, the correct answer was recoded into “1,”
whereas the incorrect answer was recoded into “0.” In addition, responses that fell into the
“don’t know” or “refused to answer” categories were recoded into “0.” The score for the five
items were summed up, with higher scores indicating greater level of factual scientific
knowledge (M = 3.44, SD = 1.25, KR-20 = .47).
Trust in scientists was measured using two items on a ten-point scale (1 = “do not trust
their information at all,” 10 = “trust their information very much”) with regard to how much
66
the respondents trust: (a) “University scientists doing research in nanotechnology” and (b)
“Scientists working for the nanotech industry.” As the two items were highly and significantly
correlated (r = .58, p < .001), they were averaged to form a composite scale (M = 6.16, SD = 2.00).
4.1.2.3. Control Variables
The control variables included in Study 1 were age, gender, and socioeconomic status.
Age was measured as a continuous variable (M = 46.15, SD = 17.07) and gender was measured as
a dichotomous variable, with males dummy coded as “0” and females dummy coded as “1”
(51.4 percent females). Respondents’ socioeconomic status was a concept that I would like to
examine in this study. Therefore, based on both conceptual and methodological considerations,
formal education (Median = 5.00, or “some college or technical school,” SD = 1.57) and
household income (Median = 6.00, or “household income between $50,000 and $75,000,” SD =
1.92) were standardized and averaged to form a composite index of socioeconomic status (r = .43,
p<.001).5 Standardization was necessary to bring the two items that were originally measured in
different metrics onto the same metric.
An unintended advantage of combining education and household income into a single index of
socioeconomic status was that it overcomes the potential problem of the sizable number of missing values
for the household income item (there were 14.8 percent of missing responses for the household income
variable). Result of an independent samples t-test showed that there was no significant difference in the
means of the level of education between those who provided their household income in the survey and
those who did not (t-statistic = -.670, df = 201, p = .494). This suggests that the chances that systematic
biases were built into the constructed variable were kept to a minimum. A more fine-grained analysis
using four sets of independent samples t-tests was also conducted in which levels of education was
broken down into smaller categories to compare differences in income level. Result show that there was
no significant difference in the means of income between individuals who “never attended school or only
attended kindergarten” and those who went through elementary school (t-statistic = 1.278, df = 12, p
=.225), between individuals who went through some high school and those who were high school
graduate (t-statistic = -1.729, df = 273, p =.085), and between individuals who did graduate work and
those had a graduate degree (t-statistic = -1.162, df = 156, p =.247). Even though the t-test between
individuals with some college or technical school and those with a 4-year college degree was significant
(t-statistic = -6.332, df = 415, p < .001), there is sufficient evidence to show that income is missing
completely at random for the public sample.
5
67
4.1.3. Analytic Strategies
Two different sets of multivariate analyses were run in this study. Ordinary regression
analyses using the Statistical Package for the Social Sciences (SPSS) for Windows Standard
Version 15 were run to examine the direct and moderating mechanisms underlying the
cognitive and attitudinal outcomes. Next, structural equation modeling, using the LISREL
version 8.70 software program, was used to investigate the direct and mediating mechanisms
underlying the cognitive and attitudinal outcomes. Before running the multivariate analyses,
the nature and treatment of missing values should be examined in greater detail to ensure that
the results are generalizable to the larger population.
4.1.3.1. Missing Values Treatment
Two considerations determine the extent to which analyses with incomplete data can be
biased: (a) the pattern of missing values and (b) the amount of missing data. First off, missing
values should be treated differently depending on different patterns of missingness determined
by different problems. Depending on how the missing cases are related to the values that would
have been observed or other variables in the data, the missing data have three different
patterns: missing completely at random (MCAR), missing at random (MAR), and not missing at
random (NMAR). Both MCAR and MAR are two kinds of ignorable missing data patterns,
whereas NMAR is a type of non-ignorable non-response. MCAR indicates that the missing
values are completely due to chance and entirely unrelated to the values (Allison, 2001; Schafer
& Graham, 2002). Put differently, if the missing observations on some variable X differ from the
observed scores on that variable only by chance, and the presence versus absence of data on X is
unrelated to any other variable, the data loss pattern is said to be MCAR. When all the missing
values can be safely assumed to be MCAR, any kind of data treatment such as listwise deletion,
pairwise deletion, mean substitution, and maximum likelihood, yields consistent solutions
68
(Arbuckle & Wothke, 1999). On the other hand, MAR indicates that the missingness and data
values are statistically unrelated but conditional on a set of predictors. In addition, NMAR
indicates that the missingness conveys probabilistic information about the values that would
have been observed. When either missing values classified as MAR or NMAR are involved,
analysis using methods such as listwise deletion, pairwise deletion, and mean substitution often
produces biased results.
With respect to the 2007 public opinion data used in this dissertation, no socially
undesirable or sensitive questions were asked in the questionnaire. Furthermore, the length of
the questionnaire was kept short to avoid any respondent fatigue, and significant amount of
time and effort were invested in call-backs and refusal conversions to minimize systematic nonresponse. Therefore, it is safe to assume in this dissertation that the missing values were MCAR,
even though this is the most heroic assumption to make.
Second, the amount of missing values in the data is also another consideration for how
the missing values should be treated. To date, no clear standard guidelines exist to determine
the acceptable amount of incomplete data. However, in order to see less biased results, Kline
(2005) argued that the amount of missing values should probably constitute less than ten
percent of the data (also see Allison, 2001; Little & Rubin, 2002).
Under the MCAR assumption, listwise deletion, pairwise deletion, and mean
substitution, should produce unbiased results. Mean substitution basically replaces missing
independent variable values with the mean score obtained from the responding cases. One
main advantage of mean substitution is that the simple regression coefficient of Y on X is the
same whether missing X scores are replaced with the mean or whether cases with missing
values are dropped from the analysis. It is appropriate to use mean substitution if the amount of
missing values is small.
69
Under the other two methods, different calculations (e.g., different correlation
coefficients) will utilize different cases and will have different sample sizes (different n’s).
Listwise deletion omits cases that do not have data on all variables in the variables list of the
current analysis (of a multiple-step analysis). An advantage of this method is that all analyses
are conducted with the same number of cases. On the other hand, pairwise deletion omits cases
that do not have data on a variable used in the current calculation only. This effect is
undesirable and may prevent a solution altogether in some procedures like structural equation
modeling. In general, listwise deletion is preferred over pairwise deletion when sample size is
large in relation to the number of cases that have missing data. Kaplan (2009) recommended
that pairwise deletion should be avoided altogether, and that mean substitution is preferred
over listwise deletion and pairwise deletion.
As the number of missing values in the data was very small (i.e., less than 2 percent for
each of the independent variables concerned) and MCAR was assumed in this dissertation, all
missing data in the ordinary regression analyses were treated with mean substitution. Although
not the most preferred option, listwise deletion was used in the structural equation model
analysis because it was the default missing values treatment provided in the LISREL software.
4.1.3.2. Ordinary Regression Analysis
To test the hypothesized direct and moderating relationships described in the previous
chapter, ordinary regression analysis was used in predicting each of the following two
attitudinal outcome variables: (1) perceived risks-versus-benefits of nanotechnology and (2)
support for federal funding of nanotechnology. Ordinary regression examines the relationships
between potential independent variables and an outcome variable by fitting a linear equation to
the observed data. This method was chosen for Study 1 because it provides the most
70
straightforward way to test for potential conditional (i.e., moderating) relationships, above and
beyond the potential unconditional (i.e., main effects) relationships that were hypothesized.
To examine their relative explanatory power, all the independent variables were entered
in blocks into the regression model according to their assumed casual order. For the first
regression model predicting public perceived risks-versus-benefits of nanotechnology, the
independent variables were entered as follow: Demographic variables (i.e., age, gender, and
socioeconomic status) were entered first, followed by the main effects of trait-like value
predispositions (i.e., religious beliefs and deference to scientific authority), science
communication (i.e., science media use), reflective integration (i.e., elaborative processing and
science discussion), factual scientific knowledge, and state-like disposition (i.e., trust in
scientists). Finally, the interaction terms were entered in the last block. Each of the interaction
terms was constructed by multiplying the standardized values of the main effect variables to
prevent potential multicollinearity problems between the interaction term and its components
(Cohen, Cohen, West, & Aiken, 2003). In the first regression model, two multiplicative terms
were included in the final regression block: (1) the interaction between science media use and
elaborative processing and (2) the interaction between science media use and science discussion.
For the second regression model predicting public support for federal funding of
nanotechnology, the independent variables were entered as follows: The first block consisted of
the demographic variables (i.e., age, gender, and socioeconomic status), the second block
consisted of the main effects of trait value predispositions (i.e., religious beliefs and deference to
scientific authority), the third block consisted of science communication (i.e., science media use),
the fourth block consisted of reflective integration (i.e., elaborative processing and science
discussion), the fifth block consisted of factual scientific knowledge, the sixth block consisted of
trust in scientists, and the seventh block consisted of perceived risks-versus-benefits of
71
nanotechnology. Finally, two interaction terms were entered in the last block: (1) the interaction
between science media use and elaborative processing and (2) the interaction between science
media use and science discussion.
A third regression model predicting the cognitive outcome variable (i.e., factual
scientific knowledge) was also run. As Study 1 was primarily interested in examining the direct
and additive effects of science media use and reflective integration on factual scientific
knowledge, the demographic variables (i.e., age, gender, and socioeconomic status) and value
predispositions (i.e., religious beliefs and deference to scientific authority) were added as
control variables in the regression analysis. The third block of independent variables of interest
consisted of science media use, the fourth block consisted of reflective integration (i.e.,
elaborative processing and science discussion), and the final block consisted of the interaction
terms (i.e., the interaction between science media use and elaborative processing, and the
interaction between science media use and science discussion).
4.1.3.3. Structural Equation Modeling
Structural equation modeling was used to test the hypotheses related to both the direct
and indirect (i.e., mediated) effects of the posited independent variables of interests on the
outcome variable. In particular, structural equation modeling offers information on the overall
strength of the indirect effects and allows estimation of all the individual links among
exogenous and antecedent endogenous variables, and the links between antecedent and
outcome variables. For the purpose of this study, the exogenous variables were age, gender,
socioeconomic status, religious beliefs, and deference to scientific authority; the antecedent
endogenous variables were science media use, elaborative processing, science discussion,
factual scientific knowledge, trust in scientists, and perceived risks-versus-benefits of
nanotechnology; and finally, the outcome variable in the model was public support for federal
72
funding of nanotechnology. An examination of the process of pathways linking exogenous
variables, antecedent endogenous variables, and support for federal funding of nanotechnology
as the final outcome variable require the estimation of both the direct and indirect relationships
among the variables, that is, methods that go beyond, or at least supplement, traditional
regression techniques.
Given a set of variables, structural equation models postulate “a pattern of linear
relationships among these variables” (MacCallum, 1995, p. 18) and test these relationships
against the data collected. LISREL version 8.70 software program, the package of reference in
most articles about structural equation modeling (Kline, 2005), was used in this study to analyze
the hypothesized relationships.6 LISREL uses the maximum likelihood (ML) method for
parameter estimations and analyzes data based on a covariance matrix. Parameter estimation is
done comparing the actual covariance matrices representing the relationships between variables
and the estimated covariance matrices of the best fitting model (Kaplan, 2009). In addition, the
ML estimators are those that maximize the likelihood of a sample that is actually observed
(Winer, Brown, & Michels, 1991).
Structural equation modeling allows researchers to examine the causal ordering of
variables more closely and to describe both the direct and indirect paths among variables (i.e.,
the processes leading to support for federal funding of nanotechnology). These are the strengths
of structural equation modeling that overcome some of the shortcomings of ordinary regression
analyses. In addition, structural equation modeling has numerous advantages over other
multivariate techniques. First, structural equation modeling estimates all coefficients in the
model simultaneously, in which any coefficient represents the relationship between two
Even though it was not the most preferred option, listwise deletion was used in the structural equation
model analysis as it was the default missing values treatment provided in the LISREL software.
6
73
variables, controlling for all other relationships and all other variables in the model. Second,
structural equation modeling is the most straightforward method to test both direct paths and
indirect relationships between variables. That is, a direct link between two variables indicates a
relationship between these variables that is not mediated or moderated by any other variables
in the model. Conversely, an indirect link is an association that is mediated through other
variables.
Jöreskog (1993) distinguished three approaches to structural equation modeling
applications: (1) strictly confirmatory, (2) alternative models, and (3) model-generating
approach. As pointed out by numerous scholars (e.g., Kline, 2005), the first two methods of
model testing are too narrow and restrictive. Therefore, this study followed Jöreskog’s
suggestion for the model-generating approach, which is probably the most common application.
This approach is usually carried out in two steps. First, an initial model is specified based on “at
least some tentative ideas of what a suitable model should be” (Joreskog, 1993, p. 313). Next,
when the initial model does not fit the data, it can be modified by the researcher and the altered
model is then tested again with the same data. As aptly summarized by Kline (2005), “the goal
of this process is more to ‘discover’ a model with two properties: It makes theoretical sense, and
its statistical correspondence to the data is reasonable.” (p. 11)
In line with the model-generating approach, the likelihood ratio test that follows a chisquare distribution is usually used to test the overall fit of the initial model. The model fit is
tested with a chi-square (χ2) goodness-of-fit test. The χ2 goodness-of-fit statistic assesses the
magnitude of discrepancy between the sample and fitted covariance matrices (Kaplan, 2009).
However, the χ2 statistic is very sensitive to sample size. If the sample size is too large, the χ2
statistic may be significant even though differences between observed and model-implied
covariance are small. To supplement the χ2 statistic, two groups of fit indices that are relevant
74
for this study were used. The first group consisted of fit indices that are based on comparative
fit to a baseline model which specifies complete independence among the observed variables:
Normed Fit Index (NFI), Non-Normed Fit Index (NNFI; identical to Tucker-Lewis Index [TLI]),
and Comparative Fit Index (CFI). Typically, for estimation using the ML method, a value close
to 1.00 on the NFI, NNFI, and CFI is considered representative of a well-fitting model (Kaplan,
2009). One major criticism of these indices is that they are designed to compare one’s
hypothesized model against a scientifically questionable baseline hypothesis. The baseline
hypothesis states that the observed variables are completely uncorrelated with each other. Sobel
and Bohrnstedt (1985) argued that we could never seriously entertain such a hypothesis.
Furthermore, the NFI and NNFI utilize the likelihood ratio chi-square and assume that the
model fits perfectly in the population. This could be considered too restrictive and we may need
to evaluate the approximate fit of the model.
Therefore, the second group consisted of fit indices that are based on errors of
approximation: Standardized Root Mean Squared Residual (SRMR) and Root Mean Square
Error of Approximation (RMSEA). These measures assess if the model fits approximately well
in the population (i.e., measures of approximate fit are concerned with the discrepancy due to
approximation). Specifically, SRMR is the average difference between the predicted and
observed variances and covariances in the model, based on standardized residuals. The smaller
the SRMR value, the better the model fit. A SRMR value of 0 indicates perfect fit. A value less
than .05 is widely considered good fit and below .08 adequate fit (Hu & Bentler, 1999; Kelloway,
1998; Kline, 2005). A RMSEA value that is less than or equal to .05 is indicative of a good model
fit. A RMSEA value that lies between .05 and .08 is indicative of a fair fit, and a value that lies
between .08 and .10 is indicative of a mediocre fit (Kaplan, 2009). These cut-offs are somewhat
arbitrary, and therefore, should serve as a rule-of-thumb instead of as fixed criteria (Bollen,
75
1989; Kaplan, 2009). Moreover, when sample sizes are greater than 250, it is recommended to
employ SRMR and RMSEA as fit indices (Browne & Cudeck, 1993). These fit indices were
designed to avoid some of the problems of sample size and distributional misspecification
associated with the conventional overall test of fit (i.e., the χ2 statistic) in the evaluation of the
model (Bentler & Bonnet, 1980). Based on the discussions above, several goodness-of-fit indices
were reported in this study to gauge model fit, including NFI, NNFI, CFI, SRMR, and RMSEA,
along with the χ2 statistics. The skewness and kurtosis statistics indicate that none of the items
in the model violate the normality assumption to a great extent.
If the model does not fit the data based on the results of the goodness-of-fit indices, the
Lagrangian Multiplier (LM) tests or modification indices (MI) can be used to determine whether
the paths can be freed or fixed to improve model fit (Bollen, 1989). All parameters added based
on the LM test should be meaningful and substantially interpretable (Joreskog, 1993).
In addition, when determining if a particular parameter is to be freed for estimation,
Kaplan (2009) suggested researchers to take into consideration both the MI and the expected
parameter change (EPC) statistic. EPC is a point estimate of the alternative hypothesis for the
parameter in question. A general rule of thumb is that when both MI and EPC are large, free the
parameter if it makes substantive sense; when MI is large but EPC is small, check the power
before freeing the parameter; when MI is small but EPC is large, it is possible to free the
parameter but check power; and when MI is small and EPC is small, do not free the parameter
for estimation (Kaplan, 2009). Therefore, for the purpose of this study, both the MI and EPC
were taken into account when deciding which parameters were to be freed for estimation.
76
4.2. Results
In this section, tests of the proposed direct, moderating, and mediating mechanisms are
reported. Before running the multivariate analyses, a bivariate correlation analysis using SPSS
was run to take a cursory look at the relationships among all the variables at the zero-order
level. Table 4.2 shows the bivariate correlations among the variables. As only pairwise or
listwise deletion for bivariate correlations were available in SPSS, the method of listwise
deletion was chosen for reasons that were discussed above. As shown in the table, most of the
independent variables were indeed highly correlated with the two attitudinal outcome variables
and some of the control variables. For instance, socioeconomic status was positively correlated
with many of the independent variables such as science media use (r = .31, p < .001), elaborative
processing (r = .20, p < .001), science discussion (r = .17, p < .001), and factual scientific
knowledge (r = .38, p < .001). Even though most of the independent variables were significantly
correlated with perceived risks-versus-benefits of nanotechnology and support for federal
funding of nanotechnology, it is likely that the significant association between the independent
variables and the outcome variables may become weaker after controlling for the demographic
and value predisposition variables. The multivariate analyses will be reported next.
4.2.1. Direct and Moderating Relationships
Three separate regression analyses were run for (a) perceived risks-versus-benefits of
nanotechnology (attitudinal outcome variable 1), (b) support for federal funding of
nanotechnology (attitudinal outcome variable 2), and (c) factual scientific knowledge (cognitive
outcome variable), respectively. Table 4.3 shows the ordinary regression model predicting
public perceived risks-versus-benefits of nanotechnology. Among the demographic variables,
females displayed significantly higher perceived risks-versus-benefits than did males (β = .12, p
< .001), and socioeconomic status showed a negative relationship with perceived risks-versus-
77
benefits (β = -.11, p < .001), as indicated in the final regression model. However, age was not
significantly related to perceived risks-versus-benefits. The demographic variables accounted
for 9.90 percent of the total variance in perceived risks-versus-benefits of nanotechnology.
With regard to the main effects of trait-like value predispositions, levels of religious
beliefs showed robust positive relationship with perceived risks-versus-benefits (β = .14, p
< .001), lending support to Hypothesis 1a. Deference to scientific authority was initially
significantly associated with perceived risks-versus-benefits, but the significant relationship
was fully mediated away by the science media use variable entered in the subsequent block.
Therefore, Hypothesis 2a was not supported. The trait-like predispositions block accounted for
an additional 4.30 percent amount of variance in perceived risks-versus-benefits of
nanotechnology.
As expected, science media use was negatively related to perceived risks-versus-benefits
of nanotechnology in the final regression model (β = -.11, p < .001), providing support for
Hypothesis 3a. The science media use block accounted for 4.00 percent of the total variance in the
outcome variable.
In terms of reflective integration, elaborative processing displayed a significant negative
relationship with perceived risks-versus-benefits (β = -.10, p < .01). On the contrary, science
discussion did not predict perceived risks-versus-benefits of nanotechnology. Therefore,
Hypothesis 4b was supported, but not Hypothesis 5b. The reflective integration block explained an
additional 2.00 percent of the variance in the attitudinal outcome variable.
In addition, factual scientific knowledge showed a significant negative association with
perceived risks-versus-benefits of nanotechnology (β = -.15, p < .001), lending support to
78
Hypothesis 12a. Factual scientific knowledge accounted for an additional 1.90 percent of the
variance in perceived risks-versus-benefits.
With respect to the state-like disposition, trust in scientists displayed a robust negative
relationship with perceived risks-versus-benefits of nanotechnology (β = -.27, p < .001). Hence,
Hypothesis 13a was supported. The state-like disposition block accounted for 5.50 percent of the
variance in the outcome variable.
As expected, the interaction between science media use and elaborative processing on
perceived risks-versus-benefits was significant (β = -.08, p < .01), after accounting for all controls.
Therefore, Hypothesis 7a was supported. Conversely, the interaction between science media use
and science discussion on perceived risks-versus-benefits was not significant, failing to support
Hypothesis 8a. As shown in Figure 4.1, among those with low level of elaboration, science media
use displayed a significant positive relationship with perceived risks-versus-benefits; on the
other hand, among those with high level of elaboration, science newspaper use displayed a
negative significant relationship with the outcome variable. The overall regression model
accounted for 28.2 percent of the variance in perceived risks-versus-benefits of nanotechnology.
Table 4.4 shows the ordinary regression model predicting public support for federal
funding of nanotechnology (attitudinal outcome variable 2). With respect to the demographic
variables, only age was significantly related to support for funding (β = -.05, p < .05) in the final
regression model. The influence of gender and socioeconomic status were initially significant,
but their effects were fully mediated by other factors that were entered subsequently into the
model. The demographic block explained 6.80 percent of the total variance in support for
funding.
The value predispositions block explained a large amount of variance in support for
funding (9.30 percent). As expected, level of religious beliefs was negatively (β = -.08, p < .01)
79
and deference to scientific authority was positively (β = .13, p < .001) related to support for
funding. Therefore, Hypotheses 1b and 2b were supported.
Furthermore, science media use positively predicted support for federal funding (β = .08,
p < .05), and explained an additional 5.70 percent of the total variance in the outcome variable.
Hence, Hypothesis 3b was supported. In line with my expectations, elaborative processing (β
= .09, p < .01) and science discussion (β = .07, p < .05) both showed significant positive effects on
support for federal funding of nanotechnology, lending support for Hypotheses 4c and 5c. The
reflective integration block accounted for 2.80 percent of the variance in support for federal
funding of nanotechnology.
Contrary to my expectation, factual scientific knowledge did not have a significant
relationship with support for funding, failing to support Hypothesis 12b. The factor explained
only .20 percent of the variance in the attitudinal outcome variable.
When it comes to state-like disposition, trust in scientists showed a robust positive
relationship with support for funding of nanotechnology (β = .19, p < .001), lending support to
Hypothesis 13b. The factor explained an additional 5.00 percent of the variance in the outcome
variable.
With regard to the risks-versus-benefits block, perceived risks-versus-benefits was
negatively (β = -.26, p < .001) associated with support for federal funding of nanotechnology,
lending support to Hypothesis 16. Perceived risks and benefits explained an additional 4.80
percent of the total variance in the outcome variable.
Finally, there was significant interaction effect between science media use and
elaborative processing on support for federal funding of nanotechnology (β = .07, p < .05).
Hence, Hypothesis 7b was supported. However, contrary to my expectation, the interaction
between science media use and science discussion on support for funding of nanotechnology
80
was not significant, failing to support Hypothesis 8b. As shown in Figure 4.2, the relationship
between science media use and support for funding was significantly stronger for people who
engage in high level of elaboration than for those who engage in low level of elaboration. The
overall regression model explained 35.0 percent of the total variance in support for federal
funding of nanotechnology.
Table 4.5 indicates the ordinary regression model predicting public level of factual
scientific knowledge (cognitive outcome). As expected, elaborative processing positively
predicted public level of factual scientific knowledge (β = .12, p < .001), lending support to
Hypothesis 4a. However, scientific discussion did not have a significant main effect on public
level of factual scientific knowledge, failing to support Hypothesis 5a. Contrary to my
expectations, the interaction between elaborative processing and science media use on factual
scientific knowledge, and the interaction between scientific discussion and science media use on
factual scientific knowledge were not significant. Therefore, Hypotheses 6a and 6b were not
supported.
Even though these relationships were not hypothesized, the regression results shown in
Table 4.5 indicate that age and religious beliefs negatively predict respondents’ level of factual
scientific knowledge. Furthermore, socioeconomic status and science media use positively
predict public level of factual scientific knowledge. Gender and deference to scientific authority
had no effect on factual scientific knowledge. The overall regression model accounted for 22.0
percent of the variance in factual scientific knowledge.
4.2.2. Direct and Mediating Relationships
To examine the proposed indirect or mediated links, structural equation modeling
techniques were employed. The structural equation model provides information on the overall
strength of indirect effects and estimation on all individual links among exogenous and
81
antecedent endogenous variables, and between antecedent and outcome variables.7 Kline (2005)
noted that even though slight differences may exist, structural equation modeling and ordinary
regression procedure generally yield similar results within large samples. Therefore, I expect
that estimations of the direct effect in the structural equation model would be similar to the
ones found in the ordinary regression models.
The final structural equation model fits the data exceptionally well (χ2 goodness-of-fit =
25.85, df = 19, p = .13, N = 1,015). The CFI was 1.00, the NFI was .99, the NNFI was .99, and the
GFI was 1.00, which were indicative of a good fit. In addition, the RMSEA value was .02 and the
SRMR value was .02, which were also indicative of a good fit. The model accounted for 21
percent of the variance in factual scientific knowledge, 26 percent of the variance in trust in
scientists, 29 percent of the variance in perceived risks-versus-benefits of nanotechnology, and
35 percent of the variance in support for federal funding of nanotechnology.
Effects of exogenous variables. Three demographic variables and two trait-like value
predisposition variables were included as exogenous variables. Table 4.6 reports the influence
of the exogenous variables on other variables in the structural equation model. As shown in the
table, older individuals tended to be factually less knowledgeable about science (γ = -.10) and to
hold less trust in scientists (γ = -.11) than did younger individuals, and by way of these
variables, age was positively related to perceived risks-versus-benefits of nanotechnology (γ
= .04). As a combination of both direct and indirect effects, younger individuals tended to report
stronger support for federal funding of nanotechnology than did older individuals (γ = -.10).
It would be worthwhile to conduct separate tests to examine if the individual mediated pathways are
significant. However, the LISREL software does not provide the option to conduct such separate
significant tests. Future studies may use alternative software such as MPLUS to conduct significant tests
for the separate indirect pathways examined in this dissertation.
7
82
In comparison to women, men tended to pay greater amount of attention to science in
the media than did women (γ = -.07). In an indirect relationship, men tended to elaborate more
about news content than did women (γ = -.03). Interesting, through direct and indirect routes,
overall women tended to engage in scientific discussion more than did men (γ = .03). As
mediated by science media use and elaborative processing, gender also had an indirect effect on
factual scientific knowledge, in which men tended to score higher on the measure of factual
scientific knowledge than did women (γ = -.01). Women tended to hold greater levels of trust in
scientists than did men (γ = .07), with part of this total relationship mediated by science media
use and elaborative processing. Finally, through both direct and indirect effects, men tended to
perceive smaller risks-versus-benefits of nanotechnology (γ = .15) and to be more supportive of
federal funding of nanotechnology than did women (γ = -.11).
Consistent with prior research, individuals with higher levels of socioeconomic status
(SES) also tended to pay closer attention to stories about science in the mass media than did
those with lower levels of SES (γ = .29). Via direct and indirect influences, SES was overall
positively related to elaborative processing (γ = .20). Interestingly, SES was only indirectly
positively related to scientific discussion, with its effect mediated by science media use (γ = .18).
Not surprising, through both direct and indirect means, individuals with higher levels of SES
tended to report higher score in factual scientific knowledge (γ = .31) and to report lower
perceived risks-versus-benefits of nanotechnology than did those with lower levels of SES (γ = .21). As mediated by other variables in the model, SES had a positive indirect effect on trust in
scientists (γ = .09) and a final positive indirect effect on support for federal funding of
nanotechnology (γ = .13).
In terms of religious beliefs, there was a direct positive relationship with elaborative
processing (γ = .07), which was consistent with findings from previous studies that showed that
83
individuals’ motivations will drive them to reflect on news gathered from the mass media.
Religious belief had a positive indirect effect on trust in scientists (γ = .01), which was not
surprising because the significant indirect link was mediated by elaborative processing. As
expected, religious beliefs appear to be an impediment to scientific knowledge and progress:
Through a combination of direct and indirect effects, highly religious individuals tended to
report overall lower levels of factual scientific knowledge (γ = -.14) and greater amount of
perceived risks-versus-benefits (γ = .12) than those who were less religious. This result provided
additional support for Hypothesis 1a. Religious belief had a direct negative effect on support for
federal funding of nanotechnology (γ = -.08), lending further support to Hypothesis 1b. Finally,
religious beliefs had an overall negative relationship with support for federal funding of
nanotechnology (γ = .13).
Deference to scientific authority was positively associated with science media use (γ
= .14), elaborative processing (γ = .22), and science discussion (γ = .15). Via the mediation of
science media use and elaborative processing, individuals who defer to scientific authority also
tended to score higher on factual scientific knowledge than did those who possess low levels of
deference (γ = .04). Not surprising, deference to scientific authority had an overall positive
effect on trust in scientists (γ = .39). Even though deference to scientific authority had no
significant direct effect on perceived risks-versus-benefits (failing to support Hypothesis 2a), it
had an indirect negative effect on perceived risks-versus-benefits of nanotechnology (γ = -.14)
that was mediated through the communication variables. Finally, deference to scientific
authority showed an overall positive effect on support for federal funding of nanotechnology (γ
= .26), part of this effect was direct (γ = .10) (providing additional support for Hypothesis 2b),
and part of it was indirect, mediated by the science media use and reflective integration
pathway and by the trust in scientist pathway.
84
Effects of endogenous variables. The relationships among the endogenous variables are
graphed in Figure 4.3 and detailed in Table 4.7. As shown from the results, science media use
played a prominent role in influencing public support for federal funding of nanotechnology.
The influence of science media use was both direct and indirect, and could be classified into two
major categories: via the informational pathways, in which the news media provide an informal
learning channel of scientific issues for the public, and via the heuristic pathways, in which the
positive media frames served as cues or shortcut for the miserly public when making judgments.
There were several informational pathways through which mass media use exerted its
influence on the outcome variable. First, consistent with previous studies on the cognitive
mediation model, science media use had a direct positive effect on factual scientific knowledge
(β = .11), and part of its influence was also indirect, as mediated by elaborative processing.
Science media use was positively related to elaborative processing (β = .34) and elaborative
processing was in turn, positively related to factual scientific knowledge (β = .12). As a result,
science media use had an overall positive effect on factual scientific knowledge. Factual
scientific knowledge was, in turn, negatively related to public perceived risks-versus-benefits of
nanotechnology (β = -.16), lending additional support to Hypothesis 12a (however, scientific
knowledge had no significant direct effect on support for federal funding, failing to support
Hypothesis 12b).
Above and beyond textbook-style scientific knowledge, science media use could also
stimulate the public to have more sophisticated knowledge about how the scientific world
operates, which in turn, could allow the public to form informed judgments about emerging
technologies. As shown in the results, elaborative processing mediated the effect of science
media use on perceived risks-versus-benefits, with elaborative processing negatively related to
perceived risks-versus-benefits (β = -.11). (The direct effect of elaborative processing on
85
perceived risks-versus-benefits provided further support for Hypothesis 4b.) Similarly,
elaborative processing also mediated the influence of science media use on public support for
federal funding of nanotechnology, with elaborative processing positively related to support for
funding (β = .10). (The direct effect of elaborative processing on support for federal funding of
nanotechnology provided additional support for Hypothesis 4c.) The reflective integration
variables were interrelated to each other: Elaborative processing was positively related to
science discussion (Φ = .13). Paralleling the previous result, science media use tended to
stimulate scientific discussion (β = .62), and science discussion in turn, was positively associated
with public support for federal funding of nanotechnology (β = .08). (The direct effect of science
discussion on support for federal funding provided further support for Hypothesis 5c. In
addition, there was no significant direct link between science discussion and perceived risksversus-benefits, failing to support Hypothesis 5b.) It is also important to note that perceived
risks-versus-benefits had a direct negative impact on support for federal funding of
nanotechnology (β = -.25), lending additional support to Hypothesis 16. Together, the significant
mediation paths lend support for Hypotheses 9a, 10a, 10b, and 11b. However, Hypotheses 9b and
11a were not supported by the results.
In addition, there were also several heuristic pathways through which the mass media
influenced public judgments about the emerging technology. First, science media use had a
direct negative influence on public perceived risks-versus-benefits (β = -.08) and a direct
positive influence on public support for federal funding of nanotechnology (β = .07), and these
results are consistent with findings from previous studies that indicated that individuals use
positive news frames in the media as cognitive shortcuts to form opinion about emerging
science and technology (e.g., Lee et al., 2005; Scheufele & Lewenstein, 2005). These findings
provided additional support for Hypotheses 3a and 3b. The second heuristic path emanated from
86
the mediating role of trust in scientists. Specifically, science media use (β = .21) and elaborate
processing of science news stories in the media (β = .17) promoted a sense of trust in scientists,
and this sense of trust, in turn, propelled the public to perceive more benefits relative to risks
for nanotechnology (β = -.27). Therefore, Hypotheses 14a and 15a were supported. In a similar
vein, the public tended to use this trust as a heuristic cue to form judgment about support for
federal funding of nanotechnology (β = .21), lending support to Hypotheses 15a and 15b. (The
significant direct effects of trust in scientists on perceived risks-versus-benefits and support for
federal funding provided additional support for Hypotheses 13a and 13b.) With a combination of
both the informational pathways and the heuristic pathways, science media use had an overall
positive effect on public support for federal funding of nanotechnology (β = .26).
A summary of the evidence supporting each of the hypotheses postulated in Study 1
was provided in Table 4.8. As shown in the table, the direct effects of the independent variables
on the outcome variables of interest were consistent in both the ordinary regression models and
the structural equation model. Put differently, the results of the direct effects found in the
ordinary regression models were replicated in the structural equation model. These findings
will be discussed in greater details in the next chapter.
87
CHAPTER 5
DISCUSSION (STUDY 1)
Driven by the debate between the scientific literacy model (Miller, 1998) and the
cognitive miser model (Fiske & Taylor, 1991) for the public understanding of science, Study 1
attempted to draw a nexus between these two approaches by examining the direct, indirect, and
additive impact of science news in the media and reflective integration on public perceived
risks-versus-benefits of nanotechnology and public support for federal funding of
nanotechnology. Drawing from extant research on the differential gains model (Scheufele, 2001)
and the cognitive mediation model (Eveland, 2001), the analysis of the moderating and
mediating mechanisms using these theoretical frameworks is a building block for designing
more effective science communication and public outreach efforts.
To reiterate, Study 1 examines the influence of value predispositions, mass media,
reflective integration, and factual scientific knowledge on public perceived risks-versus-benefits
and on public support for federal funding of nanotechnology. Several interesting and important
results were found in this study. Overall, the ordinary regression analyses provide strong
support for the hypothesis that reflective integration in the form of elaborative processing had a
significant negative influence on public perceived risks-versus-benefits. In line with the
differential gains model, the influence of science media use on both perceived risks-versusbenefits and support for federal funding of nanotechnology were moderated by elaborative
processing, after controlling for socio-demographic and value predisposition variables.
Furthermore, the structural equation model reveals an informational pathway and a heuristic
pathway through which the mass media directly and indirectly exert its influence on public
attitudes toward nanotechnology. Taken together, these findings underscore the importance of
88
cognitive processing when it comes to understanding how the mass media differentially
influence individuals’ attitudes toward emerging technologies.
5.1. Explanations for Findings on Direct Effects
In line with the cognitive miser explanation, the public primarily uses value
predispositions – religious beliefs, deference to scientific authority, and trust in scientists – to
make judgments about risks-versus-benefits and acceptance of nanotechnology. This is
evidenced by the large beta coefficients and amount of variance explained in the attitudinal
outcome variables. These findings are consistent with my expectations, based on the public’s
unfamiliarity with this new technology. Being cognitive miser means that people would rely on
their pre-existing values and beliefs as heuristics to make quick and efficient decisions.
Consistent with results from previous studies (e.g., Brossard et al., in press; Ho et al.,
2008; Nisbet, 2005), the current study showed that religious belief is positively related to public
perceived risks-versus-benefits of nanotechnology and negatively related to public support for
federal funding of the emerging technology. The historical conflict and normative
inconsistencies between science and religion (Brooke, 1998; Miller et al., 1997) may be an
explanation for the relationships found between religious guidance and acceptance of
nanotechnology. In addition, the fact that religious people may perceive nanotechnology,
biotechnology, and stem cell research together as means to enhance human qualities, hence
tampering with nature by playing God (Sjoberg, 2004; Sjoberg & Winroth, 1986) is also another
plausible explanation for the relationships.
On the other hand, individuals’ deference for scientific authority and trust in scientists
are two driving forces propelling public acceptance of nanotechnology, consistent with findings
from previous research (Brossard & Nisbet, 2007; Ho et al., 2008; Lee et al., 2005). Again, these
89
findings are not surprising because, as tools in decision-making, deference for scientific
authority and trust in scientists are efficient when knowledge and personal experience are
limited, especially when it comes to nanotechnology. Although deference to scientific authority
did not display a significant direct influence on perceived risks-versus-benefits of
nanotechnology both in the regression model and the structural equation model, its effect was
indirect as mediated by mass media (this will be discussed in greater details in the subsequent
sections on mediating mechanisms). In addition, the independent effects of deference to
scientific authority and trust in scientist on public attitudes toward nanotechnology suggest that
researchers should adopt a fine-grained approach to examine these concepts separately in
future studies as they are essentially different entities. The fact that deference to scientific
authority, as a long-term trait-like value predisposition, is shown to promote trust in nanoscientists (i.e., a state-like disposition) in the structural equation model further supports the
assertion that they are distinct concepts.
In addition, this study shows that the public utilize positive frames derived from the
mass media as heuristic cues to make decision about perceived risks-versus-benefits and
acceptance of the emerging technology, which is congruent to results of previous studies
(Brossard & Nisbet, 2007; Lee & Scheufele, 2006; Lee et al., 2005; Scheufele & Lewenstein, 2005)
and consistent with framing effects of the media (Kahneman & Tversky, 1979; Scheufele, 1999).
This could plausibly be explained by the fact that media outlets are the major gateway to
nanotechnology for most Americans (Castellini et al., 2007) and that the tone of media coverage
of nanotechnology has been overwhelmingly optimistic in the past few years (Bainbridge, 2002;
Cobb & Macoubrie, 2004; Gaskell et al., 2004).
On the other hand, factual scientific knowledge played a significant, but minor role in
influencing perceived risks-versus-benefits, consistent with findings from previous studies (Ho
90
et al., 2008; Nisbet, 2005; Scheufele & Lewenstein, 2005). Factual scientific knowledge, however,
did not have a main effect on public support for nanotechnology. Given the minimal effect of
factual scientific knowledge and the relatively larger effects of heuristic cues on public attitudes,
does it point to the demise of the scientific literacy model? Before we explore this tentative
conclusion, however, we need to look at the informational roles that the mass media and
reflective integration bring into the picture.
Besides this, reflective integration, in the form of elaborative processing, plays an
important role in determining both the cognitive outcome (i.e., factual scientific knowledge) and
the attitudinal outcomes (i.e., perceived risks-versus-benefits and support for federal funding)
of the new technology. This could be explained by the fact that people who actively process and
synthesize information from the mass media build a larger knowledge structure about science
generally, and nanotechnology specifically, in their memory. This new scientific information
could be easily accessed for people to formulate judgments about the risks and benefits of
nanotechnology and nanotechnology acceptance. Nanotechnology has been covered in
overwhelmingly positive light in the mass media and it is therefore, not surprising that these
positive information become part of the audience memory when audiences reflect and integrate
the materials they attended to in the news.
Even though scientific discussion did not have a significant main effect on public
perceived risks-versus-benefits, it had a significant positive main effect on public support for
federal funding of nanotechnology. One plausible explanation for the null effect of
interpersonal discussion on perceived risks-versus-benefits is that previous studies have shown
that interpersonal communication tended to heighten risk judgments at the personal level
instead of the societal level (Coleman, 1993; Dunwoody & Neuwirth, 1991; Griffin & Dunwoody,
2000; Morton & Duck, 2001). Whereas risk judgments at the personal level refers to perceived
91
harm on individuals themselves, risk judgments at the societal level refers to perceived harm on
other citizens or abstract population (Morton & Duck, 2001). As the measures that were used to
tap the perceived risks and benefits of nanotechnology in this study were at the societal level, it
is not surprising that the effect of interpersonal discussion is diminished. The circumstance is
rather different for public support for federal funding of nanotechnology. By making sense of
complex scientific information gathered from the mass media through talking and discussing
with other citizens, people will be able to form a more sophisticated knowledge structure about
the scientific world and make informed judgments about the emerging technology. Since
interpersonal discussions tend to reinforce mass media effects (Johnson, 1993) and the media
has portrayed nanotechnology and science in favorable terms, it is not surprising that scientific
discussion had a positive impact on public support for funding of nanotechnology.
Returning to the fundamental question of this study, the main effects of elaborative
processing and scientific discussion on public attitudes justify the validity of the scientific
literacy model. Simply put, it would be premature to nullify the contribution of scientific
knowledge at this juncture. Most extant studies, including the current one, examined textbookstyle factual scientific knowledge, which measured only one dimension of science knowledge.
People who engage in reflective integration tend to build a more sophisticated knowledge
structure about the scientific world, which in turn, has a bearing on their attitudes toward
emerging technologies. Therefore, the significant main effects of value predispositions and
cognitive processing variables suggest that the cognitive miser model and the scientific literacy
model are two complementary processes that individuals use to form opinions about
nanotechnology.
Finally, perceived risks-versus-benefits is negatively related to public support for federal
funding of nanotechnology. From the regression models, perceived risks-versus-benefits
92
partially mediated the influence of value predispositions, science communication, and reflective
integration on support for funding for the technology. This was evidenced by the reduction in
beta coefficients when perceived risks and benefits were entered into the regression equations.
This result suggests that perceived risks and perceived benefits play major roles in influencing
acceptance of nanotechnology and therefore reinforce the importance and worthiness of
examining the mechanisms through which the public form opinion about risks-versus-benefits
of the technology. It also suggests that some form of mediation is going on and it would be
worthwhile to examine the indirect effects of these factors on nanotechnology acceptance.
5.2. Explanations for Findings on Moderating Mechanisms
Perhaps the most noteworthy contribution of this study is the finding showing the
moderating role of cognitive processing in the communication-attitude link, providing partial
support for the differential gains model applied to a science communication context. First off,
the results show a transverse interaction pattern for the combined effects of science media use
and elaborative processing (i.e., intrapersonal reflection) on public perceived risks-versusbenefits. In other words, among the low elaborative processors, those who paid more attention
to science news in the media perceived significantly more risks/lesser benefits than did those
who paid less attention to science news in the media. Conversely, among the high elaborative
processors, those who paid more attention to science news in the media perceived significantly
lesser risks/more benefits than did those who paid less attention. These transverse interaction
patterns could be explained by the common journalistic practice of reporting two opposing
sides of an issue. Even though nanotechnology has been covered in positive terms, in most
cases, journalists are likely to offer both benefits and risks perspectives to achieve balance and
objectiveness in news coverage. Therefore, it is not surprising that for those people who paid
93
attention to but did not reflect on the news, they are likely to perceive greater risks/lesser
benefits than those who paid attention to and reflect on the news.
In addition, the study shows a significant interaction between science media use and
elaborative processing on public support for federal funding of nanotechnology. Specifically,
the results show a contingent interaction pattern, in which the relationship between science
media use and support for federal funding was significantly stronger for high elaborative
processors than for low elaborative processors. Given the fact that media coverage of
nanotechnology has been overwhelmingly positive, high elaborative processors are therefore
likely to absorb this information, resulting in greater support for funding of the technology than
do low elaborative processors. Consequently, only those individuals who ponder and actively
process media messages about science per se will recall information from the media and form
judgment support for federal funding of nanotechnology. These findings underscore the
importance of elaborative processing in the public opinion formation process.
On the other hand, the postulated interactions between science media use and science
discussion on public perceived risks-versus-benefits and support for federal funding of
nanotechnology did not hold. One explanation for this null effect may lie in the measurement of
science discussion in this study, in which it may be tapping only one type of discussion.
Previous research has shown that individuals who engage in heterogeneous discussion, i.e.,
discussion with diverse others, tended to be more exposed to dissimilar viewpoints and
therefore, likely to stimulate more mental activities and thinking; conversely, individuals who
engage in homogeneous discussions, i.e., discussions with people who are likeminded, would
tend to hear similar viewpoints that reinforce their existing ideas which may not stimulate as
much cognitive reflection. This distinction between homogenous and heterogeneous
94
discussions (McLeod et al., 1999; Scheufele, Nisbet, Brossard, & Nisbet, 2004) may be the reason
behind the null findings.
In addition, the posited relationships of intrapersonal and interpersonal reflections with
factual scientific knowledge were not significant. Whereas intrapersonal and interpersonal
reflections may engender a more sophisticated knowledge structure (e.g., procedural
knowledge) about the scientific world, it may not necessary bring about an increase in factual,
textbook style scientific knowledge which was measured in Study 1. As a result, these
hypothesized effects were non-significant.
Taken together, the significant interactions suggest that people still rely on new scientific
information gathered from the mass media to form attitudes toward nanotechnology. The effect
from the media was heightened when people paid attention to science news in the media and
reflected upon the messages they received. These results partially support the differential gains
model and suggest that media effects are more complex than it seems.
5.3. Explanations for Findings on Mediating Mechanisms
In addition to the moderating mechanisms, this study advances the processes
underlying public attitudes toward nanotechnology by examining the mediating mechanisms.
Notably, the impact of science media use was both direct and indirect, (a) via the informational
route, in which the news media provide an informal learning channel of scientific issues for the
public, and (b) via the heuristic route, in which the positive media frames served as cues or
shortcut for the miserly public when making judgments.
Applying the cognitive mediation model to understand the mediating mechanism, this
study shows that science media use promoted acquisition of factual scientific knowledge, both
directly and indirectly as mediated by elaborative processing. Factual scientific knowledge was,
95
in turn, negatively related to public perceived risks-versus-benefits of nanotechnology. These
results were in line with the major propositions outlined in the “cognitive mediation model”
(Eveland, 2001, 2002). Not only that, this study found that value predispositions including
religious beliefs and deference to scientific authority motivated science media use and reflective
integration. These findings were also consistent with parts of the cognitive mediation model in
which surveillance gratifications are postulated to lead to news attention and elaboration
(Eveland, 2001, 2002).
Moreover, the findings from the study show that elaborative processing mediated the
effect of science media use on perceived risks-versus-benefits, with elaborative processing
negatively related to perceived risks-versus-benefits. Likewise, elaborative processing also
mediated the influence of science media use on public support for federal funding of
nanotechnology, with elaborative processing positively related to support for funding. Echoing
this, science media use tended to stimulate scientific discussion, and science discussion in turn,
was positively associated with public support for federal funding of nanotechnology. These
additional informational pathways could be explained by the fact that media use stimulated a
more sophisticated and advanced knowledge structure about science among the public, and this
in turn propelled the public to make informed judgments about nanotechnology. This line of
reasoning parallels previous studies in the area of political communication (e.g., Sotirovic &
McLeod, 2001).
The heuristic pathways through which the mass media influenced public judgments
about nanotechnology were manifested through the direct effects; science media use had a
direct negative influence on public perceived risks-versus-benefits and a direct positive
influence on public support for federal funding of nanotechnology. These results are consistent
with findings from previous studies that indicated that individuals use positive news frames in
96
the media as cognitive shortcuts to form opinion about emerging science and technology (Lee et
al., 2005; Scheufele & Lewenstein, 2005). Another heuristic path emanated from the mediating
role of trust in scientists. Specifically, science media use and elaborate processing of science
news stories in the media promoted a sense of trust in scientists, and this sense of trust, in turn,
propelled the public to perceive more benefits relative to risks for nanotechnology. Again, these
are additional evidence suggesting that individuals draw on both heuristic cues and cognition
to form opinions about emerging technologies.
5.4. Implications
This study has important implications for theory and policy. Theoretically, this study
demonstrates that the scientific literacy model and the cognitive miser model are not mutually
exclusive. Rather, individuals tend to use a combination of both heuristic cues and cognitive
thinking to form opinions about emerging technologies. In particular, the relationships of mass
media use and reflective integration with public attitudes toward nanotechnology enable us to
draw a nexus between the two models. As exemplified by the direct effect of mass media,
individuals tended to use the positive frames about nanotechnology as heuristic cues in the
media to form judgments about the new technology. At the same time, through both the
moderating and mediating mechanisms of reflective integration, individuals tended to retrieve
new information and scientific knowledge from the mass media to form opinions about
nanotechnology. Therefore, scholars in the area of public understanding of science should
recognize that heuristics and cognitive thinking are complementary processes, and should
strike a middle-ground when examining both processes in future research.
Another major theoretical contribution of this study is the application of communication
theories into the research area of the public understanding of science. Thus far, this is the first
97
study to incorporate the notion of reflective integration into the area of science communication
and its impact on public attitudes toward nanotechnology. Moreover, this study extended the
differential gains model and the cognitive mediation model to a different context beyond
politics and examined different motivations behind mass media use in a non-political context.
This cross-domain approach not only enhance the value and generalizability of the differential
gains model, the cognitive mediation model, and the notion of reflective integration per se, it
also enable us to build a stronger and refined model in understanding how public form
attitudes toward science and emerging technologies. More importantly, this study contributes
to existing risk and science communication literature by bridging the disconnection between the
differential gains model and the cognitive mediation model and showing a more complex
process of how people form risks judgments and attitudes toward emerging technology. The
results of this study shows that people who pay attention to news and actively engage in
elaborative processing tended to possess greater cognitive sophistication about the scientific
world to perceive greater benefits than risks for nanotechnology and offer greater support for
funding of the emerging technology. The moderating role of elaborative processing, coupled
with its mediating role, thus sheds light on the additional mechanisms through which public
could form attitudes toward emerging technology. Instead of a simple direct media effects
model, the relationships between science media use, reflective integration, and public attitudes
toward nanotechnology are far more complex than previously assumed.
Above and beyond the contributions to communication theory and research in the
public understanding of science, the findings of this study point to several important practical
implications. Given that there are groups with different opinions about nanotechnology such as
the highly religious public, science communication practitioners should adopt the target
98
segmentation strategy, in which communication messages are tailored for publics from different
social backgrounds for maximum effect.
In addition, trust in nano-scientists both in academia and industry is crucial to public
support for nanotechnology. Therefore, government regulatory bodies should ensure that the
necessary guidelines are in place (e.g., guidelines to manage toxicity related to nanotechnology
and health standards for creating commercial products) so that public confidence and trust is
maintained.
Given the findings that the mass media play a key role in shaping public attitudes
toward nanotechnology via heuristic and/or informational routes, policymakers and scientists
should learn to focus on framing their messages in ways that connect with diverse audience. It
is important for public officials, scientists, and science communicators to pay attention to new
developments in media coverage of nanotechnology to monitor public opinion movements,
especially when the issue of nanotechnology enters into a different stage of the issue-attention
cycle. The mass media could also be a point of intervention for public officials as they could
provide accurate and up-to-date information about nanotechnology to the public so as to
sustain positive public opinion.
Since scientific discussion has been shown in this study to play a key role in mediating
the impact of the mass media on public formation of attitudes toward nanotechnology, it should
be worthwhile for policymakers to invest in large-scale public dialogue initiatives such as town
hall meetings, deliberative forums, and nano cafes. This inevitably generates another practical
question: In addition to dialogues with other non-likeminded citizens who hold dissimilar
viewpoints on nanotechnology, is it necessary for policymakers to invite nanotechnology
experts to these deliberative forums to have discussions with the public? Put differently, do
public attitudes toward nanotechnology differ significantly from that of the experts, to the
99
extent that dialogues between these two groups are necessary to close the attitudinal gaps
between them? Very often, scientists in the academia are preoccupied with conducting
laboratory research, publishing scientific results, and teaching college-level courses, which leave
them with little time for participating in dialogues with the public at forums or nano cafes. This
reallocation of time and effort for the scientists has to be justified with the evidence that a
attitudinal-gap indeed exists between the public and the scientists. Moreover, a attitudinal-gap
inquiry will also justify whether there is a need for scientists to participate in other outreach
efforts such as conducting seminars at churches to reach out to the highly religious citizens and
involving in thoughtful framing of messages in the news media while staying truthful to
scientific uncertainty. Therefore, using the key factors that were identified in this study, Study 2
will examine how the experts and the public differ in terms of their perceived risks-versusbenefits of nanotechnology and their level of support for federal funding of the emerging
technology. Details about Study 2 will be described in the next chapter.
100
CHAPTER 6
EXPERTS VERSUS PUBLIC ATTITUDES TOWARD NANOTECHNOLOGY
(STUDY 2)
Following the practical questions posed in Chapter 5, the aims of Study 2 are threefold.
First, Study 2 aims to examine how the experts and the public differ in terms of their perceived
risks-versus-benefits of nanotechnology and their level of support for federal funding of the
emerging technology. Second, and more importantly, Study 2 aims to examine how heuristic
cues, in the form of value predispositions and science media use variables that were identified
earlier on in Study 1, will influence experts and public perceived risks-versus-benefits of
nanotechnology and their support for federal funding of nanotechnology.8 Finally, this study
set out to determine if the experts use the same or different set of considerations to make
judgments of nanotechnology, in comparison with the public.
Study 2 is also a follow-up to a recent research study that was conducted by Scheufele
and his colleagues (2007) that examined the similarities and differences between the U.S.
nanotechnology scientists and the general public on their perceived risks and benefits. Briefly,
Scheufele et al. (2007) found that the experts were more optimistic than the public about the
potential benefits of nanotechnology. Conversely, the public were more concerned about the
potential risks of nanotechnology, including the potential loss of privacy or adverse economic
impacts, than were the experts. Despite the importance of these findings, it is important to
highlight that only descriptive statistical analyses were run for this particular study. In order to
There were no comparable measures for reflective integration in the expert survey dataset. As such,
elaborative processing and interpersonal discussion was left out from the analyses of Study 2. Future
research should attempt to examine if the experts and public would differentially use reflective
integration when forming opinions about nanotechnology.
8
101
adequately determine if the public were indeed different from the experts in terms of their
perceived risks and benefits, appropriate social and demographic factors must be controlled for.
Therefore, Study 2 will advance the findings in Scheufele et al.’s (2007) study by comparing the
public and expert opinions using multivariate statistical analyses. Furthermore, Study 2 will
also fill out a void in the previous research by examining the potential differences between the
public and the experts in terms of their level of support for federal funding of nanotechnology.
6.1. Differences in Expert and Public Judgments of Risk
Scholars in risk communication research generally believed that experts view risks
differently from members of the lay public, and that expert judgments are closer to reality than
those of the public (e.g., Cole & Withey, 1981; Sandman et al., 1987; Slovic, 1987). In other words,
expert judgments of risk are often viewed as objective and can be measured and quantified
scientifically, whereas public assessments of risk are often deemed as subjective and qualitative.
Slovic (1997) pointed out:
“Experts are seen as purveying risk assessments, characterized as objective, wise,
and rational – based upon the real risks. In contrast, the public is seen to rely
upon perceptions of risk that are subjective, often hypothetical, emotional,
foolish and irrational” (p. 278).
Cole and Withey (1981) echoed the view:
“Risk assessments obtained from a small group of ‘experts’ were highly
correlated with statistical data describing annual fatalities, thereby indicating
102
that, among experts, perceptions of risk were a function of the available
statistical evidence and little else” (p. 145) .
Likewise, Sandman, Weinstein, and Klotz (1987) maintained: “Whereas experts base
their judgments on mortality rates, the public relies on other types of information to evaluate
the riskiness of a threat” (p. 95).
The small but growing body of empirical studies on expert versus layperson
similarities/differences in risk judgment has led to the generally accepted conclusion that
experts who are knowledgeable in their field tend to perceive hazards within their area of
expertise as less risky than the lay public, and that expert judgments are more objective than
those of the general public, both across the wide spectrum of questions asked and across the
variety of substantive domains (see Rowe & Wright, 2001, for an overview). Differences in the
way that expert and the lay population judge risks have been observed across various domains,
including toxicology (Kraus et al., 1992), ecological risks to water environments (McDaniels et
al., 1997), global climate change (Lazo et al., 2000), computer technology (Gutteling &
Kuttschreuter, 2002), aviation (Thomson, Onkal, Avcioglu, & Goodwin, 2004), biotechnology
(Savadori et al., 2004), Mad cow disease (Raude, Fischler, Setbon, & Flahault, 2005), flood risks
(Michael Siegrist & Gutscher, 2006), and nanotechnology (Siegrist, Keller, Kastenholz, Frey, &
Wiek, 2007). There are only a few exceptions. For instance, Wright, Pearman, and Yardley (2000)
found that expert and the lay public shared similarities in risk judgment of hazardous events in
oil and gas production in the North Sea. In addition, Wright, Bolger, and Rowe (2002)
demonstrated that university students closely paralleled the expert underwriters when it comes
to estimating the likelihoods of potentially lethal events.
103
Despite this, methodological weaknesses in many of these studies may have attenuated
the true effect sizes of the relationships between expertise (i.e., status) and risks judgments. Put
it another way, differences in risk judgments between experts and the public could have been
wider if there are no inherent methodological problems in previous studies. There are two
noticeable problems in many extant studies. First, the representativeness and validity of the
expert sample in many previous studies are questionable. For example, Raude, Fischler, Setbon,
and Flahault (2005) examined how medical scientists and members of the general population
reacted to the mad cow disease in France. In this particular study, the scientists were sampled
from within one scientific institution (the French Institute of Health and Medical Research).
Aside from the low response rate of 19 percent in the study, sampling from one institution
could hardly be representative of the entire scientific community in France. In another
comparison study by Slovic, Fischhoff, and Lichtenstein (1985), the expert group used in the
study included such specialists as lawyers, economists, and geographers, who could arguably
appear in the lay group as well. Not only that, their sampling process only mentioned “selected
nationwide” and thus was not well defined. In several studies, details about the composition of
the expert groups were not reported (Barke & Jenkins-Smith, 1993; Flynn, Slovic, & Mertz, 1993;
Gutteling & Kuttschreuter, 1999; McDaniels et al., 1997; Wright et al., 2000).
In fact, Rowe and Wright (2001) summarized in their meta-review that the expert
samples generally represented quite diverse groupings across studies: toxicologists, computer
scientists, nuclear scientists, aquatic scientists, ecologists, loss-prevention managers in oil and
gas production, and scientists in general. To rectify this problem, Rowe and Wright (2001)
suggested that future studies provide information about the nature of the day-to-day activities
of the experts to allow readers to ascertain that the content of the risk questions posed is located
within the experts’ expertise and experience. This is important because it will enhance the
104
ecological validity of the study and ensure that we are measuring true effect size difference
between the expert and the lay public.
The second major problem that had plagued many previous studies is the failure to
control for social and demographic factors (e.g., age, gender, SES, etc.) that are associated with
risk judgments, when determining if expert and lay public differences are indeed due to
expertise. Rowe and Wright (2001) aptly pointed out:
“There are a number of key demographic and socioeconomic factors that have
been demonstrated to correlate with risk perception…….Unfortunately, none of
the empirical studies on this subject have attempted to match their expert and lay
samples on all of these factors. As such, these factors potentially confound
observed expert-lay differences that have been attributed to ‘expertise.’” (p. 348)
For example, a study by Kraus et al. (1992) claimed that group status (i.e., public versus
experts) was the most important predictor, with gender, race, and education accounting for
significant but small amounts of variance in response to questions on risk judgments. However,
the authors only provided descriptive breakdowns of the expert versus public responses, and
no higher-level statistical analysis was provided to support this claim. This problem could be
overcome by simply using ordinary regression, controlling for social and demographic factors
in the analysis.
Siegrist, Keller, Kastenholz, Frey, and Wiek (2007) conducted a study in Switzerland by
examining laypeople’s and experts’ attitudes toward 20 different nanotechnology applications
and three non-nanotechnology applications. Notably, they found that laypeople perceived
greater risks than did experts, and that risk judgment in each respective group was predicted by
105
different set of factors. Specifically, trust, perceived benefits, and general attitudes toward
technology influenced the perceived risk of laypeople; conversely, confidence in governmental
agencies was the only important predictor of risks associated with nanotechnology applications
among the experts. Despite these differences, the results also show that both laypersons and
experts judged asbestos as more risky than any single nanotechnology application. They
concluded that laypersons and experts use different cues to make risk judgments about
nanotechnology, with the experts having a unique position to independently assess risks and
benefits.
Even though Siegrist et al. (2007) contributed to our understanding of the similarities
and differences underlying layperson’s and experts’ perceived risks of nanotechnology, there
are some inherent problems in the study that may have limited the generalizability of the
results. First off, information on the sampling of experts was unclear. They indicated that
“People who attended recent conferences about nanotechnology or who work at research
laboratories in the field of nanotechnology were selected. Experts were contacted by e-mail and
asked to fill out a paper and pencil questionnaire” (p. 61). However, there was no information
about the response rate, sampling frame, and sampling technique of the expert sample. In other
words, a convenience sampling of nanotechnology experts was carried out in their study. If the
samples were not randomly selected, this might limit the generalizability of the findings.
Moreover, the sample size of both the layperson group (N = 375) and expert group (N = 46)
were rather small for surveys, potentially leading to Type II error, that is, rejection of an
alternative hypothesis when it may have found support.
Given the fact that public in the United States and Europe differ in their perceived risks
and benefits of nanotechnology (Cobb & Macoubrie, 2004; Gaskell, Eyck et al., 2005), it would
therefore, be worthwhile to conduct a separate study that compares how the public and experts
106
perceive the emerging technology in the United States. The current study seeks to examine
these differences between experts and lay public in terms of their perceived risks and benefits in
the United States, guided by two key goals: using a better sampling procedure with larger
sample sizes in both groups, and using appropriate multivariate analyses, to overcome the
inherent methodological problems in previous studies.
6.2. Expert and Public Differences in Levels of Support for Federal Funding of
Nanotechnology
Even though studies on layperson and experts differences in risk judgments of science
and technologies abound, extant literature examining the differences between expert and public
support for federal funding of science in general and nanotechnology in particular is scarce.
Nevertheless, it is worthwhile to examine the potential differences in the level of support for
federal funding of nanotechnology between the public and the experts. In addition, it is
worthwhile to assess if the experts and public use different sets of considerations to make
funding decisions about the emerging technology.
Specifically, would the experts make more rational judgments than do the public by
virtue of their expertise and knowledge (instead of using heuristic cues and value
predispositions) when making decisions about federal funding of nanotechnology? Also, would
differences in perceived risks-versus-benefits between the two groups have a differential effect
on their level of support for federal funding of nanotechnology? If indeed, the differences in
funding decisions between the public and the experts are large, and if the public do use more
affective considerations to make judgments than do the experts, then communication and
dialogue between these two groups would be necessary to bridge the gap.
107
6.3. Factors that Influence Perceived Risks-versus-Benefits and Support for Federal Funding
of Nanotechnology
As mentioned earlier, the perceived risks and benefits of individuals vary as a function
of various socio-demographic variables, such as gender, age, income, education, etc. (Bodmer,
1985; Brody, 1984; Lee et al., 2005; Miller & Kimmel, 2001; Miller et al., 1997; Siegrist, 1998, 2000;
Sparks, Shepherd, & Frewer, 1994). In particular, the most robust demographic predictor of
perceived risks and benefits is gender, where females are likely to perceive greater risks (and
less benefits) than do males (Bord & O'Connor, 1997; Davidson & Freudenburg, 1996; Gutteling
& Wiegman, 1993; Lee et al., 2005; Savage, 1993; Siegrist, 2000; Slovic, 1999). Among experts,
gender differences have also been found to influence risk judgments, in which female scientists
tend to perceive greater risks than do male scientists (e.g., Barke, Jenkins-Smith, & Slovic, 1997;
Kraus et al., 1992; Slovic et al., 1995). The impact of age has been mixed, with some studies
found significant influence of age on perceived risks (e.g., Morton & Duck, 2006) while others
have not found such differences (e.g., Lee et al., 2005; Morton & Duck, 2001). Nevertheless, both
gender and age will be included as control variables in the analysis in the current study to
ascertain that any differences in perceived risks-versus-benefits is due to scientific status (i.e.,
difference in expertise) instead of demographic factors.
According to availability heuristic (Tversky & Kahneman, 1982), people use easily
accessible instances or associations (from various sources) that could be brought readily to mind,
as cues to form judgments and make decisions. Notably, value predispositions are heuristic
cues or cognitive shortcuts that individuals often use to form opinions about risks and benefits
related to science and technology (e.g., Siegrist et al., 2007). Differences in religious, political,
and social worldviews have been found to be associated with variations in attitudes among
both the experts and the public. Religious guidance has been demonstrated to correlate at the
108
zero-order level with risks-versus-benefits of nanotechnology (e.g., Lee et al., 2005). Ascribing
trust to social institutions enables us to reduce uncertainty to an acceptable level and simplify
decisions involving large amount of information. Studies in the domains of genetic engineering
and gene technology have demonstrated that people who trusted scientific institutions
attributed more benefits and fewer risks (Siegrist, 2000; Siegrist & Cvetkovich, 2000; Tanaka,
2004).
Communication scholars have argued that the more positive framing of nanotechnology
in the mass media is likely to act as heuristic cues in influencing the risks and benefits
considerations among the public (e.g., Nisbet & Scheufele, 2007; Scheufele & Lewenstein, 2005).
Existing experimental studies have demonstrated that framing of nanotechnology have an effect
on how audience perceived risks and benefits of the technology (e.g., Cobb, 2005; Schutz &
Wiedemann, 2008). Particularly in the area of emerging technologies where most citizens have
little or no direct experience, media coverage of these technologies acts as a key heuristic for the
audience (Ho et al., 2008; Nisbet et al., 2003; Nisbet & Lewenstein, 2002). To get a good sense of
how the mass media could be used by the experts and the public to make risks and benefits
judgment about nanotechnology, science media use (or, attention to science news in the mass
media) will be examined.
In addition, results of Study 1 in Chapters 4 and 5 have shown that value
predispositions, including religious beliefs, deference to scientific authority, science media use,
and trust in scientists were key determinants of public level of support for federal funding of
nanotechnology. As such, these factors will also be examined in Study 2 to determine if they
have differential effects on funding support depending on scientific status.
109
Based on the above-mentioned considerations and the existing literature demonstrating
the differences between the public and the experts in terms of perceived risks, the following
hypotheses are postulated:
Hypothesis 17a: Experts will perceive lesser risks-versus-benefits of nanotechnology than
will the public, after taking into account all the appropriate control variables.
Hypothesis 17b: Experts will indicate greater support for federal funding of
nanotechnology than will the public, after taking into account all the appropriate control
variables.
Hypothesis 18a: The public will use more heuristic cues, in the form of value
predispositions and science news frames, when making judgments about perceived
risks-versus-benefits of nanotechnology than will experts.
Hypothesis 18b: The public will use more heuristic cues, in the form of value
predispositions and science news frames, when making judgments about support for
federal funding of nanotechnology than will experts.
In essence, the purpose of the present study is to investigate whether experts perceive,
and react differently to, nanotechnology than do the public, and, if such is the case, whether the
differences may be attributed to the main demographic predictors, value predispositions, and
science news frames, which have consistently been shown to account for individuals’ risk
judgments.
110
CHAPTER 7
METHODS AND RESULTS
(STUDY 2)
In Chapter 7, I will describe the methods and results of Study 2. In particular, the
methods section will elaborate on the data and sampling procedure employed, measures used,
and the analytical approach for testing the hypotheses posited in Study 2. Following this, I will
report the findings for the hypotheses in the results section.
7.1. Methods
Data for Study 2 came from a mail survey of 363 nanotechnology scientists and
engineers conducted by the University of Wisconsin Survey Center.9 A rigorous sampling
design was employed in which authors were identified from more than 90,000 nanotechnology
publications indexed in the ISI Web of Knowledge database between January 2005 and July
2006. To construct the target sample, the names and detailed contact information of a complete
list of roughly 1,000 U.S. scientists, who worked primarily in higher education and private
industry were compiled (contact information was obtained from the Internet and public
sources). These scientists were first or corresponding authors of their nanotechnology-related
work, which was cited at least five times in the publication database.
This sampling design of focusing on the most highly cited and most active scientists
within the field of nanotechnology would encapsulate viewpoints from scientists with an
9
Just like the 2007 public opinion data, the scientist survey was originally collected by Professor Dietram
A. Scheufele, under grants support from the National Science Foundation (SES-0531194) and the
University of Wisconsin-Madison Graduate School (135GL82). Again, I would like to acknowledge his
generosity in making these data available for my dissertation.
111
established track record in the field, and exclude scientists in unrelated disciplines who
happened to publish a nanotechnology-related topic during the timeframe outlined in the
sampling frame. In addition, the small number of graduate students who were listed as lead or
corresponding authors were excluded from the sample because most of them had relocated to
other labs or institutions by the time the survey went in the field and it was difficult to reliably
identify contact information for many of them.
The mail survey was administered in three waves, following Dillman’s Total Design
Method (AAPOR RR-3: 39.5 percent). The fieldwork was conducted from May to June 2007. The
approximate margin of error was +/- 5 percent. All respondents with valid addresses received
an initial full mailing, via First Class U.S. Mail, including a cover letter explaining the study, a
postage-paid return envelope, and a questionnaire sent on March 13, 2007. A postcard
reminder was mailed to respondents on March 20, 2007. A second full mailing was sent to all
respondents who had not returned a survey on April 20, 2007. The third wave was held for
mailing until June 6, 2007 in order to have them arrive soon after the end of the semester at most
universities, and was sent to corrected addresses when available. The University of Wisconsin
Survey Center concluded the field period and began data delivery on June 22, 2007. (Sampling
procedure of public responses to nanotechnology was described earlier in Chapter 4.)
7.1.1. Measures
Only identical or similar measures in the expert and public samples were included in the
analyses. Variables such as elaborative processing, interpersonal discussion, and factual
scientific knowledge were not included for comparisons because these questions were not
measured in the expert survey. In many ways, the expert survey questionnaire was an abridged
version of the 2007 public opinion survey questionnaire. Table 7.1 shows the descriptive
112
statistics and the actual question wording from the original expert survey questionnaire for
each item examined in this study.
7.1.1.1. Outcome Variables
Support for federal funding of nanotechnology. The first outcome variable was measured
using a single-item measure for both the expert and the public samples: “Overall, I support
federal funding for nanotechnology.” For the expert sample, the outcome variable was
measured on a 5-point scale, from 1 “do not agree at all” to 5 “agree very much” (M = 4.69, SD
= .66). For the public sample, the outcome variable was measured on a ten-point scale (1 = “do
not agree at all,” 10 = “agree very much”). This item in the public sample was therefore recoded
(e.g., values of 1 and 2 were recoded into 1, and values of 3 and 4 were recoded into 2, etc.) to
correspond to the same metric and range as the item in the expert sample (M = 3.22, SD = 1.34).
Figure 6.1 indicates that the experts were overwhelmingly more supportive of federal funding
of nanotechnology than were the public. Specifically, 95.6 percent of the experts agree with the
statement that “Overall, I support federal funding for nanotechnology” as compared with 42.6
percent of the public who agree with the statement.
Perceived risks-versus-benefits of nanotechnology. Next, perceived risks of nanotechnology was
an additive index of seven items for both the expert and the public samples: (a) “Nanotech may
lead to the loss of personal privacy because of tiny new surveillance devices,” (b) “Nanotech
may lead to an arms race between the U.S. and other countries,” (c) “Nanotech may lead to new
human health problems,” (d) “Nanotech may be used by terrorists against the U.S.,” (e)
“Because of nanotech we may lose more U.S. jobs.,” (f) “Nanotech may lead to the
uncontrollable spread of very tiny self-replicating robots,” and (g) “Nanotech may lead to more
pollution and environmental contamination.” These items were measured on five-point scales
in the expert sample from 1 “strongly disagree” to 5 “strongly agree” (M = 16.65, SD = 5.00,
113
Cronbach’s alpha = .77). For the public sample, these items were measured on ten-point scales
from 1 “do not agree at all” to 10 “agree very much.” As with support for federal funding, the
items in the public sample were recoded (e.g., values of 1 and 2 were recoded into 1, and values
of 3 and 4 were recoded into 2, etc.) to correspond to the same metric and range as those scales
in the expert sample (M = 18.87, SD = 5.77, Cronbach’s alpha = .82).
Likewise, perceived benefits of nanotechnology was measured with an additive index of
seven items for both samples: (a) “Nanotech may lead to new and better ways to treat and
detect human diseases,” (b) “Nanotech may lead to new and better ways to clean up the
environment,” (c) “Nanotech may give scientists the ability to improve human physical and
mental abilities,” (d) “Nanotech may help us develop increased national security and defensive
capabilities,” (e) “Nanotech may lead to technologies that will help solve our energy problems,”
(f) “Nanotech may revolutionize the computer industry,” and (g) “Nanotech may lead to a new
economic boom.” For the expert sample, these items were measured on five-point scales
ranging from 1 “strongly disagree” to 5 “strongly agree” (M = 28.89, SD = 4.55, Cronbach’s
alpha = .83). For the public sample, these items were measured on ten-point scales from 1 “do
not agree at all” to 10 “agree very much.” Again, these items were recoded in the public sample
such that they are transformed into the same metric and range as those scales in the expert
sample (M = 25.41, SD = 6.59, Cronbach’s alpha = .90).
Exploratory factor analyses using principal axis factoring extraction method was
conducted to determine the dimensionality of each of the composite measures for perceived
risks and perceived benefits in the public and expert samples. For each analysis, only
meaningful factor with an eigenvalue that was greater than 1.00 was extracted. For public
perceived risks, only one factor emerged in the analysis, with an eigenvalue of 3.407 which
explained 48.67 percent of the total variance. For experts’ perceived risks, one meaningful factor
114
emerged, with an eigenvalue of 2.975 which explained 42.504 percent of the total variance.
When it comes to public perceived benefits, one factor with an eigenvalue of 4.579 which
accounted for 65.42 percent of the variance was extracted. Likewise, for experts’ perceived
benefits, one factor with an eigenvalue of 3.561 emerged from the analysis, which accounted for
50.87 percent of the total variance. Therefore, the results of the exploratory factor analyses
revealed the unidimensionality of each of the composite measures for perceived risks and
perceived benefits in the public and expert samples. The results also suggested that the same
patterns emerged for the public and experts sample, and provided further justification for
combining the items into additive, composite measures.
For both the expert and public samples, the outcome variable was measured by
subtracting perceived benefits of nanotechnology from perceived risks of nanotechnology, with
higher scores indicating greater perceived risks (Expert: M = -12.24, SD = 6.24; Public: M = -6.51,
SD = 6.98).
As shown in Figures 6.2 and 6.3, experts were more optimistic about the potential
benefits than did the general public, but they were also more concerned about environmental
and health risks. For example, experts were significantly more likely than the general public to
agree that nanotechnology may lead to “new and better ways to treat and detect human
diseases” (92 percent for experts; 64 percent for public) or to “new and better ways to clean up
the environment” (83 percent for experts; 49 percent for public). Members of the general public,
in contrast, were more concerned about many of the potential drawbacks of nanotechnology,
such as the “loss of personal privacy because of tiny new surveillance devices” (44 percent for
public; 30 percent for experts) or the loss of “more U.S. jobs” (38 percent for public; 6 percent for
experts). The exceptions were two areas in which experts expressed higher levels of concern
than did the general public: the potential of nanotechnology to “lead to more pollution and
115
environmental contamination” (19 percent for experts; 14 percent for public) and “to new
human health problems” (31 percent for experts; 21 percent for public). On the whole, the
public expressed more concern about the potential risks of nanotechnology than did the experts.
7.1.1.2. Independent Variables
Religious beliefs. For both samples, respondents were asked to indicate on a ten-point
scale (1 = “no guidance at all,” 10 = “a great deal of guidance”), how much guidance does
religion provide in their everyday life (Expert: M = 3.42, SD = 2.91; Public: M = 6.00, SD = 3.01).
Deference to scientific authority. The following two items were used to tap respondents’
deference to scientific authority: (a) “Scientists know best what is good for the public,” and (b)
“Scientists should do what they think is best, even if they have to persuade people that it is
right.” For the public sample, the respondents were asked to indicate on a ten-point scale, from
1 “do not agree at all” to 10 “agree very much” the extent to which they agree with the two
statements. The items were averaged to create a composite scale (M = 4.30, SD = 2.02, r = .39, p
< .001). For the expert sample, the two items were measured on a five-point scale, from 1
“strongly disagree” to 5 “strongly agree.” Likewise, the items were averaged to create a
composite scale (M = 3.42, SD = .81, r = .24, p < .001). The new index in both samples was
standardized before data analysis to ensure that they were in the same metric.
Science media use. For the public sample, respondents were asked to indicate how much
attention they paid to the following items when they read the newspapers, watch television,
and read online content on a ten-point scale (0 = “no attention at all,” 10 = “very close
attention”): (a) “Stories related to science and technology,” (b) “Stories about scientific studies
in new areas of research such as nanotechnology,” and (c) “Stories about the social or ethical
implications of emerging technologies.” This corresponded to nine separate items. Cronbach’s
alpha reliability coefficient indicated high internal consistency among these nine items (X = .89).
116
Therefore, these items were averaged to create a composite index, with higher score indicating
greater amount of attention (M = 4.73, SD = 2.12). Comparable questions were posed to the
experts. In the expert sample, respondents were asked to indicate how much attention they paid
to the following kinds of content when they read the newspapers, watch television, and read
online content on a five-point scale (1 = “none,” 5 = “a lot”): (a) “Science and technology outside
of your own field of research;” and (b) “The social or ethical implications of emerging
technologies.” This corresponded to six items and they were averaged to create a composite
index, with higher score indicating greater amount of attention (M = 3.17, SD = .92, Cronbach’s
alpha = .75). To bring them to the same metric, the science media use index was standardized
prior to analysis.
Trust in scientists was measured using the following two items, in which respondents
were asked how much they trust: (a) “University scientists doing research in nanotechnology”
and (b) “Scientists working for the nanotech industry.” For the public sample, respondents were
specifically asked to indicate on a ten-point scale (1 = “do not trust their information at all,” 10 =
“trust their information very much”), how much they trust the above sources of information to
tell them the truth about the risks and benefits of nanotechnology. The items were averaged to
create an index, with higher scores indicating greater levels of trust (M = 6.16, SD = 2.00, r = .58,
p < .001). In a near identical question format, respondents in the expert sample were asked to
indicate, on a five-point scale (1 = “not at all,” 5 = “very much”), the degree to which the abovementioned groups currently have the necessary scientific expertise to communicate about risks
and benefits related to nanotechnology. The items were averaged to create an index, with
higher scores corresponding to greater levels of trust (M = 4.21, SD = .80, r = .61, p < .001). The
index in both samples was standardized before data analysis to ensure that they are in the same
metric.
117
7.1.1.3. Control Variables
To ensure that the two samples were comparable, only age and gender were included as
control variables in the analyses. For the public sample, age was measured as a continuous
variable (M = 46.15, SD = 17.07) and gender was measured as a dichotomous variable (51.4
percent females). Similarly, for the expert sample, age was measured as a continuous variable
(M = 44.94, SD = 10.42) and gender was measured as a dichotomous variable (14.0 percent
females).
7.1.2. Analytical Approach
To test the hypotheses, two sets of ordinary regression analyses were run, one for
perceived risks-versus-benefits of nanotechnology and the other for support for federal funding
of nanotechnology. For each of the regression models, the independent variables were entered
in blocks according to their assumed causal order.
With respect to perceived risks-versus-benefits of nanotechnology as the outcome
variable, two separate regression models were run, one for the public sample and one for the
expert sample. In both models, demographic variables were entered first, followed by the traitlike value predisposition variables, science media use, and state-like value predisposition. In
the third regression model, the public and the expert samples were aggregated for analysis. A
new dummy variable, “scientific status,” was created in which the public was coded as “0” and
the experts was coded as “1.” Demographic variables were entered first, followed by the main
effects of religious beliefs, deference to scientific authority, science media use, trust in scientists,
and scientific status. Finally, scientific status was multiplied with each of the independent
variables to create interaction terms that were entered in the last block. The independent
variables and the dummy variable were standardized before multiplication to avoid possible
multicollinearity problems between the interaction term and its components (Cohen et al., 2003).
118
In the analysis, six multiplicative terms were included in the final regression block: (1) the
interaction between religious beliefs and scientific status, (2) the interaction between deference
to scientific authority and scientific status, (3) the interaction between science media use and
scientific status, and (4) the interaction between trust in scientists and scientific status.
A similar analytical approach was used for support for federal funding of
nanotechnology as the outcome variable. Two regression models were run, one for the public
sample and one for the expert sample. In both models, the independent variables were entered
in the following order: demographic variables (block 1), trait-like value predispositions (block 2),
science media use (block 3), state-like dispositions (block 4), and finally, perceived risks-versusbenefits (block 5). In the third regression model used to test for significant interactions, the
public and expert samples were aggregated for analysis. For this aggregated sample, the
independent variables were entered in the regression model as mentioned above, followed by
the standardized interactions terms in block 6: (1) the interaction between religious beliefs and
scientific status, (2) the interaction between deference to scientific authority and scientific status,
(3) the interaction between science media use and scientific status, (4) the interaction between
trust in scientists and scientific status, and (5) the interaction between perceived risks-versusbenefits and scientific status. As the number of missing values in the data was very small (i.e.,
less than 2 percent for each of the independent variables concerned in both the 2007 public
opinion survey and the expert opinion survey) and MCAR was assumed in this dissertation, all
missing data in the ordinary regression analyses in Study 2 were treated with mean substitution.
7.2. Results
In this section, tests of the potential differences between experts and public attitudes
toward nanotechnology are reported. Before running the multivariate analyses, a bivariate
119
correlation analysis was run for the expert sample to preliminary examine the relationships
among all the variables at the zero-order level. SPSS was used for the bivariate analysis with
listwise deletion procedure. Table 7.2 shows the bivariate correlations among the variables in
the expert sample. As indicated in the table, most of the variables in the sample were not highly
correlated with one another at the zero-order level. As such, the factors may not predict experts’
perceived risks-versus-benefits and support for federal funding of nanotechnology as well as
that in the public sample. The ordinary regression results will be reported next.
7.2.1. Experts versus Public: Factors Predicting Perceived Risks-versus-Benefits of
Nanotechnology
Table 7.3 shows the ordinary regression model of factors predicting perceived risksversus-benefits of nanotechnology for the public sample. As seen from the table, most of the
factors significantly predicted public perceived risks-versus-benefits of the emerging
technology. Females displayed a significantly higher perceived risks-versus-benefits than did
males (β = .13, p < .001), but age was not significantly related to perceived risks-versus-benefits.
The demographic variables accounted for 3.50 percent of the total variance in perceived risksversus-benefits of nanotechnology.
With regard to the effects of trait-like value predispositions, levels of religious beliefs
showed robust positive relationship with perceived risks-versus-benefits (β = .18, p < .001).
Deference to scientific authority was initially related to perceived risks-versus-benefits at the
zero-order level (as evidenced by the significant correlation), but its influence was fully
mediated by trust in scientists, introduced in the subsequent blocks. The trait-like
predisposition variables explained an additional 5.70 percent of the variance in the outcome
variable.
120
For science communication, science media use was negatively associated with perceived
risks-versus-benefits (β = -.19, p < .001). Science media use alone explained 6.70 percent of the
variance in the outcome variable. Likewise, trust in scientists displayed a strong negative
relationship with the outcome variable (β = -.29, p < .001). The state-like value disposition
accounted for a substantial amount of variance in perceived risks-versus-benefits of
nanotechnology (6.70 percent). The overall regression model for the public sample accounted
for 22.6 percent of the total variance in perceived risks-versus-benefits of nanotechnology.
Table 7.4 displays the ordinary regression model of the factors predicting perceived
risks-versus-benefits of nanotechnology among the experts sample. The results show that the
perceived risks-versus-benefits was not well-predicted by the factors among the experts. Even
though age and gender had no effect on experts’ perceived risks-versus-benefits, the
demographic block accounted for 1.50 percent of the variance in the outcome variable.
Of the trait-like predispositions, religious beliefs had no significant effect on experts’
perceived risks-versus-benefits. On the other hand, deference to scientific authority had a
significant negative association with the outcome variable (β = -.13, p < .01). The trait-like
predispositions block explained an additional 3.20 percent of the variance in the outcome
variable.
In addition, science media use was negatively related to experts’ perceived risks-versusbenefits of nanotechnology (β = -.10, p < .05), accounting for an additional 1.30 percent of the
variance in the outcome variable. Similarly, trust in scientists was negatively associated with
experts’ perceived risks-versus-benefits (β = -.24, p < .001), in which trust in scientists explained
5.50 percent of the variance in the outcome variable. The overall regression model for the expert
sample accounted for 11.5 percent of the total variance in the outcome variable.
121
Results of the regression analysis of the aggregated sample of the public and experts
were shown in Table 7.5. Notably, after controlling for the demographics, trait-like value
predispositions, science media use, and trust in scientists, scientific status was significantly
related to perceived risks-versus-benefits of nanotechnology (β = -.31, p < .001). In other words,
the experts perceived significantly higher benefits and lesser risks about nanotechnology than
did the public. Scientific status alone accounted for 7.00 percent of the total variance in
perceived risks-versus-benefits. Therefore, Hypothesis 17a was supported.
To assess whether factors predicting perceived risks-versus-benefits were indeed
substantially different between the two groups, a final interactions block was entered in the
regression model. The interaction between religious beliefs and scientific status (β = -.07, p < .01)
on the outcome variable were significant, after accounting for all controls. However, the
interactions between deference to scientific authority and scientific status, between science
media use and scientific status, and between trust in scientists and scientific status, were not
significantly related to the outcome variable. The overall regression model accounted for 33.4
percent of the variance in perceived risks-versus-benefits of nanotechnology. These results lend
partial support to Hypothesis 18a. As shown in Figure 7.1, among the public, respondents with
high level of religious beliefs were significantly more likely to indicate greater perceived risksversus-benefits of nanotechnology than were those with low level of religious beliefs.
Conversely, such a difference was not apparent among the experts.
7.2.2. Experts versus Public: Factors Predicting Support for Federal Funding of
Nanotechnology
Table 7.6 shows the ordinary regression model for factors predicting public support for
federal funding of nanotechnology. As demonstrated in the table, the results show that age was
negatively related to public support for federal funding of nanotechnology (β = -.07, p < .01). On
122
the other hand, gender had no influence on the outcome variable. The demographic variables
explained 2.70 percent of the variance in the outcome variable.
With regard to the trait-like value predispositions, religious beliefs was found to be
negatively related to public support for federal funding of nanotechnology (β = -.07, p < .05).
Deference to scientific authority was positively related to the outcome variable (β = .14, p < .001).
The trait-like value predispositions accounted for an additional 10.2 percent of the variance in
the outcome variable.
Likewise, science media use was positively associated with public support for federal
funding of nanotechnology (β = .14, p < .001), accounting for an additional 6.90 percent of the
variance in the outcome variable. Trust in scientists was positively related with public support
for funding (β = .21, p < .001), explaining 6.40 percent of the variance in the outcome variable.
Finally, public perceived risks-versus-benefits negatively predicted their level of support
for federal funding of nanotechnology (β = -.26, p < .001), with the independent variable
explaining an additional 5.00 percent of the variance in the outcome variable. Taken together,
the overall regression model explained a total of 31.2 percent of the variance in public support
for federal funding of nanotechnology.
Table 7.7 indicates the results of the regression analysis for factors predicting experts’
level of support for federal funding of nanotechnology. The results show that older experts
tended to indicate less support for federal funding of nanotechnology than did younger experts
(β = -.12, p < .05). Gender had no effect on experts’ opinion about funding. The demographic
block explained 3.10 percent of the variance in the outcome variable.
In addition, the more religious the experts were, the less likely they were to indicate
support for federal funding of nanotechnology (β = -.12, p < .05). Conversely, deference to
scientific authority had no significant effect on the outcome variable. The trait-like value
123
predispositions block accounted for an additional 2.30 percent of the variance in the outcome
variable.
Both science media use and trust in scientists did not predict experts’ support for federal
funding of nanotechnology. On the other hand, experts’ perceived risks-versus-benefits
negatively predicted experts’ support for federal funding of nanotechnology (β = -.24, p < .001),
accounting for an additional 5.20 percent of the variance in the outcome variable. Overall, the
regression model predicted a mere 12.6 percent of the variance in experts’ support for federal
funding of nanotechnology.
In addition, Table 7.8 shows the regression analysis of the combined sample of the
public and the experts. Specifically, after accounting for the demographic variables, trait- and
state-like dispositions, and science media use, scientific status was significantly related to
support for federal funding of nanotechnology (β = .48, p < .001). Put differently, the experts
perceived significantly greater support for federal funding of nanotechnology than did the
public. The scientific status variable explained 16.5 percent of the variance in our outcome
variable. Hence, Hypothesis 17b was supported.
Again, to examine whether factors predicting support for federal funding of
nanotechnology were substantially different between the experts and the public, a final
interactions block was entered in the regression model. With the exception of the religious
beliefs and status interaction, all the rest of the interactions were significantly related to support
for federal funding of nanotechnology. As shown in Figure 7.2, among the public, respondents
with high deference for scientific authority were significantly more likely to indicate greater
support for federal funding of nanotechnology than were those with low deference; in contrast,
such a difference was negligible among the experts. Even though there seems to be a slight
slope for the expert sample, this could be due to “regression toward the mean,” a phenomenon
124
whereby members of a population with extreme values on a given measure for one observation
will, for purely statistical reasons probably give less extreme measurements on other occasions
when they are observed. Likewise, as indicated in Figure 7.3, among the public, those who paid
high amount of attention to science news media were significantly more likely to indicate
support for federal funding of nanotechnology than were those who paid low amount of
attention. No such significant difference was found among the experts. Figure 7.4 shows that
the public who have a high trust in scientists were significantly more likely to indicate support
for funding than were those who have a low trust in scientists; there is no effect among the
experts. Interestingly, Figure 7.5 shows that perceived risks-versus-benefits had differential
effects on the public and the experts. Specifically, the public with higher perceived risks-versusbenefits were significantly less likely to indicate support for federal funding of nanotechnology
than were those with lower perceived risks-versus-benefits. Again, such a difference was not
found among the experts. Therefore, these findings provide partial support for Hypothesis 18b.
125
CHAPTER 8
DISCUSSION (STUDY 2)
Chapter 8 will discuss the findings of Study 2 by providing explanations for experts and
public differences in perceived risks-versus-benefits and explanations for experts and public
differences in support for federal funding of nanotechnology. This chapter will also discuss the
limitations and the implications of Study 2 for theory and practice.
Using large and representative samples of the lay public and experts collected in the
United States, Study 2 set out to examine the factors influencing perceived risks-versus-benefits
of nanotechnology and support for federal funding of nanotechnology in a public sample and
an expert sample. Overall, the regression analyses provide partial support for the hypotheses
regarding the impact of scientific status (i.e., experts versus lay public) on perceived risksversus-benefits of nanotechnology. Notably, two major findings were demonstrated. First,
compared with the experts, the public judged nanotechnology as having more risks and lesser
benefits, after controlling for all appropriate factors such as demographic variables. Second,
experts, equipped with their professional training and experience, used relatively less heuristic
cues such as religious guidance, to make risks-versus-benefits judgment of nanotechnology than
did the public.
With respect to funding of nanotechnology, the regression analyses provide strong
support for the hypotheses on the effect of scientific status on support for federal funding of
nanotechnology. In particular, experts indicated greater support for federal funding of the
emerging technology than did the public, after accounting for all the appropriate control
variables. Moreover, experts drew on significantly less heuristic cues in the form of value
predispositions and science media frames to make decision about funding support for
126
nanotechnology than did the public. Taken together, these findings suggest that the experts are
in a position to independently assess risks and benefits, and indicate that experts and the public
use different considerations to make judgments about risks and benefits of the emerging
technology.
8.1. Explanations for Experts and Public Differences in Perceived Risks-versus-Benefits
As expected, the lay public perceived greater risks-versus-benefits of nanotechnology
than did the experts, consistent with the results of most previous studies (e.g., Lazo et al., 2000;
Savadori et al., 2004; Siegrist et al., 2007). Important factors including demographic and value
predisposition variables were controlled for in the analysis to ensure that differences in the
outcome variable is truly a function of scientific status, and not a consequence of the
confounding demographic factors. This status gap potentially stems from the fact that experts
deal with the technology on a daily basis and they are more knowledgeable about the
technology. Therefore, their estimation of the risks and benefits of the technology may be closer
to reality than that of the public, and experts may be more accepting of risks from
nanotechnology than are the public. To the extent that they perceive risks-versus-benefits to be
larger, the lay public may feel there is a greater need for policy intervention.
However, the hypothesis that the experts are able to independently assess risks and
benefits, and therefore employ less heuristic cues to make risks-versus-benefits judgment of
nanotechnology than did the public was, at best, partially supported in this study. Even though
in the separate regression models, value predispositions and science media use explained a
relatively larger amount of variance in public (22.6 percent of variance explained) perceived
risks-versus-benefits of nanotechnology than did the experts (11.5 percent of variance
explained), only the impact of the interaction between scientific status and religious beliefs on
127
perceived risks-versus-benefits was significant in the combined regression model. In other
words, both the experts and the public use heuristic cues – including deference to scientific
authority, science news frames, and trust in scientists – to form risks-versus-benefits judgment.
The fact that deference to scientific authority and trust in scientists did not show
differential effects on perceived risks-versus-benefits depending on scientific status is not as
surprising as it appears. The experts conduct laboratory research on nanotechnology on a dayto-day basis and therefore, they should be very confident and certain about their own and other
scientists’ capability and integrity in handling the potential risks related to the emerging
technology. As such, it is not surprising to find that deference to scientific authority and trust in
scientists influenced both the public and experts’ judgments. Perhaps significant interactions
may have been found if this study were to examine trust in other non-scientific institutions,
such as governmental agencies (e.g., the White House and the Environmental Protection
Agency) or non-profit organizations (e.g., environmental activists).
Besides this, science media use had an influence on both the public and experts’
perceived risks-versus-benefits of nanotechnology. This highlights the strong impact of mass
communication channels, potentially in the form of heuristic cues (positive media frames) and
information, on both experts and public attitude. This is not surprising since news coverage of
nanotechnology has been overwhelmingly positive (e.g., Cobb, 2005; Stephens, 2005), and
therefore, the benefits frame could influence both the experts and the public. This study may
have been able to find a differential effect of science media use on perceived risks-versusbenefits contingent on scientific status if it were to examine media use separately by medium
(i.e., television news, print news, and online news). For instance, due to the nature of the
Internet, those who attended to online science news are likely to be interested in science or to be
more knowledgeable about scientific issues from the start. Therefore, those who attended to
128
science news online would naturally gravitate towards greater perceived benefits/ lesser
perceived risks than those who do not. Of course, these are speculations that should be verified
in future research.
On the other hand, the significant interaction results show that scientific status
moderated the impact of religious beliefs on respondents’ perceived risks-versus-benefits. Put
differently, those who are highly religious are significantly more likely to indicate higher
perceived risks-versus-benefits than those who are low on religiosity among the public. No
such effect was found for the experts. Due to their expertise and knowledge, the experts would
be able to put aside their religious beliefs to give a more objective assessment of the risks and
benefits of nanotechnology than would the lay public. Echoing recommendations in Study 1,
the finding from the current study suggests that industry and university scientists may need to
work with religious institutions such as churches and synagogues to convey the message that
science is not necessary an antithesis to religion, and to provide accurate information about
nanotechnology to the public.
Therefore, the assertion that the experts are indeed more objective in their judgment,
using more objective and rational reasoning (such as statistical evidence) and concomitantly less
heuristic cues, of risks and benefits assessments than the public was partially supported by the
findings in this study. Other sets of value predispositions such as trust in governmental
agencies must be examined in future studies that compares opinions of experts and lay public
to gather more support for this conclusion.
129
8.2. Explanations for Experts and Public Differences in Support for Federal Funding of
Nanotechnology
Unlike perceived risks-versus-benefits, the lay public seemed to use significantly more
heuristic cues to form opinion about support for federal funding of nanotechnology than did
the experts, providing a strong support for the hypothesis that the experts are more objective in
their judgments when it comes to funding decisions. To recapitulate, the lay public support for
federal funding of nanotechnology were predicted by age, religious beliefs, deference to
scientific authority, science media use, trust in scientists, and perceived risks-versus-benefits.
These factors explained 31.2 percent of the variance in the outcome variable. Conversely,
experts’ level of support for federal funding of nanotechnology was predicted only by age,
religious beliefs, and perceived risks-versus-benefits with these factors explaining a small 12.6
percent of the variance in the outcome variable. This indicates that experts and the public use
different considerations to make decision about funding support for the emerging technology.
The extent to which scientific status moderated the heuristic factors in the combined
regression model provided stronger evidence for this conclusion. Notably, scientific status
moderated the influence of deference to scientific authority on support for federal funding of
nanotechnology. The influence of deference to scientific authority was significantly larger
among the public than among the experts. Likewise, trust in scientists was positively associated
with perceived risks-versus-benefits among the public, whereas no such difference was found
for the experts. A similar moderation effect on support for federal funding was found for
science media use as well.
Although perceived risks-versus-benefits had a significant influence on both the public
and experts’ support for federal funding of nanotechnology in the separate regression models,
in the combined sample, the result shows that scientific status significantly moderated the
130
impact of perceived risks-versus-benefits on funding support; the effect is larger among the
public. This finding highlights the importance of perceived risks-versus-benefits as a factor that
could create a large divide between experts’ and the lay public’s levels of support for funding of
the new technology.
Based on the results of the study, an interesting question arises: Why did heuristic cues
display differential effects on support for federal funding between the experts and the public,
but relatively smaller differential effects on perceived risks-versus-benefits? One plausible
explanation may be that support for funding is a more subjective judgment relative to perceived
risks and benefits. Hence, it is difficult to give an “objective” response to support for funding
without evoking value predispositions such as one’s religious beliefs and trust in scientists. As
such, public may place a greater emphasis on heuristic cues for funding issues instead of
perceived risks-versus-benefits. In line with this, the fact that support for federal funding of
nanotechnology is a more subjective judgment relative to perceived risks-versus-benefits may
shed light on why scientific status did not moderate the effect of religious beliefs on support for
federal funding of the emerging technology in this study.
8.3. Implications
Given that the results show a gap in perceived risks-versus-benefits and level of support
for federal funding of nanotechnology between the experts and the general public that is
indicative of a communication deficit, it should therefore be worthwhile to invest in risk
communication programs to convey expert judgments of risks and benefits to the public.
Furthermore, the results of this study strengthened the recommendations in Study 1 in which it
should be worth the effort to conduct consensus conferences for the public to have dialogues
with scientists and elites to bridge the gaps in attitudes between the two groups.
131
Investing in the mass media to convey about risks and benefits to the public should also
be useful as this study shows that science news frames are often used by the public as heuristic
cues to make judgments. Moreover, the relatively low coverage of health and environmental
risks of nanotechnology in the mass media provide industry and university scientists the
opportunity to educate the public by giving them factual information about nanotechnology
and by engaging them in a worthwhile dialogue about the emerging technology. This is
especially so to avoid the debacle that biotechnology had encountered previously (Priest, 2000).
132
CHAPTER 9
OVERALL DISCUSSION AND CONCLUSION
In Chapter 9, I will summarize the results of Study 1 and Study 2 and highlight the
major theoretical, conceptual, and theoretical contributions of this dissertation. Following this, I
will describe the limitations of both studies and the directions for future research. Finally, I will
end with a final conclusion.
9.1. Summary
This dissertation adopted a holistic approach to look at the intersection of mass media,
public opinion, and expert opinion about the emerging science of nanotechnology comprised of
three major components: assessing the moderating and mediating mechanisms behind how
public form opinion about nanotechnology, comparing the expert and public opinion of the
emerging technology, and examining the simultaneous influence of mass media on public and
expert opinion of the novel science.
Descriptive analyses of the 2004 and 2007 survey of public opinion of nanotechnology
reveal that there were only slight variations in terms of public attitudes in both years due in
part to the fact that nanotechnology is still at the early stage of the issue-attention cycle in the
mass media. Even though the public are largely unaware and under-informed about
nanotechnology, they tend to perceive greater benefits than risks related to nanotechnology and
to be supportive of federal funding of nanotechnology in both years. How did this positive
public opinion of nanotechnology come about? One major source of this positive opinion may
stem from the mass media.
133
Using a simple analysis of media content, this dissertation showed that nanotechnology
was first covered in the high circulation newspapers in the late 70s and early 80s. The issue
made its foray into the medium circulation newspapers in the late 80s. News about
nanotechnology started to appear in the low circulation newspapers in the late 90s, suggesting
that nanotechnology has gained prominence over time, evolving from a solely elite issue to a
local issue which is closer at heart to the citizens. In addition, even though the number of news
articles about nanotechnology had been climbing steadily from 1999 to August 2008, the
percentage of risk-related articles remained somewhat low, indicating that media coverage of
nanotechnology has been overwhelmingly positive, highlighting the benefits of the emerging
technology over its risks.
To take a closer look at the relationship between mass media and public opinion about
nanotechnology, this dissertation attempted to build a more sophisticated theory-driven model
by examining the influence of mass media alongside other cognitive and heuristic factors on
public attitudes of nanotechnology. In particular, this dissertation examined the moderating
and mediating mechanisms behind public attitudes toward the emerging technology, by
situating the queries within the debate between the scientific literacy model and the cognitive
miser model. Above and beyond the main effects of value predispositions, reflective integration
in the form of elaborative processing had a significant negative influence on perceived risksversus-benefits. Guided by the differential gains model, this dissertation found that the
influence of science media use on both perceived risks-versus-benefits and support for federal
funding of nanotechnology was moderated by elaborative processing. Based on the cognitive
mediation model, this dissertation also found that the mass media directly and indirectly exert
its influence on public attitudes toward nanotechnology through an informational route and a
heuristic route. Notably, people who engage in reflective integration tended to build a more
134
sophisticated knowledge structure about the scientific world, which in turn, shape their
attitudes toward emerging technologies. Therefore, the findings of the dissertation indicate that
it would be premature to invalidate the role of scientific knowledge in the area of public
understanding of science at the moment. Both the main effects of value predispositions and
cognitive processing variables suggest that the cognitive miser model and the scientific literacy
model are two complementary processes that individuals use to form opinions about
nanotechnology.
Comparison of the expert and public opinion of nanotechnology yielded several
interesting findings. The public judged nanotechnology as having more risks than benefits and
were less supportive for federal funding of the emerging technology than were the experts.
Equipped with their professional training and experience, experts used relatively less heuristic
cues, such as religious guidance, to make risks-versus-benefits judgment of nanotechnology
than did the public. Similarly, the experts draw on significantly less heuristic cues in the form of
value predispositions and science media frames to make decision about funding support for
nanotechnology than did the public. These findings indicate that the experts are in a position to
independently assess risks and benefits, and indicate that the experts and the public use
different considerations to make judgments about risks and benefits of the emerging technology.
9.2. Major Theoretical, Conceptual, and Practical Contributions
By focusing on the intersection between the mass media, public opinion, and expert
opinion of nanotechnology, this dissertation has made many important theoretical, conceptual,
and practical contributions to the field of science communication and the area of science
policymaking.
135
One key theoretical contribution that this dissertation has made pertains to the finding
that the “scientific literacy model” and the “cognitive miser model” are parallel, simultaneous
processes that individuals use to make judgments about emerging technologies. Instead of
regarding them as two distinct processes that work in isolation of each other, this dissertation
demonstrates that both heuristic cues and cognitive thinking complement each other and were
used by individuals when forming decisions about nanotechnology. In particular, the
relationships of mass media use and reflective integration with public knowledge about science
and public attitudes toward nanotechnology enable us to draw a nexus between the two
theoretical models. Science communication scholars should take into consideration both
heuristic and cognitive factors in future to develop a fuller understanding of how the public
form opinions about controversial science and emerging technologies.
Another major theoretical contribution of this dissertation is that it bridges the
disconnection between the differential gains model and the cognitive mediation model that
were originally developed from the field of political communication. The differential gains
model posits that reflective integration can moderate the influence of mass media use on
cognitive, attitudinal, and behavioral outcomes. The processes of intrapersonal and
interpersonal reflections underlie the differential gains model. The cognitive mediation model,
on the other hand, advocates a different process in which reflective integration can mediate the
impact of mass media use on cognitive, attitudinal, and behavioral outcomes. Scholars have
generated two separate lines of empirical research based on these two theoretical models
(Eveland, 2001, 2002, 2004; Eveland et al., 2003; Eveland & Thomson, 2006; Scheufele, 2001,
2002), without realizing that a nexus could be drawn between them. By testing the competing
hypotheses in a single study, this dissertation connects the two models to develop a more
complete and sophisticated model that could explain and predict how public would form
136
opinions about nanotechnology. This underscores the importance of considering both the
moderating and mediating processes of science media use and reflective integration on public
attitudes toward emerging technologies and in other areas of communication research in future
studies. Furthermore, while traditional political communication approaches often focus
narrowly on the cognitive and behavioral outcomes of elaboration, the results suggest that more
immediate outcomes, such as knowledge, matter mostly as precursors to variables like trust and
risk-benefit judgments, which in turn shape policy judgments.
Moreover, this dissertation not only develops a communication theory-centered
approach to understanding public attitudes toward science and technologies, but also
contributes theoretically to the differential gains model and the cognitive mediation model by
testing the communication models in a scientific, rather than political, context. Future studies
should continue to test the competing hypotheses of the two models in other non-political
contexts to strengthen their generalizability and validity.
Very often, researchers examine communication effects using issues that are “chronically
accessible,” that is, issues that have been heavily discussed in public discourse and that are so
prominent in the audiences’ mind to the extend that it becomes a challenge to detect significant
attitudinal changes that are due to communication effects (Iyengar & Kinder, 1987). Contrary to
these chronically accessible issues, public levels of awareness and knowledge about
nanotechnology is low at this point (Peter D. Hart Research Associates, 2007) and a majority of
citizens have hardly any predetermined ideas or firmly held attitudes about risks and benefits
of the emerging technology (Scheufele & Lewenstein, 2005). Therefore, the choice of
nanotechnology as an issue in this dissertation provided an ideal setting to explore the
processes by which audiences gather information and understand this information through
interpersonal or intrapersonal channels.
137
Conceptually, this dissertation contributes to extant science communication research by
demonstrating that deference to scientific authority and trust in scientists are two
fundamentally different concepts that should be separately examined in future studies.
Deference to scientific authority is a long-term socialized trait that directs public attitudes
toward a wide range of technical controversies, whereas trust refers to public willingness to rely
on the endorsements of experts to handle risks associated with emerging technologies. While
deference to scientific authority is a trait-like value predisposition that is general and applicable
to a wide range of scientific controversies, trust is a state-like quality that is specific to a
particular science or technology. This suggests that future studies should use different
operational definitions of the two distinct concepts when examining their impacts on public
attitudes toward emerging technologies in future.
The comparison between public and expert opinions about nanotechnology in this
dissertation have also contributed to the weight of evidence in extant risk communication
literature that experts were indeed more objective in their judgments than were the public. This
dissertation also used a more rigorous methodology in terms of sampling procedure, in terms of
sample size, and in terms of statistical analyses, than did other previous studies. Science
communication researchers could leverage on the methodological and analytical strengths of
this dissertation for similar comparative studies in future.
Besides theoretical and conceptual contributions, this dissertation also offers many
important practical implications to policymakers, scientists, and science communication
practitioners. The results from this dissertation have demonstrated that the mass media could
shape public attitudes toward nanotechnology through the heuristic and/or informational
routes. Media coverage of nanotechnology may be positive right now because it is at the early
stage of the issue-attention cycle. However, once the issue progress and is picked up by
138
mainstream media and local news, various interest groups will struggle to frame the issue and
tailor the message to suit their own interests. For instance, opponents of nanotechnology have
already begun to frame it as the “asbestos of tomorrow” in the mass media. Policymakers,
scientists, and science communication practitioners should be aware of this and exert their
influence by framing their messages in favorable terms, while staying truthful to scientific
uncertainty. For example, when scientists are speaking to a group of businessmen, they should
emphasize the economic relevance of science by pointing out that expanded government
funding would make the U.S. more economically competitive. As pointed out by Nisbet and
Scheufele (2007), “In political coverage, at the opinion pages, in television advertising, and at
the cable news shows, if scientists don’t evolve in their strategies, they will essentially be
waving a white flag, surrendering their important role as communicators.” (p. 41) At the same
time, public officials could use the mass media as an avenue, such as running campaigns and
sponsoring science programs on PBS channels, to offer accurate and up-to-date information
about nanotechnology to the public.
Next, this dissertation also shows that talking to others about scientific issues mediates
the effect of the mass media on public attitudes toward nanotechnology. This finding suggests
that it may be worthwhile for policymakers to invest in large-scale public dialogue initiatives
such as town hall meetings, deliberative forums, and nano cafes. Deliberative forums generate
conversations among highly engaged citizens and activists, and allow scientific organizations
and government officials to tap concerns early and integrate them into policy, and prevent the
types of controversies that had arisen for scientific issues such as genetically modified
technology and embryonic stem cell research. In fact, science communication practitioners
could leverage Internet resources to create online dialogue sessions for the public to engage in
discussions about emerging science and technologies. Online discussion forums have the
139
additional advantages of reaching out to citizens from diverse geographical areas to participate
and alleviating some of the dysfunctional social-psychological barriers so that the public are
willing to express their opinions (e.g., Ho & McLeod, 2008). Given that the results of this
dissertation show a gap in perceived risks-versus-benefits and level of support for federal
funding of nanotechnology between the experts and the general public that is indicative of a
communication deficit, it should be worthwhile to conduct consensus conferences for the public
to have dialogues with scientists and elites to narrow the gaps in attitudes between the two
groups. As pointed out in a 2003 speech by George M. Whitesides at Harvard University:
“Intimately related to education is public understanding, which is currently
confused on the subject of nanotechnology. This confusion means that there is a
real obligation on the part of the scientific community to try to help ‘unconfuse’
people. So long as everyone is confused, imagined risks will stand in the way of
real progress.”
The findings of this dissertation therefore highlight the importance of dialogues between
experts and the lay public, such that accurate and up-to-date information are conveyed to the
latter.
Moreover, given that there are different groups that have different opinions about
nanotechnology (such as the highly religious public), science communication practitioners
should adopt the target segmentation strategy, in which communication messages are tailored
to fit with publics from different social backgrounds for maximum effect. For example, to reach
out to the religious public, scientific institutions should strengthen partnerships with churches
by arranging scientists to speak at churches on topics related to nanotechnology and inviting
140
churches to visit research institutions and ask religious leaders to address scientists on issues of
concern. At the same time, policymakers and the relevant scientific institutions should find
ways to promote and instill trust in scientists and deference to scientific authority among the
public (e.g., arranging eminent scientists to conduct seminars for high-school students) so as to
counter the opposing force that religious guidance could potentially play in shaping opinion
about nanotechnology.
In a similar vein, trust in nano-scientists both in the academia and the industry is crucial
to public support for nanotechnology. Therefore, government regulatory bodies should ensure
that the necessary guidelines are in place (e.g., guidelines to manage toxicity related to
nanotechnology and health standards for creating commercial products) so that public
confidence and trust is maintained.
9.3. Limitations and Directions for Future Research
9.3.1. Study One
There are several limitations in Study 1 that could be overcome in future research. First,
the cross-sectional data used in Study 1 limits the extent to which I can lay claims about the
causal direction of the direct, indirect, and additive relationships in the structural equation
model and the regression models. This refers to both the relationships among predictors of
support for funding and the relationships between antecedents of support for funding and
support for funding itself. Future studies could establish time-order using panel data in order to
make more rigorous causal inferences.
Second, some concepts were operationalized with single-item measures in Study 1.
Although this meant that we could not control for unreliability in some of our measures, it is
reasonable to assume that any potential random error in these single-item measures would
141
weaken the relationships found in our regression models. In other words, if I had been able to
use multi-item measures, I would have been likely to find stronger effects for the relationships
in our regression and structural equation models.
Just like many extant studies in the area of science communication, Study 1 at best
managed to tap respondents’ level of factual scientific knowledge as opposed to other kinds of
scientific knowledge. Moreover, this study claimed that a more sophisticated knowledge of the
scientific world exists in the public minds. Yet, the assertion of a more sophisticated knowledge
is a claim that has not been directly tested in this study. Therefore, it would be worthwhile for
future studies to explore more dimensions of scientific knowledge and find ways to
operationalize them. In addition, future studies could establish the discriminant validity of
those different dimensions of scientific knowledge and examine their various impacts on public
attitudes toward nanotechnology.
Similarly, Study 1 claimed that positive framing in the news media drives audiences to
perceive more benefits over risks of nanotechnology and to be more supportive of federal
funding of the emerging technology. Although Study 1 found an impact of the mass media on
attitude, this argument was not explicitly tested in the study. Therefore, future research may
use experimental methods to tease out the interactive effects of specific news frames and
cognitive processing on individuals’ perceived risks of nanotechnology in order to validate the
claims made here.
Despite these shortcomings, the findings of Study 1 also inform future research agendas.
First, using ordinary regression analysis and structural equation modeling to test for
moderating and mediating relationships respectively are common statistical approaches used in
most communication studies. More recently, however, a more advanced statistical approach of
using structural equation model to test for either “moderated mediation” or “mediated
142
moderation” has been introduced in social science research (e.g., Preacher, Rucker, & Hayes,
2007) and has been applied in some recent communication studies (e.g., Slater, Hayes, & Ford,
2007). That is, this approach models hypotheses combining mediation and moderation. Future
research may use this integrated approach to test for the potential “moderated mediation” role
of reflective integration on public attitudes toward nanotechnology.
Next, it may be worthwhile for future studies to examine the influence of mass media
and reflective integration on other science and technologies that are at different stages of the
issue attention cycle, so as to ensure that the significant effects found in this study are not
unique to nanotechnology. For example, it may be worthwhile to examine public risk
judgments and acceptance of embryonic stem cell research and biotechnology in which the risk
aspects had been made salient in the mass media. Second, future studies may also examine
reflective integration along with other cognitive information processes such as selective
scanning to provide a more complete understanding of how people form attitudes toward
emerging technologies. Finally, future research may move beyond perceived risks-versusbenefits and support for funding of nanotechnology to examine factors that motivate public to
actively participate in issues related to nanotechnology specifically, and science more generally.
9.3.2. Study Two
Likewise, several issues in Study 2 should be addressed and overcome in future research.
Just like Study 1, Study 2 utilizes cross-sectional data, which captures only a snapshot of the
dynamic opinion formation processes. Future research should use longitudinal panel data to
track changes within individuals among the public and the expert samples over time.
Longitudinal studies are especially important for emerging technologies such as
nanotechnology when mass media coverage of the issue changes according to issue attention
cycle.
143
Second, Study 2 examined experts and public attitudinal differences within the United
States, which may limit the generalizability of the results, at least geographically. Future
research should therefore examine lay-expert differences cross-culturally, comparing across
North America, Europe, and Asia.
Third, several of the non-significant differences found in Study 2 may be a function of
the considerably smaller sample size of the experts relative to the sample size of the public.
Despite this, the size of our expert sample (N=363) was a marked improvement from previous
studies that used sample sizes that were smaller than 50 respondents when examining
layperson and experts differences (e.g., Siegrist et al., 2007).
Furthermore, future research should develop a more comprehensive model that
identifies the main factors or considerations that influence experts’ perceived risks and benefits
of the emerging technology, especially when we know that the value predispositions examined
in Study 2 are not the crucial determinants among the experts. In other words, there may be
other unidentified value predispositions that could be examined in future studies. In addition,
experts’ differences in attitudes toward nanotechnology may be a function of practical factors
such as their area of specializations and the nature of their workplace. For example, scientists
who work in nano-related toxicology research may be more concerned about the health and
environmental risks of the emerging technology than do scientists who conduct nano-related
research in other sub-domains. Therefore, the nature of their research may account for some of
the unexplained variances in experts’ perceived risks-versus-benefits. Moreover, scientists who
work in public universities often face strong competition among themselves for government
funds to conduct research in nanotechnology. As such, scientists who work in public
universities may be more concerned about federal funding support for nanotechnology than do
scientists who work in private universities. Besides the nature of their working environment, a
144
recent study has shown that scientists’ leadership and managerial roles, and their research
productivity in terms of the quantity of publications were associated with their frequency of
contact with journalists (Peters et al., 2008). Hence, scientists’ management roles and research
productivity are practical factors that may influence their attitudes toward nanotechnology. Of
course, these are speculations that should be verified in future studies.
Nevertheless, by being the first to date to examine risks-versus-benefits attitudinal
differences and support for federal funding differences between experts and the public about
nanotechnology in the United States, Study 2 no doubt provides us with a more comprehensive
understanding of opinion formation of the emerging technology.
9.4. Conclusion
In closing, I would like to underscore the importance of a theory-driven approach for
developing a sophisticated model that could provide predictions and explanations on how the
public form opinions about emerging science and technologies. By combining the moderating
and mediating mechanisms in the differential gains model and the cognitive mediation model,
this dissertation took a leap from previous studies in terms of introducing a theory-driven
approach to understanding public opinion formation about emerging technologies. Science
communication researchers should continue to extend these theoretical models to scientific
issues so that policymakers and communication practitioners could perform a more
theoretically informed evaluation of how public judgments are formed and design more
effective communication strategies to offer the latest information about science and technologies
to the public.
Last but not least, I would like to stress the importance in the comparisons of experts
and lay public attitudes toward emerging science and technologies and reiterate that
145
researchers should continue to identify other sets of considerations that were not examined in
this dissertation. This is important for informing policymakers, scientists, and practitioners on
whether dialogues between the lay public and experts and other public outreach efforts are
necessary.
146
BIBLIOGRAPHY
Allison, P. D. (2001). Missing data. Thousand Oaks, CA: Sage.
Allum, N., Sturgis, P., Tabourazi, D., & Brunton-Smith, I. (2008). Science knowledge and
attitudes across cultures: A meta-analysis. Public Understanding of Science, 17(1), 35-54.
Arbuckle, J. L., & Wothke, W. (1999). Amos 4.0 User's Guide. Chicago: Small Waters.
Bainbridge, W. S. (2002). Public attitudes toward nanotechnology. Journal of Nanoparticle
Research, 4, 561-570.
Bainbridge, W. S. (2003). Religious opposition to cloning. Journal of Evolution and Technology, 13.
Retrieved August 30, 2008 from http://www.jetpress.org/volume13/bainbridge.html
Barke, R. P., Jenkins-Smith, H., & Slovic, P. (1997). Risk perceptions of men and women
scientists. Social Science Quarterly, 78(1), 167-176.
Barke, R. P., & Jenkins-Smith, H. C. (1993). Politics and scientific expertise: Scientists, risk
perception, and nuclear waste policy. Risk Analysis, 13, 425-439.
Beaudoin, C. E., & Thorson, E. (2004). Testing the cognitive mediation model - The roles of news
reliance and three gratifications sought. Communication Research, 31(4), 446-471.
Belenky, M. F., Clinchy, B. M., Goldberger, N. R., & Tarule, J. M. (1986). Women's ways of
knowing: The development of self, voice, and mind. New York: Basic Books.
Bentler, P. M., & Bonnet, D. G. (1980). Significance test and goodness of fit in the analysis of
covariance structures. Psychological Bulletin, 88, 588-606.
Bergstein, B. (May 20, 2008). Carbon nanotubes mimic asbestos in early study. The Associated
Press.
147
Bimber, B., & Guston, D. (1995). Politics by the same means: Government and science in the
United States. In S. Jasanoff et al. (Ed.), The handbook of science and technology studies (pp.
554-571). Thousands Oaks, CA: Sage Publications.
Blumler, J. G., & Katz, E. (1974). The Uses of Mass Communication. Newbury Park, CA: Sage.
Bodmer, W. (1985). The public understanding of science. London: Royal Society.
Bollen, K. A. (1989). Structural equations with latent variables. New York: John Wiley & Sons.
Bord, R., & O'Connor, R. (1997). The gender gap in environmental attitudes: The case of
perceived vulnerability to risk. Social Science Quarterly, 78(4), 830-840.
Brody, C. J. (1984). Differences by sex in support for nuclear power. Social Forces, 63, 209-228.
Brooke, J. (1998). Science and religion: Lessons from history? Science, 282(5396), 1985-1986.
Brossard*, D., Kim*, E. (2007). Pondering media messages, talking to others and
learning: Communication processes and the production of scientific knowledge. Paper
presented to the Communication Theory & Methodology Division of the 2007 AEJMC
Convention, Washington DC, August 2007. (* equal contribution).
Brossard, D., Lewenstein, B., & Bonney, R. (2005). Scientific knowledge and attitude change: The
impact of a citizen science project. International Journal of Science Education, 27(9), 10991121.
Brossard, D., & Nisbet, M. C. (2007). Deference to scientific authority among a low information
public: Understanding U.S. opinion on agricultural biotechnology. International Journal of
Public Opinion Research, 19(1), 24-52.
Brossard, D., Scheufele, D. A., Kim, E., & Lewenstein, B. V. (in press). Religiosity as a perceptual
filter: Examining processes of opinion formation about nanotechnology. Public
Understanding of Science.
148
Brossard, D., & Shanahan, J. (2003). Do citizens want to have their say? Media, agricultural
biotechnology, and authoritarian views of democratic processes in science. Mass
Communication and Society, 3, 291-312.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J.
S. Long (Eds.), Testing structural equation models. Thousand Oaks, CA: Sage.
Caplan, A. L. (2003). Is better best? Scientific American, 68-73.
Castellini, O. M., Walejko, G. K., Holladay, C. E., Theim, T. J., Zenner, G. M., & Crone, W. C.
(2007). Nanotechnology and the public: Effectively communicating nanoscale science
and engineering concepts. Journal of Nanoparticle Research, 9, 183-189.
Chaffee, S., & Frank, S. (1996). How Americans get political information: Print versus broadcast
news. Annals of the American Academy of Political and Social Science, 546, 48-58.
Chaffee, S. H., & Kanihan, S. F. (1997). Learning about politics from the mass media. Political
Communication, 14(4), 421-430.
Chaffee, S. H., & Schleuder, J. (1986). Measurement and effects of attention to media news.
Human Communication Research, 13(1), 76-107.
Chang, K. (May 21, 2008). In study, researchers find nanotubes may pose health risks similar to
asbestos. The New York Times, p.A22.
Clark, F., & Illman, D. L. (2006). A longitudinal study of the New York Times Science Times
Section. Science Communication, 27, 496.
Cobb, M. D. (2005). Framing effects on public opinion about nanotechnology. Science
Communication, 27(2), 221-239.
Cobb, M. D., & Macoubrie, J. (2004). Public perceptions about nanotechnology: Risks, benefits,
and trust. Journal of Nanoparticle Research: An Interdisciplinary Forum for Nanoscale Science
and Technology, 6(4), 395-405.
149
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation
analysis for the behavioral sciences (3rd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
Cole, G. A., & Withey, S. B. (1981). Perspectives on risk perceptions. Risk Analysis, 1, 143-163.
Coleman, C. L. (1993). The influence of mass media and interpersonal communication on
societal and personal risk judgments. Communication Research, 20(4), 611-628.
Covello, V., & Sandman, P. (2001). Risk communication: Evolution and revolution. In A.
Wolbarst (Ed.), Solutions to an Environment in Peril (pp. 164-178). Baltimore, MD: John
Hopkins University Press.
Currall, S. C., King, E. B., Lane, N., Madera, J., & Turner, S. (2006). What drives public
acceptance of nanotechnology? Nature Nanotechnology, 1(153-155).
Davidson, D. J., & Freudenburg, W. R. (1996). Gender and environmental risk concerns: A
review and analysis of available research. Environment and Behavior, 28, 302-309.
Diamond, E. (1994). Behind the Times: Inside the New York Times. New York: Villard.
Downs, A. (1972). Up and down with ecology -- The “issue-attention cycle”. The Public Interest,
28, 38-51.
Driver, R. (1995). Constructivist approaches in science teaching. In L. P. Steffe & J. Gale (Eds.),
Constructivism in education (pp. 385-400). Hillsdale, NJ: Lawrence Erlbaum Associates.
Driver, R., Asoko, H., Leach, J., Mortimer, E., & Scott, P. (1994). Constructing scientific
knowledge in the classroom. Educational Researcher, 23(7), 5-12.
Dunwoody, S., & Neuwirth, K. (1991). Coming to terms with the impact of communication on
scientific and technological risks judgments. In L. Wilkins & P. Patterson (Eds.), Risky
Business: Communicating Issues of Science, Risks, and Public Policy (pp. 11-30). New York:
Greenwood.
150
Durant, J., Evans, G., & Thomas, G. (1992). Public understanding of science in Britain: The role
of medicine in the popular representation of science. Public Understanding of Science, 1,
161-182.
Earle, T. C., & Cvetkovich, G. T. (1995). Social Trust: Toward a Cosmopolitan Society. London:
Praeger.
Entman, R. M. (1991). Framing U.S. coverage of international news: Contrasts in narratives of
the KAL and Iran air. Journal of Communication, 41(4), 6.
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of
Communication, 43(4), 51.
European Commission. (2005). Some figures about nanotechnology R&D in Europe and beyond.
Retrieved on August 20, 2008 from http://cordis.europa.eu.int/nanotechnology
Eveland, W. P. (2001). The cognitive mediation model of learning from the news: Evidence from
non-election, off-year election, and presidential election contexts. Communication
Research, 28, 571-601.
Eveland, W. P. (2002). News information processing as mediator of the relationship between
motivations and political knowledge. Journalism and Mass Communication Quarterly, 79(1),
26-40.
Eveland, W. P. (2004). The effect of political discussion in producing informed citizens: The
roles of information, motivation, and elaboration. Political Communication, 21(2), 177-193.
Eveland, W. P., Shah, D. V., & Kwak, N. (2003). Assessing causality in the cognitive mediation
model: A panel study of motivations, information processing, and learning during
Campaign 2000. Communication Research, 30(4), 359-386.
151
Eveland, W. P., & Thomson, T. (2006). Is it talking, thinking, or both? A lagged dependent
variable model of discussion effects on political knowledge. Journal of Communication,
56(3), 523-542.
Fischhoff, B., Slovic, P., Lichtenstein, S., Read, S., & Combs, B. (1978). How safe is sage enough?
A psychometric study of attitudes toward technological risks and benefits. Policy Sciences,
9, 127-152.
Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York: McGraw-Hill.
Fleming, K., Thorson, E., & Zhang, Y. Y. (2006). Going beyond exposure to local news media:
An information-processing examination of public perceptions of food safety. Journal of
Health Communication, 11(8), 789-806.
Flynn, J., Slovic, P., & Mertz, C. K. (1993). Decidedly different: Expert and public views of risks
from a radioactive waste repository. Risk Analysis, 13(6), 643-648.
Fredin, E. S., Kosicki, G. M., & Becker, L. B. (1996). Cognitive strategies for media use during a
presidential campaign. Political Communication, 13, 23-42.
Freudenburg, W. (1992). Heurisics, biases, and the not-so-general publics: Expertise and error in
the assessment of risks. In S. Krimsky & D. Golding (Eds.), Social Theories of Risk (pp. 229250). Westport, CT: Praeger.
Freudenburg, W. (1993). Risk and recreancy: Weber, the division of labor, and rationality of risk
perceptions. Social Forces, 71, 900-932.
Frewer, L. J., Howard, C., & Shepherd, R. (1998). Understanding attitudes to technology. Journal
of Risk Research, 1, 221-235.
Friedman, S. M., & Egolf, B. P. (2005). Nanotechnology: Risks and the media. IEEE Technology
and Society, 5-11.
Gamson, W. A. (1992). Talking politics. Cambridge, UK: Cambridge University Press.
152
Gaskell, G., Allum, N., Bauer, M., Durant, J., Allansdottir, A., Bonfadelli, H., et al. (2000).
Biotechnology and the European public. Nature Biotechnology, 18(9), 935-938.
Gaskell, G., Einsiedel, E., Hallman, W., Priest, S. H., Jackson, J., & Olsthoorn, J. (2005).
Communication - Social values and the governance of science. Science, 310(5756), 19081909.
Gaskell, G., Eyck, T. T., Jackson, J., & Veltri, G. (2004). Public attitudes to nanotechnology in
Europe and the United States. Nature Materials, 3, 496.
Gaskell, G., Eyck, T. T., Jackson, J., & Veltri, G. (2005). Imaging nanotechnology: Cultural
support for technological innovation in Europe and the United States. Public
Understanding of Science, 14, 81-90.
Gelb, A. (2003). City room. New York: Marion Wood.
Giddens, A. (1991). Modernity and Self-identity. Cambridge: Polity Press.
Goffman, E. (1974). Frame analysis: An essay on the organization of experience. New York: Harper &
Row.
Griffin, R. J., & Dunwoody, S. (2000). The relation of communication to risk judgment and
preventive behavior related to lead in tap water. Health Communication, 12(1), 81-107.
Gutteling, J., & Wiegman, O. (1993). Gender-specific reactions to environmental hazards in the
Netherlands. Sex Roles, 28(7/8), 433-446.
Gutteling, J. M., & Kuttschreuter, M. (1999). The millennium bug controversy in the
Netherlands? Experts' views versus public perception. In L. H. J. Goossens (Ed.),
Proceedings of the ninth annual conference of Risk Analysis: Facing the millennium (pp. 489493). Delft, The Netherlands: Delft University Press.
153
Gutteling, J. M., & Kuttschreuter, M. (2002). The role of expertise in risk communication:
Laypeople's and experts' perception of the millennium bug in the Netherlands. Journal of
Risk Research, 5, 35-47.
Hardy, B. W., & Scheufele, D. A. (2005). Examining differential gains from Internet use:
Comparing the moderating role of talk and online interactions. Journal of Communication,
55, 71-84.
Ho, S. S., Brossard, D., & Scheufele, D. A. (2007). The polls-trends: Public reactions to global
health threats and infectious diseases Public Opinion Quarterly, 71(4), 671-692.
Ho, S. S., Brossard, D., & Scheufele, D. A. (2008). Effects of value predispositions, mass media
use, and knowledge on public attitudes toward embryonic stem cell research.
International Journal of Public Opinion Research, 20(2), 171-192.
Ho, S. S., & McLeod, D. M. (2008). Social-psychological influences on opinion expression in faceto-face and computer-mediated communication. Communication Research, 35(2), 190-207.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55.
Irwin, A. (2001). Constructing the scientific citizen: science and democracy in the biosciences.
Public Understanding of Science, 10(1), 1-18.
Irwin, A., & Wynne, B. (Eds.). (1996). Misunderstanding Science? The Public Reconstruction of
Science and Technology. Cambridge: Cambridge University Press.
Iyengar, S. (1991). Who is responsible? Chicago, IL: The University of Chicago Press.
Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and American opinion. Chicago,
IL: University of Chicago Press.
Johnson, B. B. (1993). Advancing understanding of knowledge's role in lay risk perception. Risk
- Issues in Health and Safely, 189.
154
Joreskog, K. G. (1993). Testing structural equation models. In K. A. Bollen & J. S. Long (Eds.),
Testing structural equation models (pp. 294-316). Newbury Park, CA: Sage.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.
Econometrica, 47(2), 263-291.
Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39, 341350.
Kaplan, D. (2009). Structural equation modeling: Foundations and extensions. Thousand Oaks, CA:
Sage Publications.
Kelloway, E. K. (1998). Using LISREL for structural equation modeling: A researcher's guide.
Thousand Oaks, CA: Sage.
Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: The Guilford
Press.
Kosicki, G. M., & McLeod, J. M. (1990). Learning from political news: Effects of media images
and information-processing strategies. In S. Kraus (Ed.), Mass Communication and Political
Information Processing (pp. 69-83). Hillsdale, NJ: Erlbaum.
Kraus, N., Malmfors, T., & Slovic, P. (1992). Intuitive toxicology: Expert and lay judgments of
chemical risks. Risk Analysis, 12(2), 215-232.
Lazo, J. K., Kinnell, J. C., & Fisher, A. (2000). Expert and layperson perceptions of ecosystem risk.
Risk Analysis, 20(2), 179-194.
Lee, C., & Scheufele, D. A. (2006). The influence of knowledge and deference toward scientific
authority: A media effects model for public attitudes toward nanotechnology. Journalism
and Mass Communication Quarterly, 83(4), 819-834.
155
Lee, C., Scheufele, D. A., & Lewenstein, B. V. (2005). Public attitudes toward emerging
technologies: Examining the interactive effects of cognitions and affect on public
attitudes toward nanotechnology. Science Communication, 27(2), 240-267.
Levine, J. M., & Russo, E. (1995). Impact of anticipated interaction on information acquisition.
Social Cognition, 13, 293-317.
Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data (2nd ed.). New York:
Wiley.
Luhmann, N. (1979). Trust and Power. Chichester: Wiley.
Lux Research. (2008) Overhyped technology starts to reach potential: Nanotech to impact $3.1 trillion
in manufactured goods in 2015. Retrieved on August 20, 2008 from
http://www.luxresearchinc.com/press/RELEASE_Nano-SMR_7_22_08.pdf
MacCallum, R. C. (1995). Model specification: Procedures, strategies, and related issues. In R. H.
Hoyle (Ed.), Structural equation modeling (pp. 16-36). Thousand Oaks, CA: Sage.
Maurer, M., & Reinemann, C. (2006). Learning versus knowing - Effects of misinformation in
televised debates. Communication Research, 33(6), 489-506.
McComas, K. A. (2006). Defining moments in risk communication research: 1996-2005. Journal of
Health Communication, 11(1), 75-91.
McComas, K. A., & Shanahan, J. E. (1999). Telling stories about global climate change:
Measuring the impact of narratives on issue cycles. Communication Research, 26(1), 30-57.
McCombs, M. E., & Shaw, D. L. (1972). Agenda-setting function of mass media. Public Opinion
Quarterly, 36(2), 176-187.
McDaniels, T. L., Axelrod, L. J., Cavanagh, N. S., & Slovic, P. (1997). Perception of ecological risk
to water environments. Risk Analysis, 17, 341-352.
156
McLeod, J. M., Kosicki, G. M., & Pan, Z. (1991). On understanding and misunderstanding media
effects. In J. Curran & M. Gurevitch (Eds.), Mass media and society (pp. 235-266). London:
Edward Arnold.
McLeod, J. M., Scheufele, D. A., Moy, P., Horowitz, E. M., Holbert, R. L., Zhang, W., et al. (1999).
Understanding deliberation: The effects of discussion networks on participation in a
public forum. Communication Research, 26, 623-654.
Miller, J. D. (1998). The measurement of civic scientific literacy. Public Understanding of Science,
7(3), 203-223.
Miller, J. D. (2004). Public understanding of, and attitudes toward, scientific reserach: What we
know and what we need to know. Public Understanding of Science, 13(3), 273-294.
Miller, J. D., & Kimmel, L. (2001). Biomedical communications: Purposes, audiences, and strategies.
New York: John Wiley.
Miller, J. D., Pardo, R., & Niwa, F. (1997). Public perceptions of science and technology: A comparative
study of the European Union, the United States, Japan, and Canada. Chicago: Chicago
Academy of Sciences.
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, 28(5), 602626.
Morton, T. A., & Duck, J. M. (2006). Enlisting the influence of others: Alternative strategies for
persuasive media campaigns. Journal of Applied Social Psychology, 36(2), 269-296.
National Science and Technology Council. (2000) National nanotechnology initiative:
Leading to the next industrial revolution. A report by the Interagency Working Group on
Nanoscience, Engineering and Technology, Washington, D. C.
157
National Science and Technology Council. (2007). The National Nanotechnology Initiative Strategic
Plan. A report by the Interagency Working Group on Nanoscale Science, Engineering,
and Technology, Washington, D.C.
Nelkin, D. (1979). Controversy: Politics of Technical Decisions. Newbury Park, CA: Sage.
Nelkin, D., & Lindee, S. M. (1995). The DNA mystique: The gene as a cultural icon. New York: W. H.
Freeman.
Nisbet, E. C., Nisbet, M. C., Scheufele, D. A., & Shanahan, J. E. (2004). Public diplomacy,
television news, and Muslim opinion. The Harvard International Journal of Press/Politics,
9(2), 11-37.
Nisbet, M. C. (2005). The competition for worldviews: Values, information, and public support
for stem cell research. International Journal of Public Opinion Research, 17(1), 90-112.
Nisbet, M. C., Brossard, D., & Kroepsch, A. (2003). Framing science: The stem cell controversy in
the age of press/politics. Harvard International Journal of Press/Politics, 8(2), 36-70.
Nisbet, M. C., & Huge, M. (2006). Attention cycles and frames in the plant biotechnology debate
- Managing power and participation through the press/policy connection. Harvard
International Journal of Press/Politics, 11(2), 3-40.
Nisbet, M. C., & Lewenstein, B. V. (2002). Biotechnology and the American media -- the policy
process and the elite press, 1970 to 1999. Science Communication, 23(4), 359-391.
Nisbet, M. C., & Scheufele, D. A. (2004). Political talk as a catalyst for online citizenship.
Journalism and Mass Communication Quarterly, 81(4), 877-896.
Nisbet, M. C., & Scheufele, D. A. (2007). The future of public engagement. The Scientist, 21(10),
39-44.
158
Nisbet, M. C., Scheufele, D. A., Shanahan, J., Moy, P., Brossard, D., & Lewenstein, B. V. (2002).
Knowledge, reservations, or promise? A media effects model for public perceptions of
science and technology. Communication Research, 29(5), 584-608.
Olofsson, A., Ohman, S., & Rashid, S. (2006). Attitudes to gene technology: The significance of
trust in institutions. European Societies, 8(4), 601-624.
Parrott, R., Silk, K., Krieger, J. R., Harris, T., & Condit, C. (2004). Behavioral health outcomes
associated with religious faith and media exposure about human genetics. Health
Communication, 16(1), 29-45.
Peter D. Hart Research Associates (2007). Awareness of and attitudes toward nanotechnology and
federal regulatory agencies. Retrieved October 10, 2007, from
http://www.nanotechproject.org/138/9252007-poll-reveals-public-awareness-ofnanotech-stuck-at-low-level.
Peters, H. P., Brossard, D., de Cheveigne, S., Dunwoody, S., Lallfass, M., Miller, S., et al. (2008).
Science communication: Interactions with the mass media. Science, 321(5886), 204-205.
Pew Internet & American Life Project. (2006). The Internet as a resource for news and information
about science. Retrieved from August 20, 2008, from
http://www.pewinternet.org/pdfs/PIP_Exploratorium_Science.pdf
Poland, C. A., Duffin, R., Kinloch, I., Maynard, A., Wallace, W. A. H., Seaton, A., et al. (2008).
Carbon nanotubes introduced into the abdominal cavity of mice show asbestos-like
pathogenicity in a pilot study. Nature Nanotechnology, 3, 423-428.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation
hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42(1),
185-227.
159
President’s Council of Advisors on Science and Technology (PCAST). (2005). The National
Nanotechnology Initiative at five years: Assessment and recommendations of the National
Nanotechnology Advisory Panel. Retrieved August 20, 2008, from
http://www.nano.gov/FINAL_PCAST_NANO_REPORT.pdf
Priest, S. H. (2000). U.S. public opinion divided over biotechnology? Nature Biotechnology, 18(9),
939-942.
Priest, S. H. (2001). Misplaced faith - Communication variables as predictors of encouragement
for biotechnology development. Science Communication, 23(2), 97-110.
Priest, S. H., Bonfadelli, H., & Rusanen, M. (2003). The "trust gap" hypothesis: Predicting
support for biotechnology across national cultures as a function of trust in actors. Risk
Analysis, 23(4), 751-766.
Project on Emerging Nanotechnologies. (2008). An inventory of nanotechnology-based
consumer products currently on the market. Retrieved on August 20, 2008, from
http://www.nanotechproject.org/inventories/consumer/
Raude, J., Fischler, C., Setbon, M., & Flahault, A. (2005). Scientist and public responses to BSErelated risk: A comparative study. Journal of Risk Research, 8(7-8), 663-678.
Renn, O., & Roco, M. C. (2006). Nanotechnology and the need for risk governance. Journal of
Nanoparticle Research, 8, 153-191.
Robinson, J. P., & Levy, M. R. (1986). The main source: Learning from television news. Beverly Hills,
CA: Sage.
Roco, M. C. (2003). Broader societal issues of nanotechnology. Journal of Nanoparticle Research, 5,
181-189.
160
Roco, M. C., & Bainbridge, W. S. (2002). Converging technologies for improving human
performance: Integrating from the nanoscale. Journal of Nanoparticle Research, 4(4), 281295.
Roco, M. C., & Bainbridge, W. S. (2003). Nanotechnology: Societal implications – Maximizing benefit
for humanity. A report of the National Nanotechnology Initiative Workshop, Arlington,
VA.
Rowe, G., & Wright, G. (2001). Differences in expert and lay judgments of risk: Myth or reality?
Risk Analysis, 21(2), 341-356.
Sandel, M. J. (2004). The case against perfection. Atlantic Monthly, 293(3).
Sandman, P. M., Weinstein, N. D., & Klotz, M. L. (1987). Public response to risk from geological
radon. Journal of Communication, 37(3), 93-108.
Savadori, L., Savio, S., Nicotra, E., Rumiati, R., Finucane, M., & Slovic, P. (2004). Expert and
public perception of risk from biotechnology. Risk Analysis, 24(5), 1289-1299.
Savage, I. (1993). Demographic influences on risk perceptions. Risk Analysis, 13, 413-420.
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological
Methods, 7, 147-177.
Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of Communication, 49(1),
103-122.
Scheufele, D. A. (2000). Agenda-Setting, Priming, and Framing Revisited: Another Look at
Cognitive Effects of Political. Mass Communication & Society, 3(2/3), 297.
Scheufele, D. A. (2001). Democracy for some? How political talk both informs and polarizes the
electorate. In R. P. Hart & D. Shaw (Eds.), Communication and U.S. elections: New Agendas
(pp. 19-32). Lanham, MD: Rowman and Littlefield Publishers.
161
Scheufele, D. A. (2002). Examining differential gains from mass media and their implications for
participatory behavior. Communication Research, 29(1), 46-65.
Scheufele, D. A., Corley, E. A., Dunwoody, S., Shih, T., Hillback, E., & Guston, D. (2007).
Scientists worry about some risks more than the public. Nature Nanotechnology, 2(12),
732-734.
Scheufele, D. A., & Lewenstein, B. V. (2005). The public and nanotechnology: How citizens
make sense of emerging technologies. Journal of Nanoparticle Research, 7, 659-667.
Scheufele, D. A., Nisbet, M. C., Brossard, D., & Nisbet, E. C. (2004). Social structure and
citizenship: Examining the impacts of social setting, network heterogeneity, and
informational variables on political participation. Political Communication, 21, 315-338.
Schutz, H., & Wiedemann, P. M. (2008). Framing effects on risk perception of nanotechnology.
Public Understanding of Science, 17(3), 369-379.
SearchTechnology. (2008). VantagePoint. Retrieved on August 20, 2008 from
http://www.thevantagepoint.com/
Sententia, W. (2004). Neuroethical considerations: Cognitive liberty and converging
technologies for improving human cognition. Annals of the New York Academy of Sciences,
1013, 221-228.
Shah, D. V., Cho, J. C., Eveland, W. P., & Kwak, N. (2005). Information and expression in a
digital age - Modeling Internet effects on civic participation. Communication Research,
32(5), 531-565.
Siegrist, M. (1998). Belief in gene technology: The influence of environmental attitudes and
gender. Personality and Individual Differences, 24, 861-866.
Siegrist, M. (2000). The influence of trust and perceptions of risks and benefits on the acceptance
of gene technology. Risk Analysis, 20, 195-203.
162
Siegrist, M., & Cvetkovich, G. (2000). Perception of hazards: The role of social trust and
knowledge. Risk Analysis, 20, 713-719.
Siegrist, M., Cvetkovich, G., & Roth, C. (2000). Salient value similarity, social trust, and
risk/benefit perception. Risk Analysis, 20(3), 353-362.
Siegrist, M., & Gutscher, H. (2006). Flooding risks: A comparison of lay people's perceptions
and expert's assessments in Switzerland. Risk Analysis, 26(4), 971-979.
Siegrist, M., Keller, C., Kastenholz, H., Frey, S., & Wiek, A. (2007). Laypeople's and experts'
perception of nanotechnology hazards. Risk Analysis, 27(1), 59-69.
Sinclair, J., & Irani, T. (2005). Advocacy advertising for biotechnology. Journal of Advertising,
34(3), 59-73.
Sjoberg, L. (2002). Attitudes toward technology and risk: Going beyond what is immediately
given. Policy Sciences, 35(4), 379-400.
Sjoberg, L. (2004). Principles of risk perception applied to gene technology. Embo Reports, 5, S47S51.
Sjoberg, L., & Winroth, E. (1986). Risk, moral value of actions, and mood. Scandinavian Journal of
Psychology, 27(3), 191-208.
Slater, M. D., Hayes, A. F., & Ford, V. L. (2007). Examining the moderating and mediating roles
of news exposure and attention on adolescent judgments of alcohol-related risks.
Communication Research, 34(4), 355-381.
Slater, M. D., & Rasinski, K. A. (2005). Media exposure and attention as mediating variables
influencing social risk judgments. Journal of Communication, 55(4), 810-827.
Slovic, P. (1987). Perception of risk. Science, 36, 280-285.
163
Slovic, P. (1997). Trust, emotion, sex, politics and science. In M. H. Bazerman, D. M. Messick, A.
E. Tenbrunsel & K. A. Wade-Benzoni (Eds.), Environment, ethics and behavior (pp. 277-313).
San Francisco: Lexington Press.
Slovic, P. (1999). Trust, emotion, sex, politics and science: Surveying the risk-assessment
battlefield. Risk Analysis, 19(4), 689-701.
Slovic, P., Fischhoff, B., & Lichtenstein, S. (1985). Characterizing perceived risk. In R. W. Kates,
C. Hohenemser & J. X. Kasperson (Eds.), Perilous progress: Managing the hazards of
technology (pp. 91-125). Boulder, CO: Westview.
Slovic, P., Malmfors, T., Krewski, D., Mertz, C. K., Neil, N., & Bartlett, S. (1995). Intuitive
toxicology II: Expert and lay judgments of chemical risks in Canada. Risk Analysis, 15(6),
661-675.
Sobel, M., & Bohrnstedt, G. W. (1985). Ues of null models in evaluating the fit of covariance
structure models. In N. B. Tuma (Ed.), Sociological methodology 1985 (pp. 152-178). San
Francisco: Jossey-Bass.
Sotirovic, M., & McLeod, J. M. (2001). Values, communication behavior, and political
participation. Political Communication, 18, 273-300.
Sparks, P., Shepherd, R., & Frewer, L. J. (1994). Gene technology, food production, and public
opinion: A UK study. Agricultural and Human Values, 11, 19-28.
Starr, C. (1969). Social benefit versus technological risk. Science, 165(3899), 1232-1238.
Stephens, L. F. (2005). News narratives about Nano S&T in major U.S. and non-U.S. newspapers.
Science Communication, 27(2), 175-199.
Stock, G. (2002). Redesigning Humans, Our Inevitable Genetic Future. New York: Houghton Mifflin.
Sturgis, P., & Allum, N. (2004). Science in society: re-evaluating the deficit model of public
attitudes. Public Understanding of Science, 13(1), 55-74.
164
Sturgis, P., Cooper, H., & Fife-Schaw, C. (2005). Attitudes to biotechnology: Estimating the
opinions of a better-informed public. New Genetics and Society, 24(1), 31-56.
Sztompka, P. (1999). Trust: A sociological theory. Cambridge: Cambridge University Press.
Tanaka, Y. (2004). Major psychological factors affecting acceptance of gene-recombination
technology. Risk Analysis, 24, 1575-1583.
Thomson, M. E., Onkal, D., Avcioglu, A., & Goodwin, P. (2004). Aviation risk perception: A
comparison between experts and novices. Risk Analysis, 24, 1585-1595.
Toulmin, S., & Goodfield, J. (1965). The Discovery of Time. New York: Harper & Row.
Tversky, A., & Kahneman, D. (1982). Availability: A heuristic for judging frequency and
probability. In D. Kahneman, P. Slovic & A. Tversky (Eds.), Judgment Under Uncertainty:
Heuristics and Biases (pp. 163-189). Cambridge: Cambridge University Press.
Tyler, T. R., & Cook, F. L. (1984). The mass media and judgments of risk: Distinguishing impact
on personal and societal level judgments. Journal of Personality and Social Psychology, 47,
693-708.
Valenti, J. M. (2002). Communication challenges for science and religion. Public Understanding of
Science, 11, 57-63.
von Glasersfeld, E. (1995). Radical constructivism: A way of knowing and learning. London: Falmer
Press.
Wilford, J. N. (November, 2003). The birth of Science Times: A surprise, but no accident. The
New York Times, p. 11: F12.
Winer, B. J., Brown, D. R., & Michels, K. M. (1991). Statistical principles in experimental design (3rd
ed.). Boston: McGraw-Hill.
Wright, G., Bolger, F., & Rowe, G. (2002). An empirical test of the relative validity of expert and
lay judgments of risk. Risk Analysis, 22(6), 1107-1122.
165
Wright, G., Pearman, A., & Yardley, K. (2000). Risk perception in the U.K. oil and gas
production industry: Are expert loss-prevention managers' perceptions different from
those of members of the public? Risk Analysis, 20(5), 681-690.
166
Figure 2.1 Media Issue-Attention Cycle
administrative
policy
arena
op
in
ion
Technical
framing
l/
t
r
a
n
s
i
t
i
o
n
tic
a
/ op
ini
cal
Po
liti
overtly
political policy
arena
Dramatic
Framing
overtly
political policy
arena
t
r
a
n
s
i
t
i
o
n
list
cia
spe
Media attention
Type of journalist
ist
ial
ec
Focusing Event
t
r
a
n
s
i
t
i
o
n
sp
t
r
a
n
s
i
t
i
o
n
Technical
Framing
on
Dramatic
Framing
po
li
Pre-problem stage
Cycle
Continues
administrative
policy
arena
Nisbet, M.C. & Huge, M. (2006). Attention cycles and frames in the plant biotechnology
debate: Managing power and participation through the press/policy connection. Harvard
International Journal of Press/Politics, 11(2), 3-40.
167
Figure 2.2 Public attitude towards Nanotechnology Acceptance
Do not agree at all - Agree very much
“Overall, I support federal funding for nanotechnology.”
10
9
8
7
6
5
5.9
5.36
4
3
2
1
2004
2007
Year
168
Figure 2.3 Public Attitudes toward Nanotechnology Risks
Do not agree at all - Agree very much
10
9
8
7
6.43
6
5.87
4.99 5.16
5
2004
5.33
4.89
2007
4
3.29 3.17
3
2
1
0
Loss of
Privacy
An Arms
Race
Loss of
Jobs
Selfreplicating
Robots
Note: 1 = “Do not agree at all” 10 = “Agree very much”
169
Figure 2.4 Public Perceived Nanotechnology Benefits
Do not agree at all - Agree very much
10
9
8
7
7.28
7.15
6.56
6.48
6.5
6.41
6.71
6.79
6
2004
2007
5
4
3
2
1
Better
Treatment of
Diseases
A Cleaner
Environment
Improvement in
Human
Activities
Improved
National
Security
170
Figure 2.5 Public Self-report Level of Awareness about Nanotechnology
“How much have you heard, read or seen about nanotechnology?”
10
Nothing at all - Very much
9
8
7
6
5
4
3.99
3.69
3
2
1
2004
2007
Year
171
Figure 2.6 Public Self-report Level of being Informed about Nanotechnology
“How well informed would you say you are about nanotechnology?”
10
Not informed at all - Very informed
9
8
7
6
5
4
3
3.39
3.18
2
1
0
2004
2007
Year
Note. 1 = “Not informed at all” 10 = “Very informed”
172
Figure 2.7 Public Level of General Scientific Knowledge
Knowledge score
3
2
1.98
1.77
1
0
2004
2007
Year
Note. “Antibiotics kill viruses as well as bacteria.” (False)
“Electrons are smaller than atoms.” (True)
“Ordinary tomatoes do not contain genes, while genetically modified tomatoes
do.” (False)
173
Figure 2.8 Public Level of Knowledge about Nanotechnology
6
Knowledge score
5
4
4.07
3.9
3
2
1
0
2004
2007
Year
Note. “Nanotechnology involves materials that are not visible to the naked eye.” (True)
“US corporations are not using nanotechnology yet to make products sold
today.” (False)
“Experts consider nanotechnology to be the next industrial revolution of the US
economy.” (True)
“A nanometer is a billionth of a meter.” (True)
“Nanotechnology allows scientists to arrange molecules in ways that do not
occur in nature.” (True)
“A nanometer is about the same size as an atom.” (False)
174
Figure 2.9 Public Amount of Attention Paid to Newspaper Content
No attention at all - Very close attention
10
9
8
7
6
5
4.91 4.86
5.08 5.25
4.96
4.48
4
2004
4.58 4.47
3
2
1
0
International
Affairs
National
Government
and Politics
Stories about
Stories Related to
scientific studies in
Science and
new areas of
Technology
research such as
nanotechnology
2007
175
Figure 2.10 Public Amount of Attention Paid to Television Content
10
No attention at all - Very close attention
9
8
7
6.99
6
5
6.04
5.65
4.77
5.08
5.98
4.72
4
5.67
5.68
5.05
2007
4.02
3
2.74
2
1
0
Stories Related
International
National
Affairs
Government to Science and
Technology
and Politics
2004
Specific scientific
developments
such as
nanotechnology
Science fiction
dramas
Science
documentaries
176
Figure 2.11 Public Amount of Attention Paid to Online News Content
No attention at all - Very close attention
10
8
6
4
2
4.51
4.34
4.36
4.12
2004
2007
2.25
2.34
2.38
2.09
0
-2
News about
International
Affairs
News about
National
Government
and Politics
Content Related to Content related to
Science and
specific scientific
Technology
developments such
as nanotechnology
(Note: 0 = “No attention at all” 10 = “Very close attention”)
177
Figure 2.12 Media Coverage of Nanotechnology across 21 Newspapers
300
Number of News Articles
250
200
150
100
50
0
1969
1978
1983
1988
1993
Years
1998
2003
Aug-08
178
Figure 2.13 The New York Times and the Washington Post Coverage of Nanotechnology
180
Number of News Articles
160
140
120
100
80
60
40
20
0
1969
1978
1983
1988
1993
Years
1998
2003
Aug-08
179
Figure 2.14 Emergence of Nanotechnology as an Issue across High, Medium, and Low
Circulation Newspapers
180
Phase 1
Phase 2
Phase 3
Number of News Articles
160
140
120
100
80
60
High Circulation Newspapers
40
Medium Circulation Newspapers
20
Low Circulation Newspapers
0
1969
1978
1983
1988
1993
Years
1998
2003
Aug-08
180
Figure 2.15 Percentage of Risks-Related Nanotechnology Articles across the 21 Newspapers
between January 1999 and August 2008
300
Number of
articles about
nanotechnology
100
90
250
80
70
200
60
150
50
40
100
30
20
50
10
0
0
1999 2000 2001 2002 2003 2004 2005 2006 2007 Aug08
Percentage of
articles devoted
to
nanotechnology
and risks
181
Figure 4.1. Science Media Use, Elaborative Processing, and Perceived Risks-versus-Benefits of
Perceived Risks-versus-Benefits of Nanotechnology
Nanotechnology (scale ranges only partially displayed on Y-axis)
-8
-9
High Reflective
Integrators
-10
-11
Low Reflective
Integrators
-12
-13
-14
Low Science Media Use
High Science Media Use
182
Figure 4.2 Science Media Use, Elaborative Processing, and Public Support for Federal Funding
Support for Federal Funding of Nanotechnology
of Nanotechnology (scale ranges only partially displayed on Y-axis)
6.2
High Elaboration
6
5.8
5.6
5.4
Low Elaboration
5.2
5
Low Science Media Use
High Science Media Use
183
Figure 4.3. Structural Equation Model Predicting Public Support for Federal Funding of
Nanotechnology: Relationships among Endogenous Variables
MEDIA USE
REFLECTIVE
INTEGRATION
COGNITION
STATE
DISPOSITION
OPINION
.07*
.08*
Science
Discussion
.62***
-.08**
Science Media Use
.13***
Risks-vs-Benefits
Perception
-.25***
-.11***
Support for
Federal Funding of
Nanotechnology
.34***
.10***
Elaborative
Processing
-.16***
.12***
.11***
Factual Scientific
Knowledge
-.27***
.21***
.17***
.21***
Trust in Scientists
Notes. (1) Age, gender, SES, religious beliefs, and deference to scientific authority are controlled
for in this model.
(2) The coefficients in the figure are directional standardized beta coefficients.
(3) The coefficient for the relationship between science discussion and elaborative
processing is non-directional phi-coefficient.
(4) *p<.05, **p<.01, ***p<.001.
184
Figure 6.1 Experts versus Public Support for Federal Funding of Nanotechnology
100
90
Experts
Percent of Respondents Agreeing
80
Public
70
60
50
40
30
20
10
0
Support for Federal Funding of Nanotechnology
185
Figure 6.2. Experts versus Public Perceived Benefits of Nanotechnology
Percent of respondents agreeing
100%
90%
Experts
80%
Public
70%
60%
50%
40%
30%
20%
10%
0%
Better
Treatment of
Diseases
A Cleaner
Environment
A Solution for Revolutionizing
Energy
the Computer
Problems
Industry
Perceived benefits
Improved
National
Security
Improvement in An Economic
Human
Boom
Abilities
186
Figure 6.3. Experts versus Public Perceived Risks of Nanotechnology
100%
90%
Experts
Percent of respondnts agreeing
80%
Public
70%
60%
50%
40%
30%
20%
10%
0%
Loss of Privacy
Use of the
An Arms Race
Technology by
Terrorists
Loss of Jobs
Self-replicating More Pollution
Robots
Perceived risks
New Health
Problems
187
Figure 7.1. Levels of Religious Beliefs, Scientific Status, and Perceived Risks-versus-Benefits of
Perceived Risks-versus-Benefits of Nanotechnology
Nanotechnology (scale ranges only partially displayed on Y-axis)
-4
-5
Public
-6
-7
-8
-9
-10
Experts
-11
-12
-13
-14
Low Level of Religious Beliefs
High Level of Religious Beliefs
188
Figure 7.2. Deference to Scientific Authority, Scientific Status, and Support for Federal Funding
Support for Federal Funding of Nanotechnology
of Nanotechnology (scale ranges only partially displayed on Y-axis)
5
4.8
Experts
4.6
4.4
4.2
4
3.8
3.6
Public
3.4
3.2
3
2.8
2.6
Low Deference
High Deference
189
Figure 7.3. Science Media Use, Scientific Status, and Support for Federal Funding of
Support for Federal Funding of Nanotechnology
Nanotechnology (scale ranges only partially displayed on Y-axis)
5
4.8
Experts
4.6
4.4
4.2
4
3.8
3.6
Public
3.4
3.2
3
2.8
2.6
Low Science Media Use
High Science Media Use
190
Figure 7.4. Trust in Scientists, Scientific Status, and Support for Federal Funding of
Support for Federal Funding of Nanotechnology
Nanotechnology (scale ranges only partially displayed on Y-axis)
5
4.8
Experts
4.6
4.4
4.2
4
3.8
3.6
Public
3.4
3.2
3
2.8
2.6
Low Trust in Scientists
High Trust in Scientists
191
Figure 7.5. Perceived Risks-versus-Benefits, Scientific Status, and Support for Federal Funding
Support for Federal Funding of Nanotechnology
of Nanotechnology (scale ranges only partially displayed on Y-axis)
5
4.8
Experts
4.6
4.4
4.2
4
3.8
3.6
3.4
Public
3.2
3
2.8
2.6
Low Perceived Risks-vs-Benefits
High Perceived Risks-vs-Benefits
192
Table 2.1 Comparison of 2004 and 2007 Public Opinion: Descriptive Statistics of Similar
Question Items
Items
Nanotechnology acceptance
Year
Mean
SD
N
(a) Overall, I support federal funding for
nanotechnology
(1 = Do not agree at all; 10 = Agree very much)
2004
2007
5.36
5.90
2.81
2.85
672
986
(a) Nanotech may lead to the loss of personal
privacy because of tiny new surveillance devices
2004
6.43
2.88
663
2007
5.87
2.72
996
(b) Nanotech may lead to an arms race between
the U.S. and other countries
2004
4.99
2.91
648
2007
5.16
2.70
977
(c) Because of nanotech we may lose more U.S.
jobs
2004
4.89
2.95
661
2007
5.33
2.95
986
(d) Nanotech may lead to the uncontrollable
spread of very tiny self-replicating robots
2004
3.29
2.64
638
2007
3.17
2.33
967
(a) Nanotech may lead to new and better ways to
treat and detect human diseases
2004
7.28
2.49
664
2007
7.15
2.39
990
(b) Nanotech may lead to new and better ways to
clean up the environment
2004
6.56
2.65
664
2007
6.48
2.46
985
(c) Nanotech may give scientists the ability to
improve human physical and mental abilities
(d) Nanotech may help us develop increased
national security and defensive capabilities
2004
2007
2004
6.50
6.41
6.71
2.68
2.46
2.60
664
985
666
2007
6.79
2.36
988
Perceived risks of nanotechnology
(1 = Do not agree at all; 10 = Agree very much)
Perceived benefits of nanotechnology
(1 = Do not agree at all; 10 = Agree very much)
193
General scientific knowledge
(Range = 0-3)
2004
2007
1.77
1.98
1.00
.88
706
1,015
Nanotechnology knowledge
(Range = 0-6)
2004
2007
3.90
4.07
1.55
1.36
706
1,015
2004
2007
3.69
3.99
2.54
2.46
702
1,015
2004
2007
3.18
3.39
2.27
2.18
697
1,014
2004
2007
2004
2007
2004
2007
2004
4.91
4.86
5.08
5.25
4.48
4.96
4.58
3.56
3.45
3.49
3.35
3.18
3.27
3.29
705
1,014
705
1,014
705
1,015
704
2007
4.47
3.22
1,010
2004
2007
2004
2007
4.77
5.65
5.08
6.04
3.23
3.00
3.25
2.76
704
1,006
704
1,013
Level of awareness
(a) How much have you heard, read or seen about
nanotechnology?
(1 = Nothing at all; 10 = Very much)
Level of being informed
(a) How well informed would you say you are
about nanotechnology?
(1 = Not informed at all; 10 = Very informed)
Attention to newspaper content
(a) International Affairs
(b) National government and politics
(c) Stories related to science and technology
(d) Stories about scientific studies in new areas of
research such as nanotechnology
(0 = No attention at all; 10 = Very close attention)
Attention to television content
(a) International Affairs
(b) National government and politics
194
(c) Science and technology
(d) Specific scientific developments, such as
nanotechnology
(e) Science fiction dramas, such as "Lost",
"Surface" or "CSI: Crime Scene Investigation"
(f) Science documentaries on stations such as PBS,
the Learning Channel or Discovery Channel
(0 = No attention at all; 10 = Very close attention)
2004
2007
2004
2007
2004
2007
2004
2007
4.72
5.98
4.02
5.05
2.74
5.67
5.68
6.99
2.89
2.54
2.93
2.55
2.71
3.13
3.11
2.87
704
1,012
703
1,004
703
1,009
705
1,013
2004
2007
2004
2007
2004
2007
2004
2.25
4.34
2.34
4.51
2.38
4.36
2.09
2.92
3.63
2.94
3.60
2.74
3.39
2.56
705
1,014
704
1,014
704
1,014
704
2007
4.12
3.37
1,011
Attention to online news content
(a) News about international affairs
(b) News about national government and politics
(c) Content related to science and technology
(d) Content related to specific scientific
developments, such as nanotechnology
(0 = No attention at all; 10 = Very close attention)
Note. The sample sizes varied due to the different number of missing values in each item.
195
Table 4.1. Descriptive Statistics of Question Items in the 2007 Public Opinion Survey
Question Item
Descriptive Statistics
Attitudinal Outcome Variables
Support for Federal Funding of Nanotechnology
Now, thinking about funding and support for nanotech
research, please tell me how much you agree or disagree
with each of the following statements using a ten-point
scale, where 1 means you do not agree at all and 10 means
you agree very much?
1. “Overall, I support federal funding for nanotechnology.”
(1 = Do not agree at all; 10 = Agree very much)
Mean
SD
N
5.90
2.85
986
Perceived benefits of nanotechnology
Here are a number of statements people have made about
nanotechnology and how it will develop. Thinking about
the future, can you tell me how much you agree with the
following statements?
1. “Nanotech may lead to new and better ways to treat and
detect human diseases.”
2. “Nanotech may lead to new and better ways to clean up
the environment.”
3. “Nanotech may give scientists the ability to improve
human physical and mental abilities.”
4. “Nanotech may help us develop increased national
security and defensive capabilities.”
5. “Nanotech may lead to technologies that will help solve
our energy problems.”
6. “Nanotech may revolutionize the computer industry.”
7. “Nanotech may lead to a new economic boom.”
(1 = Do not agree at all; 10 = Agree very much)
Mean
SD
N
7.15
2.39
990
6.48
2.46
985
6.41
2.46
985
6.79
2.36
988
6.73
2.48
987
7.55
6.49
2.34
2.51
992
987
Perceived risks of nanotechnology
Here are a number of statements people have made about
nanotechnology and how it will develop. Thinking about
the future, can you tell me how much you agree with the
following statements?
1. “Nanotech may lead to the loss of personal privacy
because of tiny new surveillance devices.”
2. “Nanotech may lead to an arms race between the U.S.
and other countries.”
3. “Nanotech may lead to new human health problems.”
4. “Nanotech may be used by terrorists against the U.S.”
5. “Because of nanotech we may lose more U.S. jobs.”
Mean
SD
N
5.87
2.72
996
5.16
2.70
977
4.92
5.34
5.33
2.23
2.60
2.95
973
989
986
196
6. “Nanotech may lead to the uncontrollable spread of
very tiny self-replicating robots.”
7. “Nanotech may lead to more pollution and
environmental contamination.”
(1 = Do not agree at all; 10 = Agree very much)
3.17
2.33
967
4.30
2.15
969
Religious beliefs
1. “How much guidance does religion provide in your
everyday life?”
(1 = No guidance at all; 10 = A great deal of guidance)
Mean
6.00
SD
3.01
N
1,010
Deference to scientific authority
1. “Scientists know best what is good for the public.”
2. “Scientists should do what they think is best, even if
they have to persuade people that it is right.”
(1 = Do not agree at all; 10 = Agree very much)
Mean
3.38
5.21
SD
2.10
3.01
N
1,013
1,009
Science Media Use
Please tell me how much attention you pay to the
following kinds of stories when you read the newspaper:
1. “Stories related to science and technology.”
2. “Stories about scientific studies in new areas of research
such as nanotechnology.”
3. “Stories about the social or ethical implications of
emerging technologies.”
Please tell me how much attention you pay to the
following types of content on television:
1. “Science and technology.”
2. “Specific scientific developments, such as
nanotechnology.”
3. “Information about the social and ethical implications of
emerging technologies.”
When you go online to learn about things, how much
attention do you pay to the following types of news and
information on the Internet?
1. “Content related to science and technology.”
2. “Content related to specific scientific developments,
such as nanotechnology.”
3. “Content related to the social or ethical implications of
emerging technologies.”
(0 = No attention at all; 10 = Very close attention)
Mean
SD
N
4.96
4.47
3.27
3.22
1,015
1,010
4.48
3.06
1,006
5.98
5.05
2.54
2.55
1,012
1,004
5.24
2.63
1,009
4.36
4.12
3.39
3.37
1,014
1,011
3.98
3.15
1,013
Independent Variables
197
Elaborative processing
Now I would like to ask you a few questions about when
you encounter news or information in the media about
science. Please use a 10 point scale where 1 means you do
not agree at all and 10 means you agree very much, to tell
me how much you agree with the statement:
1. “After I encounter news about a scientific development,
I am likely to stop and think about it.”
2. “If I need to act on science information, the more
viewpoints the media give me the better.”
(1 = Do not agree at all; 10 = Agree very much)
Mean
SD
N
6.88
2.45
1,012
7.47
2.57
1,007
Science Discussion
Now I would like to ask you how much you talk about
news with other people. Using a scale from 1 to 10, where
1 means never and 10 means all the time, please tell me
how often you talk with family, friends, or co-workers
about:
1. “Stories related to science and technology.”
2. “Stories about scientific studies in new areas of research
such as nanotechnology.”
3. “Stories about the social or ethical implications of
emerging technologies.”
(1 = Never; 10 = All the time)
Mean
SD
N
4.78
3.93
2.26
2.32
1,013
1,006
4.43
2.69
1,009
Correct
Answers
(%)
N
1,015
1,015
1,015
1,015
Factual scientific knowledge
1. “Lasers work by focusing sound waves.” (False)
2. “Antibiotics kill viruses as well as bacteria” (False)
3. “Electrons are smaller than atoms.” (True)
4. “Ordinary tomatoes do not contain genes, while
genetically modified tomatoes do.” (False)
5. “More than half of human genes are identical to those of
a chimpanzee.” (True)
(1 = True, 2 = False)
67.3
66.5
64.2
75.8
Incorrect
Answers/
Don’t
Know/
Refused
(%)
32.7
33.5
35.8
24.2
70.5
29.5
1,015
Trust in scientists
How much do you trust:
1. “University scientists doing research in
nanotechnology?”
2. “Scientists working for the nanotech industry?”
(1 = Do not trust their information at all; 10 = “Trust their
information very much”)
Mean
SD
N
6.80
2.33
1,012
5.52
2.05
1,005
198
Control Variables
Age
(Range = 18 to 96)
Mean
46.15
SD
17.07
N
995
Gender
Male
(%)
48.6%
Female
(%)
51.4%
N
(0 = Male; 1 = Female)
Socioeconomic Status
Proportion
(%)
1. “What is the highest grade or year of school you
completed?”
1 = Never attended school or only attended
kindergarten
2 = Grades 1 through 8 (elementary)
3 = Grades 9 through 11 (some high school)
4 = Grade 12 or GED (high school graduate)
5 = College 1 year to 3 years (some college or technical
school)
6 = 4-year college graduate
7 = Graduate work
8 = Graduate degree
3.30
1.20
29.1
26.2
2. About how much was your total family income last year
before taxes, was it:
1 = Less than $10,000
2 = $10,000 - $15,000
3 = $15,000 - $20,000
4 = $20,000 - $30,000
5 = $30,000 - $50,000
6 = $50,000 - $75,000
7 = $75,000 - $100,000
8 = Over $100,000
Proportion
(%)
5.20
2.10
4.00
11.2
18.2
18.1
18.5
22.7
1,015
N=
1,010
0.00
20.0
1.10
19.0
N = 865
Note. The sample sizes varied due to the different number of missing values in each item.
199
Table 4.2 Bivariate Correlations among the Variables in the 2007 Public Opinion Survey
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(1)
1.00
(2)
.07*
1.00
(3)
-.24***
-.12***
1.00
(4)
.22***
.18***
-.17***
1.00
(5)
.04
-.04
.05
-.12***
1.00
(6)
-.05
-.10**
.31***
-.05
.15***
1.00
(7)
-.01
.00
.20***
.01
.21***
.41***
1.00
(8)
-.05
.01
.17***
-.01
.15***
.62***
.40***
1.00
(9)
-.24***
-.08*
.38***
-.24***
.07*
.24***
.21***
.16***
1.00
(10)
-.12***
.04
.18***
-.11***
.36***
.30***
.31***
.26***
.15***
1.00
(11)
.09**
.17***
-.29***
.25***
-.14***
-.29***
-.26***
-.17***
-.31***
-.38***
1.00
-.15***
-.10***
.24***
-.21***
.28***
.34***
.31***
.29***
.23***
.44***
-.45***
(12)
(12)
1.00
Note.
Variable names: (1) age, (2) gender (female), (3) socioeconomic status, (4) religious beliefs, (5)
deference to scientific authority, (6) science media use, (7) elaborative processing, (8) science
discussion, (9) factual scientific knowledge, (10) trust in scientists, (11) perceived risks-versusbenefits of nanotechnology, and (12) support for federal funding of nanotechnology; Listwise
Solution (N = 958); *p < .05, **p < .01, ***p < .001.
200
Table 4.3. Ordinary Regression Model Predicting Public Perceived Risks-versus-Benefits of
Nanotechnology (Attitudinal Outcome Variable 1) (standardized regression coefficients)
ZeroOrder
Model
1
Model
2
Model
3
Model
4
Model
5
Model
6
Block 1: Demographics
Age
Gender (female=1)
Socioeconomic status
Incremental R2 (%)
.10***
.17***
-.28***
.02
.14***
-.26***
9.90***
-.01
.11***
-.24***
-.00
.09**
-.17***
-.00
.10**
-.16***
-.03
.10**
-.12***
-.05
.12***
-.11***
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
.24***
-.15***
.17***
-.12***
4.30***
.17***
-.09**
.18***
-.07*
.16***
-.07*
.14***
.02
Block 3: Science Media Use
Science media use
Incremental R2 (%)
-.30***
-.21***
4.00***
-.18***
-.16***
-.11**
Block 4: Reflective Integration
Elaborative processing
Science discussion
Incremental R2 (%)
-.26***
-.18***
-.16***
.03
2.00***
-.14***
.03
-.10**
.04
Block 5: Cognition
Factual scientific knowledge
Incremental R2 (%)
-.31***
-.16***
1.90***
-.15***
Block 6: State-Like Disposition
Trust in scientists
Incremental R2 (%)
-.37***
-.27***
5.50***
--
-.08**
--
-.02
Block 7: Interactions
Science media use ×
Elaborative processing
Science media use ×
Science discussion
Incremental R2 (%)
Total R2 (%)
.50*
28.2***
Note. N = 1,015. Cell entries for all models are final standardized regression coefficients for Blocks 1, 2, 3,
4, 5, and 6, while cell entries are before-entry standardized regression coefficient for Block 7.
*p<.05. **p<.01. ***p<.001.
201
Table 4.4 Ordinary Regression Model Predicting Public Support for Federal Funding of
Nanotechnology (Attitudinal Outcome Variable 2) (standardized regression coefficients)
ZeroOrder
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7
Block 1: Demographics
Age
Gender (female=1)
Socioeconomic status
Incremental R2 (%)
-.15***
-.10***
.23***
-.09**
-.07*
.20***
6.80***
-.09**
-.04
.17***
-.09***
-.02
.10**
-.09**
-.03
.08**
-.08**
-.03
.07*
-.06*
-.06*
.06
-.07**
-.03
.03
Block 2: Trait-Like
Predispositions
Religious beliefs
Deference to scientific
Authority
Incremental R2 (%)
-.21***
.29***
-.12***
.27***
-.12***
.24***
-.13***
.21***
-.13***
.21***
-.11***
.13***
-.08**
.13***
.25***
5.70***
.16***
.15***
.11**
.08*
.17***
.07
2.80***
.16***
.07
.12***
.06
.09**
.07*
.05
.20
.04
.01
9.30***
Block 3: Science Media Use
Science media use
Incremental R2 (%)
.33***
Block 4: Reflective Integration
Elaborative processing
Science discussion
Incremental R2 (%)
.31***
.28***
Block 5: Cognition
Factual scientific knowledge
Incremental R2 (%)
.22***
Block 6: State-Like Disposition
Trust in scientists
Incremental R2 (%)
.43***
Block 7: Perceived Risksversus-Benefits
Perceived risks-versusBenefits
Incremental R2 (%)
-.45***
Block 8: Interactions
Science media use ×
Elaborative processing
Science media use × Science
discussion
Incremental R2 (%)
.26***
5.00***
.19***
-.26***
4.80***
--
.07*
--
.03
Total R2 (%)
.40*
35.0***
Note. N = 1,015. Cell entries for all models are final standardized regression coefficients for Blocks 1, 2, 3, 4, 5, 6, and 7,
while cell entries are before-entry standardized regression coefficient for Block 8. *p<.05. **p<.01. ***p<.001.
202
Table 4.5. Ordinary Regression Model Predicting Public Level of Factual Scientific Knowledge
(Cognitive Outcome) (standardized regression coefficients)
ZeroOrder
Block 1: Demographics
Age
Gender (female=1)
Socioeconomic status
Incremental R2 (%)
-.25***
-.09**
.37***
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
-.23***
.06*
Block 3: Science Media Use
Science media use
Incremental R2 (%)
.26***
Block 4: Reflective Integration
Elaborative processing
Science discussion
Incremental R2 (%)
.22***
.16***
Block 5: Interactions
Science media use × Elaborative processing
Science media use × Science discussion
Incremental R2 (%)
Total R2 (%)
---
Model 1
Model 2
Model 3
Model 4
-.17***
-.04
.32***
16.4***
-.15***
-.01
.30***
-.15***
-.00
.26***
-.15***
-.01
.24***
-.14***
.04
2.00***
-.14***
.01
-.15***
-.00
.16***
2.30***
.13***
.12***
-.01
1.10***
.01
-.03
.20
22.0***
Note. N = 1,015. Cell entries for all models are final standardized regression coefficients for
Blocks 1, 2, 3, and 4, while cell entries are before-entry standardized regression coefficient for
Block 5.
*p<.05. **p<.01. ***p<.001.
203
Table 4.6. Influence of Exogenous Variables on Other Variables
Science Media Use
Elaborative Processing
Science Discussion
Factual Scientific Knowledge
Trust in Scientists
Perceived Risks-versusBenefits
Support for Federal Funding
of Nanotechnology
Age
Gender
SES
Religious
Beliefs
Deference
to
Scientific
Authority
--
-.07
.29
--
.14
--
--
--
--
--
--
-.07
.29
--
.14
--
--
.10
.07
.17
--
-.03
.10
--
.05
--
-.03
.20
.07
.22
--
.08
--
--
.06
--
-.05
.18
--
.08
--
.03
.18
--
.15
-.10
--
.26
-.15
--
--
-.01
.06
.01
.04
-.10
-.01
.31
-.14
.04
-.11
.09
--
--
.32
--
-.02
.09
.01
.06
-.11
.07
.09
.01
.39
--
.16
-.09
.11
--
.04
-.01
-.12
.01
-.14
.04
.15
-.21
.12
-.14
-.07
-.08
--
-.08
.10
-.03
-.03
.13
-.02
.16
-.10
-.11
.13
-.10
.26
Notes. (1) Coefficients in the first row indicate direct effects, coefficients in the second row
indicate indirect effects, and coefficients in the third row indicate total effects.
(2) All coefficients are at least 1.96 times larger than their standard error.
204
(3) Direct and indirect effects may not always add up to total effects due to rounding
error and non-significant pathways.
205
Table 4.7. Relationships among Endogenous Variables
Science Media Use (1)
Elaborative Processing (2)
Science Discussion (3)
Factual Scientific Knowledge (4)
Trust in Scientists (5)
Perceived Risks-versus-Benefits
(6)
Support for Federal Funding of
Nanotechnology (7)
(1)
(2)
(3)
(4)
(5)
(6)
--
--
--
--
--
--
--
--
--
--
--
--
--
--
--
--
--
--
.34
--
--
--
--
--
--
--
--
--
--
--
.34
--
--
--
--
--
.62
--
--
--
--
--
--
--
--
--
--
--
.62
--
--
--
--
--
.11
.12
--
--
--
--
.04
--
--
--
--
--
.16
.12
--
--
--
--
.21
.17
--
--
--
--
.06
--
--
--
--
--
.27
.17
--
--
--
--
-.08
-.11
--
-.16
-.27
--
-.13
-.06
--
--
--
--
-.21
-.17
--
-.16
-.27
--
.07
.10
.08
--
.21
-.25
.19
.08
--
.04
.07
--
.26
.18
.08
.04
.28
-.25
Notes. (1) Coefficients in the first row indicate direct effects, coefficients in the second row
indicate indirect effects, and coefficients in the third row indicate total effects.
(2) All coefficients are at least 1.96 times larger than their standard error.
(3) Direct and indirect effects may not always add up to total effects due to rounding
error and non-significant pathways.
206
Table 4.8. Summary of Findings in the Ordinary Regression Models and Structural Equation
Model
Hypotheses
Regression
Models
Structural
Equation Model
H1a (Religious beliefs Perceived risks-vs-benefits)
Supported
Supported
H1b (Religious beliefs Support for federal funding)
Supported
Supported
H2a (Deference Perceived risks-vs-benefits)
Not supported
Not supported
H2b (Deference Support for federal funding)
Supported
Supported
H3a (Science media use Perceived risks-vs-benefits)
Supported
Supported
H3b (Science media use Support for federal funding)
Supported
Supported
H4a (Elaborative processing Factual scientific
knowledge)
Supported
Supported
H4b (Elaborative processing Perceived risks-vs-benefits)
Supported
Supported
H4c (Elaborative processing Support for federal
funding)
Supported
Supported
H5a (Science discussion Factual scientific knowledge)
Not supported
Not supported
H5b (Science discussion Perceived risks-vs-benefits)
Not supported
Not supported
H5c (Science discussion Support for federal funding)
Supported
Supported
H12a (Factual scientific knowledge Perceived risks-vsbenefits)
Supported
Supported
H12b (Factual scientific knowledge Support for federal
funding)
Not supported
Not supported
H13a (Trust in scientists Perceived risks-vs-benefits)
Supported
Supported
H13b (Trust in scientists Support for federal funding)
Supported
Supported
H16 (Perceived risks-vs-benefits Support for federal
funding)
Supported
Supported
Direct Effects
207
Additive/Interaction Effects
H6a (Elaborative processing × Science media use Factual
scientific knowledge)
Not supported
--
H6b (Scientific discussion × Science media use Factual
scientific knowledge)
Not supported
--
H7a (Elaborative processing × Science media use Perceived risks-vs-benefits)
Supported
--
H7b (Elaborative processing × Science media use Support for federal funding)
Supported
--
H8a (Science discussion × Science media use Perceived
risks-vs-benefits)
Not supported
--
H8b (Science discussion × Science media use Support for
federal funding)
Not supported
--
H9a (Elaborative processing as mediator of science media
use on factual scientific knowledge)
--
Supported
H9b (Scientific discussion as mediator of science media use
on factual scientific knowledge)
--
Supported
H10a (Elaborative processing as mediator of science media
use on perceived risks-vs-benefits)
--
Supported
H10b (Elaborative processing as mediator of science media
use on support for federal funding)
--
Supported
H11a (Science discussion as mediator of science media use
on perceived risks-vs-benefits)
--
Not supported
H11b (Science discussion as mediator of science media use
on support for federal funding)
--
Supported
H14a (Trust in scientists as mediator of science media use
on perceived risks-vs-benefits)
--
Supported
H14b (Trust in scientists as mediator of science media use
on support for federal funding)
--
Supported
H15a (Trust in scientists as mediator of elaborative
processing on perceived risks-vs-benefits)
--
Supported
H15b (Trust in scientists as mediator of elaborative
processing on support for federal funding)
--
Supported
Indirect Effects
208
Table 7.1. Descriptive Statistics of Question Items in the 2007 Experts Survey
Question Item
Descriptive Statistics
Outcome Variables
Support for Federal Funding of Nanotechnology
Thinking of nanotechnology and nanotechnology research,
please indicate your agreement or disagreement with the
following statements.
1. “Overall, I support federal funding for nanotechnology.”
(1 = Strongly disagree; 5 = Strongly agree)
Mean
SD
N
4.69
.66
358
Perceived benefits of nanotechnology
Listed below are a number of predictions people have
made about nanotechnology. Thinking of nanotechnology
and nanotechnology research, please indicate your
agreement or disagreement with the following statements.
1. “Nanotech may lead to new and better ways to treat and
detect human diseases.”
2. “Nanotech may lead to new and better ways to clean up
the environment.”
3. “Nanotech may give scientists the ability to improve
human physical and mental abilities.”
4. “Nanotech may help us develop increased national
security and defensive capabilities.”
5. “Nanotech may lead to technologies that will help solve
our energy problems.”
6. “Nanotech may revolutionize the computer industry.”
7. “Nanotech may lead to a new economic boom.”
(1 = Strongly disagree; 5 = Strongly agree)
Mean
SD
N
4.58
.67
358
4.32
.82
358
3.74
1.11
358
4.10
.93
358
4.27
.90
358
4.24
3.66
.92
1.06
357
358
Perceived risks of nanotechnology
Listed below are a number of predictions people have
made about nanotechnology. Thinking of nanotechnology
and nanotechnology research, please indicate your
agreement or disagreement with the following statements.
1. “Nanotech may lead to the loss of personal privacy
because of tiny new surveillance devices.”
2. “Nanotech may lead to an arms race between the U.S.
and other countries.”
3. “Nanotech may lead to new human health problems.”
4. “Nanotech may be used by terrorists against the U.S.”
5. “Because of nanotech we may lose more U.S. jobs.”
6. “Nanotech may lead to the uncontrollable spread of
very tiny self-replicating robots.”
Mean
SD
N
2.81
1.21
357
2.27
1.16
357
2.96
2.69
1.85
1.51
1.06
1.21
.98
.90
357
357
356
357
209
7. “Nanotech may lead to more pollution and
environmental contamination.”
(1 = Strongly disagree; 5 = Strongly agree)
2.62
1.06
357
Religious beliefs
1. “How much guidance does religion play in your
everyday life?”
(1 = No guidance at all; 10 = A great deal of guidance)
Mean
3.42
SD
2.96
N
351
Deference to scientific authority
Here’s a list of statements other scientists have made about
the relationship between science and the public and about
how scientists should communicate with the public.
Thinking of science more generally, please indicate your
agreement or disagreement with the following statements.
1. “Scientists know best what is good for the public.”
2. “Scientists should do what they think is best, even if
they have to persuade people that it is right.”
(1 = Strongly disagree; 10 = Strongly agree)
Mean
SD
N
2.90
3.94
1.10
.96
361
361
Science Media Use
In general, how much attention do you pay to the
following kinds of content when you read the newspaper?
1. “Science and technology outside of your own field of
research.”
2. “The social or ethical implications of emerging
technologies.”
In general, how much attention do you pay to the
following kinds of content on television?
1. “Science and technology outside of your own field of
research.”
2. “The social or ethical implications of emerging
technologies.”
In general, how much attention do you pay to the
following kinds of content online, i.e., when you use online
news sites, blogs, etc.?
1. “Science and technology outside of your own field of
research.”
2. “The social or ethical implications of emerging
technologies.”
(1 = None; 5 = A lot)
Mean
SD
N
3.12
1.75
361
2.78
1.66
357
3.42
1.19
349
3.03
1.16
348
3.59
1.08
358
3.06
1.11
358
Independent Variables
210
Trust in scientists
To what degree do the following groups currently have the
necessary scientific expertise to communicate about risks
and benefits related to nanotechnology?
1. “University scientists doing research in
nanotechnology.”
2. “Nano scientists working for big companies.”
(1 = Not at all; 5 = “Very much”)
Mean
SD
N
4.40
.79
359
4.02
.99
359
Age
(Range = 28 to 84)
Mean
44.94
SD
10.72
N
343
Gender
Male
(%)
85.6
Female
(%)
14.4
N
Control Variables
(0 = Male; 1 = Female)
354
Note. The sample sizes varied due to the different number of missing values in each item.
211
Table 7.2 Bivariate Correlations among the Variables in the 2007 Experts Survey
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(1)
1.00
(2)
-.03
1.00
(3)
-.08
.13*
1.00
(4)
.00
-.06
-.17**
1.00
(5)
-.02
-.05
.07
-.05
1.00
(6)
-.13*
-.05
-.08
.19***
.04
1.00
(7)
.09
.05
.07
-.17**
-.11*
-.28***
1.00
(8)
-.15**
-.12*
-.14**
.08
.10
.18***
-.30***
(8)
1.00
Note.
Variable names: (1) age, (2) gender (female), (3) religious beliefs, (4) deference to scientific
authority, (5) science media use, (6) trust in scientists, (7) perceived risks-versus-benefits of
nanotechnology, and (8) support for federal funding of nanotechnology; Listwise Solution (N =
328); *p < .05, **p < .01, ***p < .001.
212
Table 7.3. Ordinary Regression Model with Perceived Risks-versus-Benefits of Nanotechnology
as Outcome Variable for the Public Sample (standardized regression coefficients)
Zero-Order
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
.10***
.17***
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
.24***
-.15***
Block 3: Science Communication
Science media use
Incremental R2 (%)
-.30***
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
-.37***
Total R2 (%)
Note. N = 1,015.
*p<.05. **p<.01. ***p<.001.
Model 1
Model 2
Model 3
Model 4
.08**
.16***
3.50***
.05
.12***
.04
.10***
.00
.13***
.19***
-.13***
5.70***
.19***
-.09**
.18***
.01
-.26***
6.70***
-.19***
-.29***
6.70***
22.6***
213
Table 7.4. Ordinary Regression Model with Perceived Risks-versus-Benefits of Nanotechnology
as Outcome Variable for the Expert Sample (standardized regression coefficients)
Zero-Order
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
.10*
.06
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
.07
-.18***
Block 3: Science Communication
Science media use
Incremental R2 (%)
-.11*
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
-.28***
Total R2 (%)
Note. N = 363.
*p<.05. **p<.01. ***p<.001.
Model 1
Model 2
Model 3
Model 4
.10*
.07
1.50
.11*
.05
.10*
.04
.07
.03
.04
-.17***
3.20**
.04
-.18***
.03
-.13**
-.11*
1.30*
-.10*
-.24***
5.50***
11.5***
214
Table 7.5. Ordinary Regression Model with Perceived Risks-versus-Benefits of Nanotechnology
as Outcome Variable with the Public and Expert Samples Combined (standardized regression
coefficients)
ZeroOrder
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
.20***
.27***
Model 1
Model 2
Model 3
Model 4
Model 5
.17***
.24***
9.70***
.11***
.17***
.11***
.15***
.08***
.17***
.03
.10***
.25***
-.11***
7.30***
.26***
-.09***
.24***
-.02
.14***
-.03
-.20***
3.70***
-.14***
-.15***
-.24***
4.90***
-.25***
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
.34***
-.15***
Block 3: Science Communication
Science media use
Incremental R2 (%)
-.23***
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
-.31***
Block 5: Scientific Status
Scientific Status (1 = Expert)
Incremental R2 (%)
-.41***
-.31***
7.00***
---
-.07**
-.04
---
.04
.03
.80**
Block 6: Interactions
Religious beliefs × Status
Deference to scientific authority ×
Status
Science media use × Status
Trust in scientists × Status
Incremental R2 (%)
Total R2 (%)
33.4***
Note. N = 1,378. Cell entries for all models are final standardized regression coefficients for
Blocks 1, 2, 3, and 4, while cell entries are before-entry standardized regression coefficient for
Block 5. Deference to scientific authority, science media use, and trust in scientists variables in
both samples were standardized before analysis to ensure that they are in the same metric.
215
*p<.05. **p<.01. ***p<.001.
216
Table 7.6. Ordinary Regression Model with Support for Federal Funding of Nanotechnology as
Outcome Variable for the Public Sample (standardized regression coefficients)
ZeroOrder
Model 1
Model 2
Model 3
Model 4
Model 5
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
-.14***
-.10***
-.13***
-.09**
2.70***
-.12***
-.05
-.11***
-.03
-.07**
-.06*
-.07**
-.02
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
-.20***
.29***
-.13***
.28***
10.2***
-.13***
.24***
-.12***
.14***
-.07*
.14***
.27***
6.90***
.19***
.14***
.29***
6.40***
.21***
Block 3: Science Communication
Science media use
Incremental R2 (%)
.32***
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
.42***
Block 5: Perceived Risks-versus
Benefits
Perceived risks-versus-benefits
Incremental R2 (%)
Total R2 (%)
Note. N = 1,015.
*p<.05. **p<.01. ***p<.001.
-.43***
-.26***
5.00***
31.2***
217
Table 7.7. Ordinary Regression Model with Support for Federal Funding of Nanotechnology as
Outcome Variable for the Expert Sample (standardized regression coefficients)
ZeroOrder
Model 1
Model 2
Model 3
Model 4
Model 5
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
-.14**
-.10*
-.14**
-.11*
3.10**
-.15**
-.08
-.15**
-.08
-.13**
-.07
-.12*
-.07
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
-.14**
.09*
-.13*
.06
2.30*
-.14**
.07
-.13*
.05
-.12*
.02
.10
.90
.09
.07
.11*
1.20*
.05
Block 3: Science Communication
Science media use
Incremental R2 (%)
.09*
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
.16***
Block 5: Perceived Risks-versusBenefits
Perceived risks-versus-benefits
Incremental R2 (%)
Total R2 (%)
Note. N = 363.
*p<.05. **p<.01. ***p<.001.
-.29***
-.24***
5.20***
12.6***
218
Table 7.8. Ordinary Regression Model with Support for Federal Funding of Nanotechnology as
Outcome Variable with the Public and Expert Samples Combined (standardized regression
coefficients)
ZeroOrder
Model
1
Model
2
Model
3
Model
4
Model
5
Model
6
Block 1: Demographics
Age
Gender (female=1)
Incremental R2 (%)
-.26***
-.25***
-.23***
-.22***
11.5***
-.18***
-.14***
-.17***
-.12***
-.15***
-.14***
-.15***
-.12***
-.06**
-.02
Block 2: Trait-Like Predispositions
Religious beliefs
Deference to scientific authority
Incremental R2 (%)
-.37***
.21***
-.26***
.18***
10.2***
-.27***
.16***
-.26***
.10***
-.24***
.10***
-.07**
.11***
.19***
3.50***
.15***
.12***
.12***
.20***
3.30***
.15***
.15***
-.17***
2.20***
-.21***
Block 3: Science Communication
Science media use
Incremental R2 (%)
.23***
Block 4: State-Like Dispositions
Trust in scientists
Incremental R2 (%)
.31***
Block 5: Perceived Risks-versusBenefits
Perceived risks-versus-benefits
Incremental R2 (%)
Block 6: Scientific Status
Scientific Status (1 = Expert)
Incremental R2 (%)
Block 7: Interactions
Religious beliefs × Status
Deference to scientific authority
× Status
Science media use × Status
Trust in scientists × Status
Perceived risks-versus-benefits ×
Status
Incremental R2 (%)
Total R2 (%)
-.34***
.53***
.48***
16.5***
---
.02
-.08***
----
-.06**
-.11***
.10***
2.00***
48.8***
Note. N = 1,378. Cell entries for all models are final standardized regression coefficients for
Blocks 1, 2, 3, 4, and 5 while cell entries are before-entry standardized regression coefficient for
219
Block 6. Deference to scientific authority, science media use, trust in scientists, and perceived
risks and benefits variables in both samples were standardized before analysis to ensure that
they are in the same metric.
*p<.05. **p<.01. ***p<.001.
220
Appendix A
2004 Public Opinion Survey
In the fall of 2004, a representative national telephone survey with a sample size of N = 706, was
conducted. The cooperation rate (based on standard definitions developed by the American
Association for Public Opinion Research) was 43 percent (AAPOR definition CR-1). The survey
was based on a carefully constructed probability sample that minimizes sampling and nonresponse biases. In this survey, we are particularly concerned about systematic non-response as
a result of the scientific nature and novelty of the survey topic. In other words, it is possible that
the people who chose to respond to our survey are overall more interested in nanotechnology
and related issues and that people who were less aware or less interested in the issue refused to
participate in the survey. (Or, we are aware that individuals who would choose to respond to
our survey might be overall more interested in nanotechnology and related issues than those
who did not participate, mainly because of the scientific nature and novelty of the survey topic.)
This would not only skew the descriptive statistics, but potentially also introduce biases in the
multivariate relationships reported in this dissertation. Therefore, significant amounts of
resources in multiple call-backs for non-contacts and initial refusals were invested in order to
minimize non-response potentially due to the survey topic. (The data for the 2004 public
opinion survey were originally collected by Professor Dietram A. Scheufele, under grants
support from the National Science Foundation, Grant No. SES-0403783.)
221
Appendix B
The following comprehensive search term was used to gather the nanotech-related articles
using the Vantage Point software program:
“atleast3(nanotech!) OR nanosci! OR nanoscal! OR nanocrystal* OR nanotube*OR nanomat! OR
(nanometer* NOT W/15 light or laser or wavelength or UV) OR nanodot* OR nanomed! OR
nanopart! OR nanowir! OR nanoeng! OR nanocomp! OR nanoelectric! OR nanoelectronic! OR
nanobot* OR nanomachine* OR fullerene* OR buckminsterfullerene* OR fullerite* OR
buckyball* OR buckypaper* OR buckytube* OR molecular assembl! OR molecular manufactur!
OR micromachine* OR quantum dot* OR quantum wire* OR quantum well* OR sub micron OR
(individual atom* w/5 manipulate or move or build) OR (scanning w/3 microscope*) OR
(tunneling w/3 microscope*) AND NOT nanosecond* AND NOT apple AND NOT ipod AND
NOT mp3 AND NOT digest AND NOT news w/2 brief* AND NOT business w/2 brief* AND
NOT news summary”