Disentangling the influence of value predispositions and risk/benefit perceptions on support for nanotechnology among the American public

Risk Analysis DOI: 10.1111/risa.12141

Disentangling the Influence of Value Predispositions andRisk/Benefit Perceptions on Support for NanotechnologyAmong the American Public

Jiyoun Kim,1,∗ Sara K. Yeo,1 Dominique Brossard,1 Dietram A. Scheufele,1

and Michael A. Xenos2

Using nanotechnology as a case study, this article explores (1) how people’s perceptions ofbenefits and risks are related to their approval of nanotechnology, (2) which information-processing factors contribute to public risk/benefit perceptions, and (3) whether individuals’predispositions (i.e., deference to scientific authority and ideology) may moderate the rela-tionship between cognitive processing and risk perceptions of the technology. Results indi-cate that benefit perceptions positively affect public support for nanotechnology; perceptionsof risk tend to be more influenced by systematic processing than by heuristic cues, whereasboth heuristic and systematic processing influence benefit perceptions. People who are moreliberal-minded tend to be more affected by systematic processing when thinking about thebenefits of nanotechnology than those who are more conservative. Compared to less defer-ent individuals, those who are more deferent to scientific authority tend to be less influencedby systematic processing when making judgments about the benefits and risks of nanotech-nology. Implications are discussed.

KEY WORDS: Benefit perception; information processing; nanotechnology; risk perception

1. INTRODUCTION

Individuals often employ schematic mentalshortcuts, or heuristics, when forming attitudes aboutunfamiliar subjects.(1,2) This is true for issues inemerging technologies.(3) People are “cognitive mis-ers” and only process as much new informationas they think is necessary to make a judgment orform an attitude.(4,5) The more one knows abouta topic, however, the more likely one is to usenew information to make a decision about the is-

1Department of Life Sciences Communication, University ofWisconsin–Madison, Madison, WI 53706, USA.

2Department of Communication Arts, University of Wisconsin–Madison, Madison, WI 53706, USA.

∗Address correspondence to Jiyoun Kim, Department of Life Sci-ences Communication, University of Wisconsin–Madison, Madi-son, WI 53706, USA; [email protected].

sue at hand, a phenomenon known as systematicprocessing.(1,6) Risk communication scholars haveexpanded this model to predict how people may pro-cess risk-related information.(7) However, the extentto which individuals systematically process new in-formation or use heuristic cues when judging thebenefits and risks related to emerging and, poten-tially controversial, new technologies is still underdebate.

Building on earlier studies, we examine howheuristic and systematic processing may shape pub-lic risk perceptions for complex scientific innova-tions. Using nanotechnology as a case study, this ar-ticle explores how people’s risk/benefit perceptionsare related to their approval of nanotechnology. Wealso explore information-processing factors, whichcontribute to public risk/benefit perceptions and ex-amine whether individuals’ predispositions, such as

1 0272-4332/13/0100-0001$22.00/1 C© 2013 Society for Risk Analysis

2 Kim et al.

deference to scientific authority and ideology, mod-erate the relationship between cognitive processingand risk perceptions of nanotechnology.

2. LITERATURE REVIEW

2.1. Nanotechnology and the American Public:A Primer

Nanotechnology is the science of materials onthe nanoscale, which is one-billionth of a meter.To put this in perspective, a single sheet of paperis about 100,000 nanometers thick.(8) Materials onthe nanoscale behave differently because of theirsize-to-volume ratios, and nanotechnology takes ad-vantage of these properties by using nanomaterialsin new applications.(8) Although nanotechnologyis present in over 1,600 products in the UnitedStates,(9) the American public is still relativelyunfamiliar with it.(9–12) In fact, a 2013 report foundthat roughly 70% of Americans have heard “nothingat all” or “very little” about the technology.(13)

Nanotechnology presents a challenge for severalreasons. First, many scientific fields are encompassedin the term “nanotechnology,” including materialsscience and engineering, chemistry, and molecularbiology. Nanotechnology applications range fromhealth and medicine to food products. Thus, it islikely that the type of application and underlyingscientific field are important for public judgments ofrisks and benefits regarding nanotechnology. Sincenanotechnology is such a broad field, it may bedifficult for citizens to anchor perceptions of risksand benefits when forming attitudes and opinions.(14)

In addition, nanotechnology is an imperceptibletechnology, one that cannot be seen with the nakedeye. This invisible nature means that there are fewreal-world examples to which evaluations of risksand benefits can be anchored.

Although the majority of the American pub-lic may have little familiarity with the issue, newscoverage of nanotechnology has been increasing inprint and online settings.(15,16) Considering that mostAmericans get their technology-related informationfrom newspapers, television, and the Internet,(9,17)

an increase in nanotechnology news coverage mayeventually result in an increase in familiarity with thetechnology. Indeed, people who claim to be famil-iar with nanotechnology often cite newspapers, tele-vision, and the Internet as informational sources.(10)

In addition, the use of these sources for information

about science is related (either directly or indirectly)to increased support for nanotechnology.(18)

This relationship between attention to science inthe media, which we term science media use, andsupport for nanotechnology may exist because ofpositively framed news coverage of emerging tech-nologies. Past content analyses of online and printmedia have shown that news items tend to highlightthe benefits of nanotechnology over its risks.(15,19)

Indeed, media initially tend to highlight the inno-vative features and economic potential of emergingtechnologies.(20) In a recent study of South Carolinacitizens, researchers showed perceptions of the risksrelated to nanotechnology were scarce among thepublic.(21) Yet, results from current research on nan-otechnology content in the media are contradictory.Some studies show that there has been an increas-ing tendency for media reports to focus on the poten-tial risks of nanotechnology,(22) with risk informationgaining some prevalence in online media.(16,23) Otherstudies have found few articles on the risks associ-ated with nanotechnology published in newspapers.Among those published, most contained narrativesabout runaway technology, health, or regulation.(24)

Yet, articles about potential benefits counterbal-anced those that highlighted risks. As is evident fromthese inconsistent findings, the extent to which me-dia use is currently shaping public perceptions of therisks and benefits related to nanotechnology is stillunclear.

2.2. Risk/Benefit Perceptions and Support ofNew Technologies

Research has established that preexisting atti-tudes may influence new attitudes or behaviors re-lated to an issue.(25) In fact, an attitude can bethought of as a learned psychological tendencyto evaluate an issue or entity either positively ornegatively.(26) This evaluative response about an ob-ject can then translate into other attitudes or actionsrelated to the issue at hand. Negative risk perceptionsabout an emerging technology may influence howmuch the public supports or accepts it. For example,previous scholarship has noted that increased riskperceptions surrounding genetically modified organ-isms decreased support for the technology.(27,28) OneEuropean study highlighted more complex mecha-nisms. Although some subjects relied on both riskand benefit perceptions (i.e., trading off betweenthese two aspects) when rendering a judgment about

Information Processing and Risk Perceptions 3

genetically modified foods, others relied predomi-nantly on benefit perceptions when supporting thetechnology.(29)

The public is not homogenous and tends not torespond to risk in a uniform fashion. Human percep-tions of risk are integral to these responses becauseempirical facts tend to be interpreted through indi-vidual values and personal experience.(30) As a re-sult, there is often a difference in how risks are per-ceived between lay audiences and experts. Expertsmay perceive fewer risks than the public and viceversa depending on the issue at hand. For example,studies have shown that compared to experts, lay in-dividuals generally consider nuclear power and nan-otechnology, among other technologies, to be morerisky.(31,32)

Previous scholarship has shown connectionsbetween risk/benefit perceptions and support fornanotechnology.(33,34) Although attitudes toward anissue may only be one factor in determining support,it is clear that risk and benefit perceptions play animportant role in the formation of attitudes towardemerging technologies such as nanotechnology. Onthe basis of the results of earlier studies, we proposethe following hypotheses:

HYPOTHESIS 1: Perceived benefits of nan-otechnology are positively related to generalsupport for nanotechnology.

HYPOTHESIS 2: Perceived risks of nanotech-nology are negatively related to general sup-port for nanotechnology.

2.3. Attitudes and Information Processing

Most Americans get science information fromthe media, and mass media has generally portrayednanotechnology in a positive light.(15,16) However,individuals tend to use the least amount of cogni-tive effort and information to make decisions andjudgments.(4,5) As a result, people often use cogni-tive heuristic cues when assessing new, unfamiliarinformation. Cognitive heuristics are simple mentalmodels that assist individuals in everyday decision-making processes. They require little cognitive ef-fort and are a practical way of forming judgmentsand attitudes. Heuristics allow us to process informa-tion quickly, albeit shallowly, with little effort in or-der to cope with the constant onslaught of data thatwe experience daily. Common cognitive heuristics in-clude affect,(35) familiarity,(12) and availability(12) (for

a more detailed review of cognitive heuristics, seeRef. 36). When information is limited, individualstend to rely on affect heuristics to shape their atti-tudes. Affect heuristics have been shown to be partic-ularly important in the context of nanotechnology(37)

because of the scarcity of information available to layindividuals.

However, the use of heuristics can often leadto biased judgments. Biases that arise due to theuse of heuristics when assessing nanotechnology riskinformation include confirmation bias, risk aver-sion, and alarmist bias.(36) The use of cognitiveheuristics and the resulting biases reinforce thefact that a communication strategy based solelyon educating the public using accurate informa-tion is one that will not be successful in generat-ing consensus, or lowering risk perceptions, aboutnanotechnology.

Although cognitive heuristics can result in ac-curate and appropriate judgments, they can also re-sult in systematic biases and individuals forming un-predictable attitudes in response to risk information.This is especially the case for issues in science andtechnology, where laypersons are often exposed tosuch information through the media. Construction ofattitudes and preferences are informed by percep-tions of risks and benefits,(38) which rely on cognitiveheuristics to process information.

Researchers have proposed several theoriesthroughout the 20th century that explain attitude for-mation through message-based persuasion. Notably,McGuire(39–41) suggests that attitude formation takesplace in a three-step process where an individualpays attention to a message, comprehends the mes-sage content, and accepts or rejects the message’sconclusion. According to this theory, comprehen-sion of a message may be influenced by a host ofother factors. For example, a message from an ex-pert may lead to a more in-depth understanding ofan issue and, in turn, positively influence acceptanceof the message content. This model suggests that at-titude formation depends solely on systematic pro-cessing, or one’s analytical scrutiny of informationpresented in a message.(6,42) Other models of per-suasion, however, theorize that attitude formationmay depend not only on cognitively processing in-formation in messages, but also on peripheral mech-anisms that do not rely on analyzing a message’sargument.(26)

One such model that proposes dual pathways toattitude formation is the heuristic-systematic model

4 Kim et al.

(HSM).3 The HSM postulates that an individualforms attitudes about an issue by judging a mes-sage’s validity with two modes of processing.(26) Onemode of message persuasion can occur through sys-tematic processing, which requires cognitive elabo-ration of message content. This process is time con-suming and cognitively taxing as an individual mustcomprehend and dissect mediated information.(26,44)

Alternatively, heuristic processing allows an indi-vidual to save time and cognitive energy by re-lying on less information when making judgmentsabout a message. This is achieved by using sim-ple decision rules, known as schemata, and mentalshortcuts (all known as heuristics) when processinginformation.(6,12,26)

Although systematic and heuristic processingcan occur at the same time, certain factors mayaffect which mode an individual uses when pro-cessing new information. For example, high motiva-tion to learn facts may increase systematic process-ing, but low cognitive ability may lead to increaseduse of heuristic cues when making judgments aboutnew information.(6,26,31) Although individuals mayuse both processes simultaneously, cognitive miserstend to heuristically process new information to savetime and energy.(4) In other words, people generallyfavor using mental shortcuts such as preexisting be-liefs when judging the validity of new informationrather than spending resources on scrutinizing mes-sages, and this is most likely the case when consider-ing nanotechnology.

On the basis of the HSM, research has exam-ined how people may process risk-related informa-tion. The issue itself seems to be a key determi-nant of how information processing can influencerisk perceptions. For instance, empirical data haveshown that heuristic processing of news stories aboutcancer is negatively related to perceptions of risk,whereas systematic processing is positively relatedto risk perceptions.(45,46) This may occur because themore an individual learns and cognitively processesrisk information related to a specific issue, the morerisky they find that issue to be. On the other hand, ifan individual is predisposed to find something non-

3The elaboration likelihood model (ELM) also considers two dis-tinct paths in persuasion processing, that is, a central path (activethinking mediated condition), and a peripheral route (simple in-ferences without complex cognitive processing). Although boththe HSM and the ELM share the assumption that informationprocessing can occur through dual paths, unlike the HSM, theELM does not assume that the two types of processing can oc-cur simultaneously. In our study, we use the HSM as a theoreticalframework for that reason.(26,43)

risky, she will process information about the issue ina heuristic manner. Both forms of information pro-cessing are related to increased perceptions of therisks of industry on town-life.(47) An explanation forwhy both systematic and heuristic processing wouldincrease risk perception in this case is less clear.

It has also been concluded that individuals tendto rely more on systematic processing of risk in-formation when they perceive a greater need forinformation.(7) Further, Griffin et al. found a posi-tive link between systematic processing of risk in-formation and the strength of people’s evaluations,attitudes, and behavioral beliefs.(42) Also, individ-ual risk perceptions affect their emotional risk re-sponses, which in turn influence their information-seeking intention.(48)

Past research has also linked heuristic processingto inflated risk perceptions.(36,37,49) Specifically, thisscholarship found that heuristic processing that re-lies on shortcuts, mental cues, and mass media maycause individuals to overly dramatize risk-relatedinformation.(26,50) For example, heuristic cues such aslabels on products have been found to increase riskperceptions while lowering benefit perceptions.(51)

Other research suggests that heuristic processing isassociated with an overestimation of risk with smallprobabilities and an underestimation of risks withlarge probabilities.(52–54) This may be because un-usual risks tend to persist in an individual’s mindand are more likely to be cognitively availablewhen making risk assessments. However, some re-cent research has suggested that heuristic processingcan actually lead to better judgments and decisionmaking.(55,56)

Because of these conflicting results, we ex-amine how individuals process information aboutnanotechnology-related risks and benefits in order toreach a judgment about the technology. We, there-fore, posit the following research questions:

RESEARCH QUESTION 1: To what extent doindividuals use systematic processing factorsand heuristic cues when reaching a judgmentabout nanotechnology benefits?

RESEARCH QUESTION 2: To what extent doindividuals use systematic processing factorsor heuristic cues when reaching a judgmentabout nanotechnology risks?

Empirical evidence suggests that people are morelikely to be cognitive misers than systematic pro-cessers when forming attitudes about complex tech-nologies such as nanotechnology,(34) and a varietyof heuristic cues could influence risk perceptions

Information Processing and Risk Perceptions 5

and support of emerging technologies.(33,56–60) Pre-vious studies have highlighted that value pre-dispositions play an important role in publicattitudes toward technologies and can be used asheuristics.(57,61,62) For example, ideology(33) and af-fective variables, for example, negative emotiontoward nanotechnology, influence variation in at-titudes toward nanotechnology.(63) Deference toscientific authority, which is a strongly held valueacquired through the education system, is a vital vari-able in predicting positive attitudes toward a con-troversial issue such as the use of genetic engineer-ing in food production(57) or stem cell research.(60)

In fact, for many citizens, deference to scientificauthority serves as a convenient shortcut that re-places information from mass media or techni-cal knowledge of issues such as genetically modi-fied foods. In addition, technological optimism andfaith in scientists have all been shown to influ-ence support for technologies, such as agriculturalbiotechnology.(29,64)

Previous studies highlight the complex linkbetween information processing and risk/benefitperceptions(46,47) because the HSM allows the si-multaneous engagement of both processes.(6,26) Theheuristic and systematic processes are not mutuallyexclusive, and thus, certain factors may affect whichmode an individual uses when processing new infor-mation. In detail, when people have higher motiva-tion and ability to process messages in a cognitiveway, they tend to use fewer heuristic cues. Neverthe-less, the systematic process can be slanted when anindividual’s heuristic cues influence their systematicprocessing.(26) Thus, in addition to the potential maineffects of heuristic cues and systematic processing onperceived risks and benefits of nanotechnology, wealso posit that heuristic cues will moderate the effectof systematic processing on perception of emergingtechnology.

Specifically, based on previous studies,(65) we ex-pect that individuals’ ideological values will moder-ate the systematic process on perception of bene-fits/risks of the technology.

A recent study points out that more liberal-minded individuals tend to be more supportive ofemerging technologies.(33) We assume that individu-als form their attitudes using both heuristic and sys-tematic processing, but that heuristic and system-atic processes are not mutually exclusive, and thus,certain factors may affect which mode an individualuses when processing new information. Due to thepaucity of empirical findings on the moderating role

of heuristic cues in the relationship between system-atic processing and risk perceptions of nanotechnol-ogy, we put forth the following research questions in-stead of formal hypotheses:

RESEARCH QUESTION 3: Do heuristic cuesmoderate the effect of systematic processingon perceptions of benefit for nanotechnol-ogy?

RESEARCH QUESTION 4: Do heuristic cuesmoderate the effect of systematic processingon perceptions of risk for nanotechnology?

3. METHOD

The data used in this study were collected aspart of a larger online experiment on nanotech-nology in society using KnowledgePanel. Knowl-edgePanel is run by GfK Knowledge Networks andis a probability-based web panel designed to be rep-resentative of the United States. The goal of thebroader study was to examine various communica-tive elements, including civility and incivility, rea-son, emotion, and heterogeneity of opinions in onlinemedia. The experiment dealt with two technologicalissues, nuclear energy and nanotechnology. Partici-pants were randomly assigned to one of these issues,then exposed to blog posts about the issue, and askedseveral posttest questions about the blog posts.4

This study was part of a larger study that in-cluded random assignment to experimental stim-uli. This study is, however, particularly interestedin how individuals’ heuristic and systematic process-ing works on public risk and benefit perceptions of

4This study relied on secondary data analysis. In the original study,participants were randomly assigned to one of four experimentalgroups, each with different definitions of nanotechnology: a tech-nical focus, a technical-application focus, a technical-benefit/risk,and a combination of these foci. After reading the definition ofnanotechnology, participants were required to complete a ques-tionnaire that assessed their attitudes toward the technology. Fol-lowing this, they were asked to read a neutral news blog post withmanipulated comments about technology from a generic fictionalscience blog in a Canadian newspaper (the blog post was writtenby a science writer and contained equivalent risk and benefit in-formation about the technology). All participants were randomlyassigned to one of the 16 of manipulated blog post conditions (in-cluding civility and incivility, reason, emotion, and heterogeneityof opinions in online setting; blog post about either nanotechnol-ogy or nuclear energy). Respondents answered several questionsabout the blog post, including application-specific statements ofrisk and benefit perception. In this study, potential confoundingeffects of the original experimental manipulations are assessedand controlled in the regression models.

6 Kim et al.

nanotechnology and is not focused on the effect ofthe experimental design (e.g., different definition orthe influence of blog post as a stimulus) on their per-ception. Thus, the goal of this study was not to as-sess the effect of these stimuli and data were usedas a secondary source. Potential confounding effectsof the original experimental manipulations were as-sessed and controlled in the regression models.

GfK Knowledge Networks obtains its initial sam-ple using random-digit dialing (RDD) and address-based sampling (ABS). Households that do not al-ready have Internet access are provided a laptopcomputer and online access. The sample for thisstudy was drawn at random from active panel mem-bers. Data were collected in July 2010. Those con-tacted who did not consent to participate were ter-minated from the survey. The completion rate was54.2% and a total of 2,338 people completed thesurvey.5

Participants answered sets of questions aboutmedia use habits and elaborative information pro-cessing, among others. At the end of the survey,participants were asked demographic and value pre-disposition questions, including about religiosity andideology. To avoid any potential bias judgment by re-sponse order effects,(66) each item, for example, therisk or benefit perception questions, was displayedin random order. On average, participants completedthe study in approximately 28 minutes.

3.1. Measures

Sociodemographic variables were used to con-trol potential demographic influences on our threedependent variables: (1) support for nanotechnology;(2) benefit perceptions; and (3) risk perceptions.(67)

These controls included age, which was categorizedinto a seven-point scale (1 = 18–24; 2 = 25–34; 3 =35–44; 4 = 45–54; 5 = 55–64; 6 = 65–74; 7 = above 75;M = 3.77, SD = 1.63), gender (50.6% female), andeducation, which was an ordinal variable measuredwith four categories (1 = less than high school; 2 =high school; 3 = some college; 4 = Bachelor’s degreeor higher; M = 2.86, SD = 0.98).6

5Data were weighted using weights provided by the data collectionagency to be representative of the American population.

6This survey provided respondents with four different definitionsof nanotechnology before asking sets of attitude quesitons aboutthe technology, for example, support, and risk/benefit percep-tions: (1) technical; (2) technical and applications; (3) technicaland risks and benefits; (4) a combination of technical, applica-tions, and risks and benefits. To control the experiments’ effects,

3.1.1. Dependent Variables

The first purpose of this study was to explorethe potential determinants of support for nanotech-nology (i.e., general support for nanotechnology).Specifically, we wanted to explore the influence ofrisk/benefit perceptions on support for nanotechnol-ogy. In this study and as measured elsewhere,(18,34)

support for nanotechnology was measured with amean index of two items measured on 10-point scales(1 = do not agree at all; 10 = agree very much) thatasked how much respondents agree with the follow-ing statements: “Overall, I support federal fundingfor nanotechnology,” and “Overall, I support the useof nanotechnology” (M = 5.04, SD = 2.50, Pearson’sR = 0.80).

As a second step, this study explored whatinfluenced the formation of (1) perceivednanotechnology-related benefits and (2) per-ceived nanotechnology-related risks. Althoughmany researchers interested in risk and benefitperceptions of nanotechnology have relied on asingle item to gauge individuals’ perceptions ofrisks and benefits for an issue, for example, theextent to which the risks attached to a technologywould outweigh its benefits,(68) recent researchhas concluded that this measure may not alwaysbe suitable for nanotechnology, and that a setof application-specific statements may be moreappropriate.(33) Since this study is more interestedin how information processing influences risk andbenefit perceptions, respectively, we measure thesetwo concepts separately. Benefit perceptions ofnanotechnology was measured using an averagedindex of four items measured on 10-point scales(1 = do not agree at all; 10 = agree very much) thatasked respondents their level of agreement with thefollowing statements: “Nanotech may lead to newand better ways to treat and detect human diseases,”“Nanotech may lead to new and better ways to cleanup the environment,” “Nanotech may lead to tech-nologies that will help solve our energy problems,”and “Nanotech may lead to more efficient ways

we added these definitions as control variables in our researchmodel. The definitions of nano are nominal variables, hence weadded these variables as dummy variables and considered the“technical” definition as a reference group. Also, because thesesurvey data were collected in an experimental environment, thatis, exposure to one of 16 blog post conditions, the possible influ-ences of the study’s stimuli were controlled for. That is, althoughour survey includes the experimental designs, that is, definitionof nanotech and blog stimuli, we only dealt with the experimen-tal items as control variables due to the interest of this study.

Information Processing and Risk Perceptions 7

to clean water” (M = 5.98, SD = 1.97, Cronbach’sα = 0.91). Risk perceptions of nanotechnology was anaveraged index of three items measured on 10-pointscales (1 = do not agree at all; 10 = agree very much)that asked how much respondents agree with thefollowing statements: “Nanotech may lead to newhuman health problems,” “Nanotech may lead tomore pollution and environmental contamination,”and “Nanotech may lead to contamination of watersupplies” (M = 5.80, SD = 2.09, Cronbach’s α =0.91).

3.1.2. Independent Variables

The first part of our study examined potentialdeterminants of public support for nanotechnology.Nano news attention was measured using a five-pointscale (1 = none; 5 = a lot) that asked respondentshow much attention they pay to news stories aboutnanotechnology in the newspaper, traditional or on-line television, and online-only sources. These threeitems were averaged to form an index ranging from“1” to “5” (M = 2.30, SD = 0.93, Cronbach’s α =0.84). When considering the social, ethical, and reg-ulatory aspects of nanotechnology, nano-scientiststend to support more regulation of nanotechnologywhen they perceive higher level of risks.(69,70) Thelay public also believes that regulations to controlnanotechnology’s potential risks are important.(71)

Thus, confidence in the regulatory system was mea-sured with a 10-point scale (1 = do not agree atall; 10 = agree very much) asking respondents theirlevel of agreement with the following statement: “Iam confident in the safety and regulatory approvalsystems governing nanotechnology” (M = 4.20, SD= 2.31). Ideology was measured using two six-pointscales ranging from 1 (very conservative) to 6 (veryliberal) on economic and social issues (M = 3.33, SD= 1.24, Pearson’s R = 0.75). Deference to scientificauthority was measured using a single 10-point scale(1 = do not agree at all; 10 = agree very much) ask-ing how much respondents agree with the followingstatement: “Scientists know best what is good for thepublic” (M = 3.89, SD = 2.12). Particularly, in orderto explore how people’s perceptions of benefit andrisk are related to their support for nanotechnology,we chose the same measures for benefit/risk percep-tions that we used in our second and third models asdependent variables.

The second part of our study explored the role ofheuristic and systematic processing in the formationof public attitudes toward nanotechnology, specif-ically benefit and risk perceptions of nanotechnol-

ogy. For the heuristics-related variable, we used thesame measures for value predispositions included inour first model, specifically ideology and deference toscientific authority. In addition, systematic processing-related variables were included in this model. Elabo-rative processing (a measure of how much individu-als actively think about or process news information)was assessed using three items measured on 10-pointscales (1 = do not agree at all; 10 = agree very much)asking respondents their level of agreement with thefollowing statements: “I try to make sense of what Iencounter in the media by comparing it to my ownexperiences,” “After getting information from themedia, I use it to help organize my thoughts,” and“Often when I’ve learned about something in thenews, I’ll recall it later and think about it” (M = 5.24,SD = 2.20, Cronbach’s α = 0.87). This study assumesthat an individual’s inclination toward “elaborativeprocessing” and “need for information” are indica-tive of systematic information processes. Individualswho attempt to better understand, elaborate on, andmake sense of news media content are more likely tosystematically process any information provided.(72)

Further, a strong perception of information insuffi-ciency, that is, a higher “need for information,” willbe positively linked to greater motivation for system-atic processing.(7,45,47) Need for information was mea-sured using a single 10-point scale item (1 = do notagree at all; 10 = agree very much) asking how muchrespondents agree with the following statement: “Iwould need more information about nanotechnologybefore I could make any decisions about it” (M =6.74, SD = 2.81). We used the same measure of nanonews attention included in our first model.

3.2. Analytical Techniques

For both studies (part I and part II), we usedhierarchical ordinary least squares regression mod-els to test our hypotheses and research questions.Specifically, to find the predictors’ relative explana-tory variances for the final model, we entered theindependent variables into blocks based on their as-sumed causal order.

Interaction terms were constructed in order toassess the degree of moderators’ influences on (1)public benefit perceptions of nanotechnology and (2)risk perceptions of the technology. To avoid issuesof multicollinearity between an interaction term andits component, each of the interaction terms wascreated by multiplying the standardized value of itscomponents.(73)

8 Kim et al.

Table I. Regression Predicting Support for Nanotechnology(N = 2,177)

β

Control:a

Block 1: Experimental DesignDefinition Dummy application 0.10***

Definition Dummy risks&benefits − 0.01Definition Dummy combination 0.02Comment civility − 0.05***

Comment emotion − 0.03*

Comment homogeneity − 0.01Blog Topic Nano 0.06***

Incremental R2 (%) (�F)b 2.0*** (6.2)Test:Block 2: Demographics

Age − 0.05**

Genderc − 0.05**

Education 0.12***

Incremental R2 (%) (�F) 11.8*** (98.4)Block 3: Value Predispositions

Ideology 0.08***

Incremental R2 (%) (�F) 2.8*** (72.1)Block 4: Attitude Toward Science

Deference to scientific authority 0.15***

Confidence regulatory system 0.13***

Nano news attention 0.20***

Incremental R2 (%) (�F) 23.7*** (285.0)Block 5: Perceptions

Benefit perception 0.37***

Risk perception − 0.06***

Incremental R2 (%) (�F) 9.0*** (192.3)Total R2 (%) 49.2***

Note: Cell entries are standardized regression coefficients for eachblock in which all independent variables have been included.*p < 0.05; **p < 0.01; ***p < 0.001.aBecause our survey data were collected in experimental environ-ments, the influences of experimental design are controlled in aregression equation.b�F = F change value.cThe variable gender was coded 0 = male, 1 = female.

4. RESULTS

As Table I illustrates, after controlling for allother independent variables, individuals’ ideologywas significantly related to support for nanotechnol-ogy. That is, more liberal individuals were more sup-portive of nanotechnology (β = 0.08, p < 0.001).

A higher level of deference to scientific authorityis positively linked to nanotechnology-related sup-port (β = 0.15, p < 0.001). In addition, individualswho pay more attention to nanotechnology news inmass media are more likely to support the technol-ogy (β = 0.20, p < 0.001).

Our first hypothesis was supported, with the datashowing a significant positive relationship between

benefit perceptions and support for nanotechnology(β = 0.37, p < 0.001). As expected, the greater theperceived nanotechnology risks, the less supportiveindividuals are of the technology (β = –0.06, p <

0.001). Hence, our second hypothesis was also sup-ported.

As a second step, our study sought to identifythe determinants of public perceptions about nan-otechnology (RQ1 and RQ2), and more specificallythe predictive variables of benefit/risk perceptions.A total of 30.1% of the variance of benefit percep-tions was explained by our model, whereas a to-tal of 23.8% of the variance was explained by theindependent variables in the model predicting riskperceptions.

In response to our first research question, peopletend to use both systematic processing and heuristiccues when perceiving nanotechnology-related bene-fits. With respect to heuristic cues, deference to sci-entific authority is positively and significantly associ-ated with benefit perceptions (β = 0.24, p < 0.001).As expected, individuals who are more deferent toscientific authority perceive more benefits than thosewith less scientific deference. Thus, this study infersthat individuals form benefit perceptions while re-lying on certain predispositions, for example, defer-ence to scientific authority.

In addition, all of the systematic processing vari-ables in the model show significant relationships withbenefit perceptions. Together, the systematic pro-cessing variables account for 2.7% of the variance inthe model whereas the heuristic processing variablesexplain 7% of the variance. Individuals who activelyprocess information perceive more benefits in nan-otechnology than those who do not (β = 0.20, p <

0.001). In addition, there is a significant but negativerelationship between need for information and ben-efit perceptions (β = –0.05, p < 0.05).

Interestingly, this study found significant interac-tion effects between the heuristic processing factorsand cognitive processing factors in the benefit per-ceptions model (RQ3).7 As shown in Fig. 1, liberal-leaning individuals are less likely to perceive bene-fits when they have a greater need for information(β = –0.05, p < 0.01). Also, as shown in Fig. 2, peo-ple who are less deferential to scientific authority aremore likely to perceive benefits when they have agreater need for information (β = –0.13, p < 0.001).In addition, people who are less deferential toward

7Following the method proposed by Aiken and West,(74) each in-teraction is presented in Figs. 1–4.

Information Processing and Risk Perceptions 9

Fig. 1. Relationship between systematicprocessing (i.e., need for information)and benefit perception, which is moder-ated by heuristic trait (i.e., ideology).

Fig. 2. Relationship between systematic processing (i.e., need for information) and benefit perception, which is moderated by heuristic trait(i.e., deference to scientific authority).

scientific authority tend to use elaborative processingwhen they perceive benefits of nanotechnology (β =–0.05, p < 0.05; see Fig. 3).

More precisely, there is a negative interactioneffect between perceived need for information andideology when predicting benefit perceptions of

nanotechnology (β = –0.05, p < 0.01). That is, thereare negative relationships between perceived needfor information and benefit perceptions of nanotech-nology, but individuals’ ideology moderates theselinks. Our data show that the relationship is largelymoderated among liberal-minded respondents.

10 Kim et al.

Fig. 3. Relationship between systematic processing (i.e., elaborative process) and benefit perception, which is moderated by heuristic trait(i.e., deference to scientific authority).

Similarly, there is a significant negative interactioneffect between perceived need for information andscientific deference in predicting benefits (β = –0.13,p < 0.001).

Elaborative processing is positively associatedwith nanotechnology benefit perceptions, but thislink is moderated by deference to scientific author-ity (β = –0.05, p < 0.05). Particularly, the relation-ship is stronger among those with lower deferenceto scientific authority compared to those with greaterdeference.

Our second research question asked whether in-dividuals use heuristic or systematic processing whenrendering a judgment about risks related to nan-otechnology. According to our data, systematic pro-cessing factors influence risk perceptions more thanheuristic shortcuts (Table II). Although individuals’ideology (β = 0.05, p < 0.01) and deference to scien-tific authority (β = –0.07, p < 0.01) are significantlyrelated to risk perceptions, the systematic processing-related variables have more explanatory power inpredicting risk perceptions related to nanotechnol-ogy (7.9% of variance for cognitive processing in pre-dicting risk perceptions compared to only 0.3% of thevariance for heuristic factors). In particular, if peopleassume that they need more information related tonanotechnology, they tend to perceive the technol-ogy to be riskier (β = 0.26, p < 0.001).

Noticeably, we find that deference to scientificauthority is a significant moderator of the relation-ship between need for information and risk percep-tion (β = –0.09, p < 0.001). Individuals’ need forinformation shows a significant and positive influ-ence on risk perceptions; however, this relationshipis moderated by scientific deference (Fig. 4).

5. DISCUSSION

This study analyzes public attitudes relatedto nanotechnology (i.e., public support and publicrisk/benefit perceptions), how individuals’ preexist-ing attitudes can affect their general attitudes, andhow heuristic and systematic processing inform pub-lic risk and benefit perceptions related to nanotech-nology. We found that perception of nanotechnologybenefits is a main predictor of support for the tech-nology. Interestingly, our data indicate that individu-als tend to be more influenced by systematic process-ing factors than heuristic cues when developing riskperceptions, while both types of information process-ing affect benefit perceptions. In addition, significantinteractions show that the relationships between sys-tematic processing and risk/benefit perceptions tendto be moderated by value predispositions, primarilythe level of deference to scientific authority.

Information Processing and Risk Perceptions 11

Table II. Regression Predicting Benefit/Risk Perceptions of Nanotechnology (N = 2,147)

Benefit (β) Risk (β)

Control: a

Block 1: Experimental DesignDefinition Dummy application − 0.01 0.02Definition Dummy risks&benefits − 0.03 0.03Definition Dummy combination 0.02 0.02Comment civility − 0.00 − 0.05Comment emotion 0.02 − 0.00Comment homogeneity 0.02 − 0.01Blog Topic Nano − 0.03 0.35***

Incremental R2 (%) (�F) b 0.4 (1.1) 12.5***(43.6)Block 2: Response Extremity

Risk perception of benefit/benefit perception of risk 0.11*** 0.12***

Media use − 0.08*** 0.03Incremental R2 (%) (�F) 2.1*** (23.3) 1.5*** (18.2)Test:Block 3: Demographics

Age 0.06** − 0.03Genderc − 0.05** 0.06***

Education 0.11*** − 0.02Incremental R2 (%) (�F) 6.7*** (52.8) 0.8***(7.1)Block 4: Science Communication

Nano news attention 0.20*** 0.00Incremental R2 (%) (�F) 9.5*** (248.9) 0.0 (0.2)Block 5: Heuristic Traits

Ideology − 0.00 0.05**

Deference to scientific authority 0.24*** − 0.07**

Incremental R2 (%) (�F) 7.0*** (99.7) 0.3* (4.2)Block 6: Systematic Processing

Need for information − 0.05* 0.26***

Elaborative process 0.20*** 0.11**

Incremental R2 (%) (�F) 2.7*** (40.4) 7.9*** (108.5)Block 7: Interactions

Need for information*Ideology − 0.05** − 0.03Need for information*Deference to scientific authority − 0.13*** − 0.09***

Elaborative process*Ideology − 0.02 − 0.01Elaborative process*Deference to scientific authority − 0.05* − 0.02

Incremental R2 (%) (�F) 1.7*** (12.7) 0.8** (5.7)Total R2 (%) 30.1*** 23.8***

Note: Cell entries are final standardized regression coefficients for blocks 1, 2, 3, 4, 5, and 6 in which all independent variables have beenincluded while cell entries for block 7 are before-entry standardized regression coefficients that control for variables entered in the previousblock but not for interactions entered in the same block.*p < 0.05; **p < 0.01; ***p < 0.001.aBecause our survey data were collected in experimental environments, the influences of experimental design and the potential responseextremity are controlled in a regression equation.b�F = F change value.cThe variable gender was coded 0 = male, 1 = female.

Before discussing our results in more detail, weneed to acknowledge some limitations of our study.First, researchers often use multiple scales and com-bined questions to measure variables with greaterpredictive power. In this study, single-item measureswere employed for some of the variables, which maylimit the impact of our findings. Future research

should attempt to replicate our results with more so-phisticated, multi-item measures. We also acknowl-edge that using “need for information” as a proxy forsystematic information processing may be disputed.However, we believe this is a valid use of this mea-sure as individuals who process information in a cog-nitive manner tend to use as much information as

12 Kim et al.

Fig. 4. Relationship between systematic processing (i.e., need for information) and risk perception, which is moderated by heuristic trait(deference to scientific authority).

possible for decision making. Unlike cognitive mis-ers, systematic processors tend to put more effortinto decision making. Therefore, providing individu-als who process information systematically with moreinformation generally allows them to make rationaljudgments about an issue.

The third concern is related to incommensurabil-ity among the dependent variables. That is, the itemabout “support for nanotechnology” was measuredby generic categories, for example, “support the useof nanotechnology” and “support federal funding fornanotechnology,” whereas the risk/benefit percep-tions items were measured by a set of application-specific statements. As Cacciatore et al. have con-cluded, specific applications of a technology can workas primers in ensuing judgment, such as the transitionfrom risk/benefit perceptions to general attitudes to-ward nanotechnology.(33)

Furthermore, although our goal was to analyzethe relationship between information processingand risk/benefit perception of nanotechnology,we were less interested in the exact topic. Thatis, this study measured overall risk perceptionsusing a set of application-specific statements, forexample, nanotech in health, environment, etc.,but did not present each risk perception of specificapplications of nanotechnology separately. Hence,we suggest that future research should explore

application-specific risk and benefit perceptions ofnanotechnology.

In order to conceptualize heuristic-systematicprocessing, we measured each variable, that is, ide-ology, deference to scientific authority, need for in-formation, and elaborative process, separately. It ispossible to combine these variables into one con-struct instead of treating them separately as we havehere. If we had opted to do the former, we couldpotentially have focused more on the componentsof the larger construct, that is, heuristic-systematicprocessing. However, in doing so we would likelyhave missed the significant influence of each indi-vidual variable. Therefore, we decided to treat thevariables separately in this study. Consequently, weobserved that heuristic-systematic processing wasexplained by the incremental variance in themodel, as well as the unique effect of eachvariable.

For the sake of parsimony, our study includedkey variables of choice and omitted some that wouldbe important to integrate. Future research shouldhence also integrate affect as a key heuristic as an im-portant predictor of support.(75,76) Despite these lim-itations, we believe that this study offers reliable in-sights that contribute to the understanding of publicattitudes toward emerging technologies as they relateto perceptions of risks and benefits.

Information Processing and Risk Perceptions 13

Unlike past studies suggesting Americans aregenerally optimistic about nanotechnology,(68,71) ourdata show that Americans have a relatively skep-tical attitude toward the technology. About 32%believed that the risks outweighted the benefits ofnanotechnology, whereas only 23% believed thebenefits outweighted risks. This is, in fact, nota surprise. Although initial media coverage wasmainly positive, focusing on the potential bene-fits of nanotechnology,(77,78) recent coverage hastended to cover more potential risks of nanotech-nology, and the dominant positive tone has beenchallenged.(15,16,22) If media effects on public atti-tudes related to emerging technological issues re-main powerful, we might expect a continuing down-ward trend of public support for nanotechnology inthe future.

As Brossard et al.(58) suggest, people use per-ceptual filters as interpretative frameworks whenforming attitudes toward a controversial technology.Thus, we examined how people perceive nanotech-nology, and explored determinants of public supportfor nanotechnology. Our findings show that individ-uals’ value predispositions (i.e., ideology, and atti-tudes toward science) are significant influences onnanotechnology support.

Our data confirm that public risk/benefit percep-tions play an important role in shaping public accep-tance of the technology. As expected, there is a pos-itive link between benefit perceptions and supportfor nanotechnology. Although there is also a signif-icant relationship between risk perception and nan-otechnology support, the expected relationship be-tween benefit perceptions and support is stronger.Nevertheless, as Currall et al.(79) point out, much ofthe current dialogue about the future of nanotech-nology focuses on its potential risks, and the federalgovernment tends to provide more funding for short-and long-term studies of nanotechnology risks.(80) Ofcourse, we are not implying that focusing on risk-related studies is not (or less) important, but ourfindings suggest that highlighting the benefits of nan-otechnology (while not downplaying the risks) couldbe one of the key strategies to enhance public sup-port if this was a goal of a communication exercise.Of course, the pros and cons of this type of strategyshould be carefully assessed in light of ethical consid-erations.

As a second step in our analysis, we were inter-ested in exploring different attitude formation pro-cessing pathways (heuristic vs. systematic) in orderto understand how individuals reach judgment about

the risks and benefits of nanotechnology based onmass media messages. Interestingly, our data showthat systematic processing of information has a re-lationship with risk perceptions while both system-atic and heuristic processing influence benefit per-ceptions. Indeed, perceived need for informationnoticeably and positively affects perception of riskrelated to nanotechnology. This suggests that peo-ple tend to adopt a more cautious and conservativeapproach when forming a risk-related attitude whenthey have less information about a controversialissue.

According to previous research,(45–47,49) the re-lationship between individuals’ systematic process-ing and benefit/risk perceptions is quite complex, andthe relationship does not always follow the same pat-tern. It can also be topic-dependent. Noticeably, oneof the interesting findings of this study is the pos-itive interaction effect between systematic process-ing and individuals’ deference to scientific authorityon benefit/risk perceptions. Our study suggests thatsystematic processing of information is positively as-sociated with public benefit/risk perceptions of nan-otechnology, but this relationship is moderated byindividuals’ deference to scientific authority. Over-all, individuals with lower deference perceived fewerbenefits to nanotechnology. However, among theseindividuals, those who had a higher need for infor-mation and scored higher on the elaborative process-ing scale perceived greater benefits. These data couldindicate that individuals who are less deferent to sci-entific authority are less likely to use heuristic cues,and more likely to process information systematicallywhen it comes to nanotechnology.

6. CONCLUSION

This study assesses information-processing ap-proaches that individuals employ when formingrisk/benefit perceptions about the technology, andthe subsequent support for it. First, the public’s in-creasing perception of risks related to nanotechnol-ogy will surely have implications for future outreachefforts, as well as policy and regulatory efforts. Sec-ond, our finding that individuals tend to employ sys-tematic processing when forming risk perceptions re-lated to nanotechnology contributes to the growingliterature on information processing and risk percep-tions. It may be somewhat encouraging that peopleare likely to scrutinize mass-mediated informationand be cognitively involved when forming opinionson the riskiness of nanotechnology.

14 Kim et al.

In addition to theoretical implications, this studyis also important for practical reasons. We can usethese findings to inform designs of effective risk andbenefit communication campaigns. For citizens whotend to process information systematically, commu-nicators may simply be able to provide more infor-mation to facilitate rational decision making. How-ever, value predispositions need to be taken intoaccount when designing campaigns aimed at individ-uals who process information and reach judgmentsusing primarily heuristic cues.

ACKNOWLEDGMENTS

This study is based on work supported by a grantfrom the UW-Madison Nanoscale Science and En-gineering Center in Templated Synthesis and As-sembly at the Nanoscale (Grant No. SES-DMR-0832760). Any opinions, findings, and conclusionsor recommendations expressed in this material arethose of the authors and do not necessarily reflect theviews of the National Science Foundation.

REFERENCES

1. Chaiken S. Heuristic versus systematic information processingand the use of source versus message cues in persuasion. Jour-nal of Personality and Social Psychology, 1980; 39(5):752–766.

2. Chaiken S. The heursitc model of persuasion. Pp. 3–39 inZanna MP, Olson JM, Herman CP (eds). Social Influence: TheOntario Symposium, Vol. 5. Ontario Symposium on Personal-ity and Social Psychology. Hillsdale, NJ: Lawrence ErlbaumAssociates, 1987.

3. Scheufele DA. Messages and heuristics: How audiences formattitudes about emerging technologies. Pp. 20–25 in Turney J(ed). Engaging Science: Thoughts, Deeds, Analysis and Ac-tion. London: Wellcome Trust, 2006.

4. Fiske ST, Taylor SE. Social Cognition, 2nd ed. New York:McGraw-Hill, 1991.

5. Popkin S. The Reasoning Voter: Communication and Per-suasion in Presidential Campaigns. Chicago: University ofChicago Press, 1991.

6. Chaiken S, Eagly A. Heuristic and systematic information pro-cessing within and beyond the persuasion context. Pp. 212 inUleman JS, Bargh JA (eds). Unintended Thought. New York:Guilford Press, 1989.

7. Kahlor L, Dunwoody S, Griffin RJ, Neuwirth K, Giese J.Studying heuristic-systematic processing of risk communica-tion. Risk Analysis, 2003; 23:355–368.

8. National Nanotechnology Initiative. What is Nanotechnol-ogy? Available at: http://www.nano.gov/nanotech-101/what/definition, Accessed on October 25, 2013.

9. Project on Emerging Nanotechnologies. Consumer Prod-ucts Inventory: An Inventory of Nanotechnology-Based Con-sumer Products Introduced on the Market. Available at:http://www.nanotechproject.org/cpi/, Accessed on October 25,2013.

10. Peter D. Hart Associates. Report Findings Based on aNational Survey of Adults. Washington, DC: WoodrowWilson International Center for Scholars Project on

Emerging Nanotechnologies, 2006. Available at: http://www.nanotechproject.org/file download/files/HartReport.pdf,Accessed on March 28, 2011.

11. Peter D. Hart Associates. Awareness of and AttitudesToward Nanotechnology and Federal Regulatory Agen-cies. Washington, DC: Project on Emerging Nanotech-nologies the Woodrow Wilson International Center forScholars, 2007. Available at: http://www.nanotechproject.org/process/files/5888/hart nanopoll 2007.pdf, Accessed on March28, 2011.

12. Tversky A, Kahneman D. Judgment under uncertainty:Heuristics and biases. Science, 1974; 185(4157):1124–1131.

13. Hart Research Associates. Awareness and Impressionsof Synthetic Biology: A Report of Findings, Based ona National Survey Among Adults. Washington, DC:Project on Synthetic Biology Project the Woodrow Wil-son International Center for Scholars, 2013. Available at:https://www.cbd.int/doc/emerging-issues/emergingissues-2013-07-WilsonCenter-SynbioSurvey-en.pdf, Accessed on Septem-ber 26, 2013.

14. Pidgeon N, Harthorn B, Satterfield T. Nanotechnologyrisk perceptions and communication: Emerging technolo-gies, emerging challenges. Risk Analysis, 2011; 31(11):1694–1700.

15. Dudo A, Dunwoody S, Scheufele DA. The emergence of nanonews: Tracking thematic trends and changes in media cover-age of nanotechnology. Journalism and Mass CommunicationQuarterly, 2011; 88:55–75.

16. Cacciatore MA, Anderson AA, Choi DH, Brossard D,Scheufele DA, Liang X, Ladwig PJ, Xenos M, Dudo A. Cov-erage of emerging technologies: A comparison between printand online media. New Media & Society, 2012; 14(6):1039–1059.

17. Nelkin D. Science controversies: The dynamics of public dis-putes in the United States. Pp. 444–456 in Jasanoff S, MarkleGE, Petersen JC, Pinch T (eds). Handbook of Science andTechnology Studies. Thousand Oaks, CA: Sage PublicationsInc., 1995.

18. Lee CJ, Scheufele DA. The influence of knowledge and def-erence toward scientific authority: A media effects model forpublic attitudes toward nanotechnology. Journalism and MassCommunication Quarterly, 2006; 83(4):819–834.

19. Ladwig P, Anderson AA, Brossard D, Scheufele DA, ShawB. Narrowing the nano discourse? Materials Today, 2010;13(5):52–54.

20. Nisbet MC, Huge M. Attention cycles and frames in theplant biotechnology debate: Managing power and participa-tion through the press/policy connection. Harvard Interna-tional Journal of Press-Politics, 2006; 11(2):3–40.

21. Priest S, Lane T, Greenhalgh T, Hand LJ, Kramer V. Envi-sioning emerging nanotechnologies: A three-year panel studyof South Carolina citizens. Risk Analysis, 2011; 31(11):1718–1733.

22. Weaver DA, Lively E, Bimber B. Searching for a frame: Newsmedia tell the story of technological progress, risk, and regula-tion. Science Communication, 2009; 31(2):139–166.

23. Anderson AA, Brossard D, Scheufele DA. The changinginformation environment for nanotechnology: Online audi-ences and content. Journal of Nanoparticle Research, 2010;12(4):1083–1094.

24. Friedman SM, Egolf BP. A longitudinal study of newspaperand wire service coverage of nanotechnology risks. Risk Anal-ysis, 2011; 31(11):1701–1717.

25. Ajzen I, Fishbein M. The influence of attitudes on behav-ior. Pp. 173–221 in Albarracın D, Johnson BT, Zanna MP(eds). The Handbook of Attitudes. Mahwah, NJ: Erlbaum,2005.

26. Eagly A, Chaiken S. The Psychology of Attitudes. Fort Worth,TX: Harcourt Brace Jovanovich College Publishers, 1993.

Information Processing and Risk Perceptions 15

27. Ferber D. Risks and benefits: GM crops in the cross hairs. Sci-ence, 1999; 286(5445):1662–1666.

28. Gaskell G, Bauer MW, Durant J, Allum NC. Worlds apart?The reception of genetically modified foods in Europe and theU.S. Science, 1999; 285(5426):384–387.

29. Gaskell G, Allum N, Wagner W, Kronberger N, Torgersen H,Hampel J, Bardes J. GM foods and the misperception of riskperception. Risk Analysis, 2004; 24(1):185–194.

30. Slovic PF, Fischhoff B, Lichtenstein S. Facts and Fears: Under-standing Perceived Risk Societal Risk Assessment: How Safeis Safe Enough. New York: Plenum Press, 1980.

31. Siegrist M, Cousin ME, Kastenholz H, Wiek A. Public accep-tance of nanotechnology foods and food packaging: The influ-ence of affect and trust. Appetite, 2007; 49(2):459–466.

32. Slovic P. Perception of risk. Science, 1987; 236(4799):280–285.33. Cacciatore MA, Scheufele DA, Corley EA. From enabling

technology to applications: The evolution of risk perceptionsabout nanotechnology. Public Understanding of Science, 2011;20(3):385–404.

34. Scheufele DA, Lewenstein BV. The public and nanotechnol-ogy: How citizens make sense of emerging technologies. Jour-nal of Nanoparticle Research, 2005; 7:659–667.

35. Finucane ML, Alhakami A, Slovic P, Johnson SM. The affectheuristic in judgments of risks and benefits. Journal of Behav-ioral Decision Making, 2000; 13(1):1–17.

36. Berube DM, Faber B, Scheufele DA, Cummings CL, Gard-ner GE, Martin KN, Martin MS, Temple NM. CommunicatingRisk in the 21st Century: The Case of Nanotechnology. Ar-lington: National Nanotechnology Coordination Office, 2010.

37. Kahan D, Slovic P, Braman D, Gastil J, Cohen G, Kysar D. Bi-ased Assimilation, Polarization, and Cultural Credibility: AnExperimental Study of Nanotechnology Risk Perceptions. TheCultural Cognition Project. New Haven, CT: Yale Law School,2008.

38. Slovic P. The construction of preference. American Psycholo-gist, 1995; 50(5):364–371.

39. McGuire WJ. Personality and attitude change: Aninformation-processing theory. Pp. 171–196 in GreenwaldAG, Brock TC, Ostrom TM (eds). Psychological Foundationsof Attitudes. San Diego, CA: Academic Press, 1968.

40. McGuire WJ. Personality and susceptibility to social influence.Pp. 1130–1187 in Borgatta EF, Lambert WW (eds). Handbookof Personality Theory and Research. Chicago: Rand McNally,1968.

41. McGuire WJ. Attitude change: The information processingparadigm. Pp. 108–141 in McClintock CG (ed). Experimen-tal Social Psychology. New York: Holt, Rinehart & Winston,1972.

42. Griffin RJ, Neuwirth K, Giese J, Dunwoody S. Linking theheuristic-systematic model and depth of processing. Commu-nication Research, 2002; 29(6):705–732.

43. Petty RE, Brinol P,Priester JR. Mass media attitude change:Implications of the elaboration likelihood model of persua-sion. Pp. 125–163 in Bryant J, Oliver MB (eds). Media Ef-fects: Advances in Theory and Research. New York: Rout-ledge, 2009.

44. Petty RE, Cacioppo J. Communication and Persuasion: Cen-tral and Peripheral Routes to Attitude Change. New York:Springer, 1986.

45. Trumbo C. Heuristic systematic information processing andrisk judgment. Risk Analysis, 1999; 19(3):391–400.

46. Trumbo C. Information processing and risk perception: Anadaptation of the heuristic systematic model. Journal of Com-munication, 2002; 52(2):367–382.

47. Johnson B. Testing and expanding a model of cognitive pro-cessing of risk information. Risk Analysis, 2005; 25(3):631–650.

48. Kahlor L. PRISM: A planned risk information seeking model.Health Communication, 2010; 25:345–356.

49. Lopes L. The rhetoric of irrationality. Theory and Psychology,1991; 1(1):65–82.

50. Dunwoody S. The media and public perceptions of risk:How journalists frame risk stories. Pp. 75–100 in BromleyD, Segerson K (eds). The Social Response to EnvironmentalRisk. Boston, MA: Kluwer, 1992.

51. Siegrist M, Keller C. Labeling of nanotechnology consumerproducts can influence risk and benefit perceptions. RiskAnalysis, 2011; 31(11):1762–1769.

52. Lichtenstein S. Judged frequency of lethal events. Journalof Experimental Psychology: Human Learning and Memory,1978; 4(6):551–578.

53. Facione NC. Perceived risk of berate cancer: Influence ofheuristic thinking. Cancer Practice, 2002; 19(5):256–262.

54. Sunstein CR. Hazardous heuristics. University of ChicagoLaw Review, 2003; 70(2):751–782.

55. Gigerenzer G, Peter M, Todd PM; the ABC Research Group.Simple Heuristics that Make Us Smart. Oxford, UK: OxfordUniversity Press, 1999.

56. Gigerenzer G, Brighton H. Homo heuristics: Why biasedminds make better inferences. Topics Cognitive Science, 2009;1:107–143.

57. Brossard D, Nisbet MC. Deference to scientific authorityamong a low information public: Understanding U.S. opinionon agricultural biotechnology. International Journal of PublicOpinion Research, 2007; 19(1):24–52.

58. Brossard D, Scheufele DA, Kim E, Lewenstein BV. Religios-ity as a perceptual filter: Examining processes of opinion for-mation about nanotechnology. Public Understanding of Sci-ence, 2009; 18(5):546–558.

59. Gaskell G, Ten Eyck T, Jackson J, Velti G. Correspondence:Public attitudes to nanotechnology in Europe and the UnitedStates. Naturematerials, 2004; 3:496. doi:10.1038/nmat1181.

60. Ho SS, Brossard D, Scheufele DA. Effects of value predispo-sitions, mass media use, and knowledge on public attitudes to-ward embryonic stem cell research. International Journal ofPublic Opinion Research, 2008; 20(2):171–192.

61. Kahan DM, Braman D, Slovic P, Gastil J, Cohen G. Culturalcognition of the risks and benefits of nanotechnology. NatureNanotechnology, 2009; 4(2):87–90.

62. Siegrist M, Keller C, Kastenholz H, Frey S, Wiek A. Laypeo-ple’s and experts’ perception of nanotechnology hazards. RiskAnalysis, 2007; 27(1):59–69.

63. Lee CJ, Scheufele DA, Lewenstein BV. Public attitudes to-ward emerging technologies: Examining the interactive effectsof cognitions and affect on public support for nanotechnology.Science Communication, 2005; 27(2):240–267.

64. Besley JC, Shanahan J. Media attention and exposure in rela-tion to support for agricultural biotechnology. Science Com-munication, 2005; 26:347–367.

65. Dosman DM, Adamowicz WL, Hrudey SE. Socioeconomicdeterminants of health- and food safety-related risk percep-tions. Risk Analysis, 2001; 21(2):307–317.

66. Schuman H, Presser S. Questions and Answers in AttitudeSurveys. New York: Academic Press, 1981.

67. Flynn J, Slovic P, Mertz CK. Gender, race, and perception ofenvironmental health risks. Risk Analysis, 1994; 14:1101–1107.

68. Cobb MD, Macoubrie J. Public perceptions about nanotech-nology: Risks, benefits and trust. Journal of Nanoparticle Re-search, 2004; 6(4):395–405.

69. Corley EA, Scheufele DA, Hu Q. Of risks and regulations:How leading U.S. nanoscientists form policy stances aboutnanotechnology. Journal of Nanoparticle Research, 2009;11:1573–1585.

70. Bowman DM, Hodge GA. Nanotechnology: Mapping the wildregulatory frontier. Futures, 2006; 38(9):1060–1073.

71. Besley JC, Kramer VL, Priest SH. Expert opinion onnanotechnology: Risks, benefits, and regulation. Journal ofNanoparticle Research, 2008; 10:549–558.

16 Kim et al.

72. Ho SS, Scheufele DA, Corley EA. Value predispositions, massmedia, and attitudes toward nanotechnology: The interplay ofpublic and experts. Science Communication, 2011; 33(2):167–200.

73. Cohen J, Cohen P. Applied Multiple Regression/CorrelationAnalysis for the Behavioral Sciences, 2nd ed. Hillsdale, NJ:Erlbaum, 1983.

74. Aiken LS, West SG. Multiple Regression: Testing and Inter-preting Interactions. Thousand Oaks, CA: Sage, 1991.

75. Slovic P. The Feeling of Risk: New Perspectives on Risk Per-ception. London: Earthscan, 2010.

76. Slovic P, Finucane ML, Peters E, MacGregor DG. Risk asanalysis and risk as feelings: Some thoughts about affect, rea-son, risk, and rationality. Risk Analysis, 2004; 24(2):311–322.

77. Friedman SM, Egolf BP. Nanotechnology risks and themedia. IEEE Technology and Society Magazine, 2005; 24:5–11.

78. Gaskell G, Ten Eyck T, Jackson J, Velti G. Nanotechnol-ogy: Cultural support for technological innovation in Europeand the United States. Public Understanding of Science, 2005;14:81–90.

79. Currall SC, King EB, Lane N, Madera J, Turner S. Commen-tary: What drives public acceptance of nanotechnology? Na-ture Nanotechnology, 2006; 1(3):153–155.

80. Maynard A. Nanotechnology: A Research Strat-egy for Addressing Risk, 2006. Available at: www.nanotechproject.org/file download/77, Accessed on March 28,2011.