on: 12 march 2014, at: 10:18websites publisher: routledge ... · shark bytes: message sensation...

21
This article was downloaded by: [Michigan State University] On: 12 March 2014, At: 10:18 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Journal of Risk Research Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjrr20 Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinski a , Lindsay Neuberger b , Meredith L. Gore c d , Bret A. Muter c & Brandon Van Der Heide e a Department of Communication , College of Communication Arts & Sciences, Michigan State University , East Lansing , MI , USA b Nicholson School of Communication, University of Central Florida , Orlando , FL , USA c Department of Fisheries & Wildlife , Michigan State University , East Lansing , MI , USA d School of Criminal Justice, Michigan State University , East Lansing , MI , USA e School of Communication, The Ohio State University , Columbus , OH , USA Published online: 18 Jan 2013. To cite this article: Maria Knight Lapinski , Lindsay Neuberger , Meredith L. Gore , Bret A. Muter & Brandon Van Der Heide (2013) Shark bytes: message sensation value and emotional appeals in shark diving websites, Journal of Risk Research, 16:6, 733-751, DOI: 10.1080/13669877.2012.737822 To link to this article: http://dx.doi.org/10.1080/13669877.2012.737822 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

Upload: others

Post on 03-Jul-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

This article was downloaded by: [Michigan State University]On: 12 March 2014, At: 10:18Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Risk ResearchPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/rjrr20

Shark bytes: message sensation valueand emotional appeals in shark divingwebsitesMaria Knight Lapinski a , Lindsay Neuberger b , Meredith L. Gore cd , Bret A. Muter c & Brandon Van Der Heide ea Department of Communication , College of Communication Arts& Sciences, Michigan State University , East Lansing , MI , USAb Nicholson School of Communication, University of CentralFlorida , Orlando , FL , USAc Department of Fisheries & Wildlife , Michigan State University ,East Lansing , MI , USAd School of Criminal Justice, Michigan State University , EastLansing , MI , USAe School of Communication, The Ohio State University ,Columbus , OH , USAPublished online: 18 Jan 2013.

To cite this article: Maria Knight Lapinski , Lindsay Neuberger , Meredith L. Gore , Bret A. Muter &Brandon Van Der Heide (2013) Shark bytes: message sensation value and emotional appeals in sharkdiving websites, Journal of Risk Research, 16:6, 733-751, DOI: 10.1080/13669877.2012.737822

To link to this article: http://dx.doi.org/10.1080/13669877.2012.737822

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

Page 2: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 3: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Shark bytes: message sensation value and emotional appeals inshark diving websites

Maria Knight Lapinskia*, Lindsay Neubergerb, Meredith L. Gorec,d, Bret A. Muterc

and Brandon Van Der Heidee

aDepartment of Communication, College of Communication Arts & Sciences, Michigan StateUniversity, East Lansing, MI, USA; bNicholson School of Communication, University ofCentral Florida, Orlando, FL, USA; cDepartment of Fisheries & Wildlife, Michigan StateUniversity, East Lansing, MI, USA; dSchool of Criminal Justice, Michigan State University,East Lansing, MI, USA; eSchool of Communication, The Ohio State University, Columbus,OH, USA

(Received 27 December 2011; final version received 1 September 2012)

Websites designed to promote risky activities provide a novel context forstudying the role of emotional appeals and message sensation value (MSV) inrisk messages in order to ultimately understand the type of messages that moti-vate people to engage in risk behaviors. Framed in theories of message designand emotion, this study investigates representations of threat, efficacy, and theextent to which risk messages appeal to a range of positive and negative emo-tions through the examination of 53 shark diving websites using content analy-sis and computer generated linguistic analysis. Results indicate that fewwebsites provide explicit threat information (i.e. severity and susceptibility) butmany do present implicit threats. Efficacy-related messages were present on allwebsites. Positive emotion was more common than negative emotion and thereis little representation of the traditional components of MSV. Implications fortheory development and communication about risk-seeking are addressed.

Keywords: risk; message sensation value; website content; wildlife; emotion

‘For the first time, the great white shark is well within the reach of the average person.’

– Shark diving promotional website

Introduction

Risk messages commonly portray risks as something that produces negative affectand which ought to be avoided (Witte 1992). Indeed, ‘risks, by definition, have avery particular negative valence’ (Griffin, Dunwoody, and Neuwirth 1999, S235),especially within the context of choice and decision-making. The extant risk com-munication literature addresses negative dimensions emotion (Peters, Burraston, andMertz 2004), for example Witte and Allen (2000) focused on fear, GriffinDunwoody, and Neuwirth (1999), worry. Research indicates specific negative

*Corresponding author. Email: [email protected]

Journal of Risk Research, 2013Vol. 16, No. 6, 733–751, http://dx.doi.org/10.1080/13669877.2012.737822

� 2013 Taylor & Francis

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 4: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

emotions are predominantly associated with avoidance as opposed to approachbehaviors (Nabi 2002a). Nonetheless, there are cases in which risks are approached(e.g. cliff diving and high-stakes investing; Zaleskiewicz 2001). March and Shapira(1987) discussed the relationship between risk and return in the managerial contextwhere taking risks in the face of adversity is often considered a desirable trait.Although counter to traditional avoidance of risk, there can be positive outcomesassociated with risk-taking, as well as denying or minimizing risk.

Risk communicators are tasked with designing, implementing, and evaluatingmessages designed to foster decisions in support of and opposition to risk. Theknowledge base about these two behavioral outcomes is disproportionate in favorof risk avoidance (McComas 2006). Empirical studies of the approach-orientationof risk-taking behaviors suggest such behaviors are not a function of inaccurate riskperception (Choi et al. 2004) but rather positive emotions resulting from approach-ing a risk. People who are most likely to engage in risky behaviors do so in partbecause of the positive emotions they feel when engaging in the behavior (e.g.high-sensation seekers; Zuckerman 1979; stimulation seekers; Nicholson et al.2005). Moreover, such people are susceptible to certain types of messaging.Specifically, high-sensation seekers are responsive to messages high in messagesensation value (MSV) and focused on the positive emotions associated with risk-taking (Stephenson and Palmgreen 2001).

Beyond understanding the influence of these factors on people susceptible toengaging in risk-taking behavior, however, the knowledge base about communicat-ing risk-promoting messages from an emotional perspective is scant. It is erroneousto assume that risk-promoting messages with emotive dimensions should simplyincorporate antithetical attributes of risk-avoiding messages. Additional insight isneeded about the characteristics of risk-promoting messages with emotive dimen-sions so that communication activities can more accurately fit within decision-mak-ing discourse. Herein, we aim to help fill this void through the examination ofshark-diving messages on websites using emotion-based risk-communicationapproaches as the guiding framework.

Theoretical background

Message design and emotion

Models of emotion and communication are based on the premise that characteristicsof messages are linked with perceptions on the part of individuals (Witte 1992;Nabi 1999). Fundamentally, discrete emotions yield different type of appraisals andbehavioral responses (Lerner and Keltner 2000, 2001). In the case of risk messages,perception of risk is linked to the emotions associated with message content andemotion is seen as a critical motivator of response to risk information (Nabi 1999;Witte 1992). The extended parallel processing model (EPPM) (Witte 1992), forexample, addresses the message and perceptual variables that motivate functionalresponse to risks. It proposes that in order for risk messages to motivate self-protective behaviors, they should contain two main components: threat and efficacy(Witte 1992). The threat component of messages should motivate two beliefs in thepeople who see the message: perceived severity, described as beliefs regarding theseriousness of the threat, and perceived susceptibility, described as beliefs regardingone’s chances of experiencing the threat (Witte 1992). These factors combine to

734 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 5: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

influence the extent to which people perceive they are at risk of a threat and experi-ence the emotion of fear.

People experience risk messages in the context of existing emotional statesthat may be driven, in part, by risk perceptions prior to receiving a message.For example, the utility of fear-based messaging appears to be limited tocases where perception of threat prior to intervention is extremely high.Muthusamy, Levine, and Weber (2009) found that individuals with high levelsof fear did not increase threat perceptions after being presented with a high-threat message. Their results suggest that at some level, there is little potentialto scare individuals who are already sufficiently fearful of a particular threat.Thus, one way to understand the extent to which emotions are likely to beelicited by message content is to first examine whether threats are explicitlydescribed in messages and the extent and nature of severity and susceptibilitymessages.

Although the EPPM is focused on negative emotions, other research in riskcommunication and emotion suggests that perceived risks can be associated withpositive emotions (Zuckerman 1979; Nabi 1999) and these emotions influencedecision-making about risks (Slovic and Peters 2006; Slovic et al. 2005). Emo-tional response to messages has typically focused on general valence (i.e. positiveor negative) as opposed to discrete emotion (e.g. anxiety) and has often neglectedpositive emotion. It is clear, however, that different discrete emotions yield differ-ent behavioral responses (Lerner and Keltner 2001; Nabi 2002b). Reviewing theliterature on negative emotion and attitude change, Breckler (1993) reported fear isthe only discrete negative emotion that has been thoroughly understood throughempirical evidence, a claim echoed by Nabi (2002b). Guilt, anger, disgust, andsadness are understudied (e.g. Shimp and Stuart 2004; Turner 2007). Lerner andKeltner (2001) is an exception. Positive emotions, such as happiness, pride, andrelief, have been examined in the context of persuasive messages, but little isknown about whether they are useful in risk messaging. Happiness is the mostthoroughly examined positive emotion (Lazarus 1991). How other positiveemotions function in messages is not well understood with some exceptions (Kin-nick, Krugman, and Cameron 1996; Igartua, Cheng, and Lopes 2003). Given theimportant role of fear in risk messages, other emotional responses bear additionalscrutiny.

There are at least two ways positive emotion might be elicited through riskmessages; one is through the ‘thrill’ an individual experiences when perceiving a riskand the second is through a sense of ‘empowerment’ as a response to efficacy mes-sages. The efficacy component of a message is conceptualized as a message designedto convince individuals they are able to perform some behavior (i.e. perceived self-efficacy) and that the behavior or the desired outcome of the message (i.e. therecommended response) will avert or reduce the threat (i.e. perceived response effi-cacy; Witte 1992). It may be the case that efficacy perceptions not only work toreduce the fear felt by dealing with the threat (i.e. danger control) but also createpositive emotional response (e.g. hope and excitement) associated with averting thethreat. Thus, the efficacy component of a message may yield positive emotions or atleast minimize negative emotions (Nabi 1999). Similarly, response efficacy, or know-ing something is available to reduce a threat, may provide reassurance and promotepositive emotional response.

Journal of Risk Research 735

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 6: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Risk-taking, sensation seeking, and MSV

Axiomatic in the risk communication literature is the belief that not everyoneresponds to risk information in the same way; individual difference variables areinfluential in this process. People who are high-sensation seekers, for example,might be thrilled by the thought of a severe and probable threat or at least not expe-rience the same level of fear as others (Witte and Morrison 1995). In his seminalwork, Zuckerman (1979) characterized sensation seeking as a personality traitexemplified by an intense need for stimulation and suggested high-sensation seekersdesire intense experiences. High-sensation seekers are attracted to stimuli that arenovel, vivid, and exciting while those who are low-sensation seekers are over-whelmed by this type of information (Donohew, Lorch, and Palmgreen 1991). Assuch, Palmgreen et al. (1991, 219) conceptualized MSV as ‘the degree to which for-mal and content audio-visual features of a … message elicit sensory, affective, andarousal responses.’

Messages with high sensation value may be more likely to utilize intense, gro-tesque, or horrifying images or show people engaged in actions or activities that aredirectly the focus of the message. These message features may evoke arousal in par-ticipants; Stephenson and Palmgreen (2001) differentiated between MSV as a mes-sage attribute as defined above, and MSV as a receiver response. Perceived MSV,they argued, is a receiver’s response to potentially arousing stimulus material. MSVresearch has largely focused on video public service announcements (PSA’s) andthe sensory cues, arousal cues, novelty, and intense language portrayed within thosePSAs. Tailoring MSV to people’s levels of sensation seeking improves reception ofthose messages (Donohew, Lorch, and Palmgreen 1991; Everett and Palmgreen1995). Thus, risk-promoting messages should include audience tailored MSV cuesas well as affective appeals other than fear (e.g. positive emotions such as thrill).

Shark diving as a context for studying emotion in risk communication

Shark attacks are relatively improbable (Pringle 2001) and are akin to other lowprobability risks such as airplane crashes and lightning strikes (Slovic 1987). Evenso, sharks have been portrayed by the media as primitive, scary, human-eating ani-mals that are unintelligent and elusive. Anyone who has seen Jaws or Shark Weekmight think a swim in the ocean is likely to result in a fatal shark attack (Dobson2008). Davey et al. (1998) proposed an evolutionary basis to people’s perception ofthe risks associated with sharks, going back to the time when humans were prey forsome predators.1

Because sharks are perceived as threatening animals (Davey et al. 1998), sharkdiving is generally considered a high-risk behavior. Cage diving with sharks is awildlife tourism event where people, under the guidance of a tour operator, are sub-merged in cage where the water may be chummed (i.e. baited with fish parts andblood) to attract sharks. Shark diving might be considered in a class of risks wherelevels of perceived threat are likely to be extremely high prior to any additionalmessage exposure (like HIV/AIDS; Muthusamy, Levine, and Weber 2009). A multi-million dollar wildlife tourism industry has made it possible for people to observe,feed, and swim with sharks in 40 countries around the world (Topelko and Dearden2005).

Shark diving is a case of premeditated risk-taking; for most people, diving withsharks takes some planning and preparation. This is consistent with other risk behav-

736 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 7: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

iors like hang gliding and elective surgery but unique from spontaneousrisk-taking such as car surfing. One way people can gather information about a risk-related decision prior to engaging in a behavior (e.g. screening for a disease) throughsearching for content on the world-wide web (Williams 2003). The web has becomea critical source of information about risks, making analysis of the content of web-based risk information important (Lapinski 2006). Shark diving is a risk behavior forwhich there is extensive online content devoted to providing information about thisbehavior; much of it is constructed to persuade people to engage in the behavior.

The lack of understanding about emotion in risk-promoting messaging and thecontext of shark diving messages on websites provide an interesting case study forexploring how risk-promotion messages are structured in a real-world context.Based on the lack of research in this area and the novelty of this application, weset three objectives and multiple research questions. First, we aimed to characterizethe extent and nature of severity and susceptibility messages, including whetherthreats are explicitly described. Given that the ultimate goal of shark diving web-sites is to persuade people to dive with sharks, it may be that risk information isminimally portrayed to downplay potential threats. Contrarily, these websites mayexplicitly portray the risks of shark diving to satiate risk-seekers or possibly do sobecause of the potential legal repercussions of not providing such information. Assuch, we posed the following research questions:

RQ1: Are threats (i.e. severity and susceptibility) to humans stated on the shark divingwebsites?

RQ2: How do shark diving websites portray severity of shark-related threats tohumans?

RQ3: How do shark diving websites portray susceptibility of shark-related threats tohumans?

Second, we aimed to explore the extent and nature of efficacy portrayals in thewebsites. Efficacy messages often focus on what a person can do to protect them-selves from a threat. There are other types of recommended responses and actorswho might protect people from threats associated with shark diving. Without bias-ing the analysis toward particular recommended responses, we were interested inexamining all efficacy messages by addressing the following questions:

RQ4: What is the nature of the self-efficacy messages on the websites?RQ5: What is the nature of the response efficacy messages on the websites?

The third objective was to examine the extent to which risk-promoting messagesportray emotion and components of MSV (e.g. arousal cues, sensory cues, novelty,and intense language). Importantly, it may be the case that because risk promotionwas the focus of many of the sites, the content uses positive emotions in ways notcommonly associated with risk messaging. Further, evidence for MSV cues wouldprovide an understanding of how such cues are presented when promoting riskbehaviors. Our corresponding research questions for this objective were:

RQ6: What are the affective cues (positive and negative) portrayed on the websites?RQ7: What are the ways (other than emotional cues) in which message sensation

value is portrayed on the websites?

Journal of Risk Research 737

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 8: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Method

A social scientific content analysis (Fico, Lacy, and Riffe 2008) of shark dive web-sites occurred in two separate phases: first, a group of wildlife and communicationscientists coded and analyzed the websites (hereafter referred to as content analysis)and second, websites were coded using linguistic analysis software (i.e. LinguisticInquiry and Word Count (LIWC); hereafter referred to as linguistic analysis). Thismulti-methodological approach was used to answer the study research questions in away that was designed to be both valid and replicable (Tashakkori and Teddlie 1998).

Sampling

The Internet search engine Google.com was used in a search during a 24-hour per-iod in December 2008 to draw the sample for the study based on McMillian’s(2000) recommendations for sampling Internet-based content.2 The target sampleincluded online portals of organizations providing shark diving opportunities andwas reviewed for the evidence of several criteria designed to limit the sample. Tomeet the criteria for inclusion in the sample, websites had to be active, in the Eng-lish language, and include information about diving with great white (Carcharodoncarcharias), tiger (Galeocerdo cuvier), or bull (Carcharhinus leucas) sharks; speciesmost frequently implicated in attacks on people (International Shark Attack File2009). The sample included only websites that provided an actual opportunity todive with a shark in the ocean; other shark-related pages were excluded.3 Completewebsites (N= 53) were downloaded and printed for analysis to allow for stable andreliable coding (McMillian 2000).

Content analysis

Analysis of the sampled websites relied on two teams of researchers with expertisein risk communication and wildlife, all of whom are authors on this manuscript.Research questions and the theoretical framework discussed above served as thebasis for this analysis and the process occurred in several stages. First, a subsampleof websites (n= 3) was reviewed by all researchers to develop an analysis protocol.Next, several websites (n= 2) were reviewed and evaluated by the team as a wholeto refine the analysis protocol. This refinement ensured that other investigatorswould make similar classifications regardless of their attitudes toward or knowledgeof the topic area (Fico, Lacy, and Riffe 2008). The subsample of websites used inthe preliminary analysis was included in the final analysis.

The team was broken into dyads with one researcher from each discipline. Eachresearcher independently coded approximately 20% of the sample for all the fea-tures in the protocol; there was approximately 20% overlap in sites coded betweenthe dyads. Then, the dyads reviewed each website together and then came to anagreement on the characteristics of each website; instances of all study variableswere counted and recorded by coders. The team met as a whole to review the find-ings of the analysis conducted by each dyad. This process was repeated until nodiscrepancies were found (Wimmer and Dominick 2003). The nature of the codingprotocol precluded calculation of intercoder reliability; to ensure comparable ratingsamong coders, the team had regular calibration meetings to discuss decisions aboutspecific values. Disagreements were discussed until agreement on values by consen-sus was reached (Trochim 2001).

738 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 9: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Homepage characteristics and risk factors (e.g. portrayed severity, susceptibility,diving ease, response efficacy, and implicit and explicit threats portrayed) wereevaluated using previous research as a model for coding these variables (Lapinski2006; Gore et al. 2007; Muter, Gore, and Riley 2009; see Table 1 for constructexplication). Specific dimensions of MSV (e.g. physiological arousal content, sen-sory stimuli and intensity of language) were examined using the framework pro-vided by Morgan and colleagues (2003) and adapted for the website content.Presence or absence and extremity of arousing content (e.g. intense images, video,second-half punch-surprising or novel content below the page break, and intense orsensory text) were analyzed. Specific occurrences of the variables of interest wererecorded for use as examples, and the complete text of all websites was retained forcomputer-based linguistic analysis. The complete coding protocol is available fromthe first author.

Linguistic analysis and emotion coding

Linguistic analyses of the text from the study websites were conducted using thetext analysis software LIWC (Pennebaker, Francis, and Booth 2007). LIWC is a lin-guistic analysis software that categorizes the content of text files according to awell-developed and researched dictionary containing nearly 4500 words and wordstems (see Pennebaker, Francis, and Booth 2007). After a text file (e.g. website con-tent) is given as input, LIWC scans the file for words in its dictionary and reportsthe occurrence rate (i.e. number of words in a specific category divided by totalnumber of words in the file) for every category. For example, if analyzing a text filecontaining the sentence ‘I cried,’ ‘cried’ would be identified as being in the sadnesscategory and the occurrence rate for sadness would be 50%.

Table 1. Theoretical variables and sample messages from study websites; conceptualdefinitions drawn from Witte (1992).

VariableConceptualdefinition Operational definition Sample message

Severity Seriousness of thethreat

Messages promoting theseriousness or danger ofshark diving

‘… row upon row of teethand capable of sensing theblood … they are the mostfeared creature beneath thewaves’

Susceptibility Chances ofexperiencing thethreat

Messages promoting thelikelihood that target willbe vulnerable to sharkthreat

‘Sharks are unpredictableanimals’

Self Efficacy Ability to performa behavior

Messages promotingconsumer ability to engagebehaviors to reduce risk orto engage in shark diving

‘… the great white sharkis well within the reach ofthe average person’

ResponseEfficacy

Ability ofrecommendedresponse to avertor reduce thethreat

Messages promoting theeffectiveness of responsesfor keeping shark diverssafe from threats

‘We have a 100%operational record’

Journal of Risk Research 739

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 10: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

In this study, LIWC’s most recent (2007) dictionary was used to analyze contentin conjunction with a supplementary emotion dictionary created by the researchteam and described below. LIWC, although a well-developed and useful tool, has abroad category scheme for emotion words (e.g. positive and negative emotion). TheLIWC dictionary does not provide an option for more specific coding into discreteemotion categories. This research was interested in a more refined measurement ofspecific textual emotional indicators and thus more precisely defined emotion cate-gories were defined for a supplementary dictionary.

Emotion categorization schemes have been extensively studied and refinedthrough both expert analysis and coding by study participants; the categories definedby Ekman (1992); Lazarus (1991, 1998); and Shaver et al. (1987) were used as abasis to create a supplementary dictionary which resulted in 16 distinct emotion

Table 2. Discreet emotion categories and word examples from LIWC linguistic analysis;conceptual definitions drawn from Ekman (1992), Lazarus (1991, 1998), and Shaver et al.(1987).

Emotion Conceptual definition Sample words

Anger a demeaning offense against me and mine …something interfering with personal goals and personalharm perceived

Harm, hate, abuse

Anxiety uncertain, existential threat … ambiguous danger/threat Apprehensive,nervous, uncertain

Compassion being moved by another’s suffering and wanting tohelp

Caring, considerate,helpful

Disgust taking in or being close to an indigestible object oridea

Gross, horrible,unattractive

Envy wanting what someone else has Jealous, longing,missing

Fear an immediate, concrete, and overwhelming danger …interpretation of events as potentially dangerous orthreatening

Intimidate, scare,phobia

Guilt having transgressed a moral imperative … self-blame Fault, sin, sorryHappiness/Joy

making reasonable progress toward the realization of agoal … positive outcomes … getting somethingdesired or desirable

Content, pleasant,satisfied

Hope possibility of amelioration/success Determined,opportunity, trust

Jealousy resenting a third party ForbidLove desiring or participating in affection, usually but not

necessarily reciprocated – traditionally directed at aspecific person and involving intimacy

Affection, flatter,liking

Pride enhancement of one’s ego-identity by taking credit fora valued object or achievement, either our own or thatof someone or group with whom we identify

Arrogant, courage,success

Relief a distressing goal-incongruent condition that haschanged for the better or gone away

Freeing, relieving,sigh

Sadness discovering that one is powerless, helpless, or impotentto change unhappy circumstances … havingexperienced an irrevocable loss

Depressing,melancholy, sorrow

Shame having failed to live up to ego-ideal Broken, crushed,humiliated

Thrill/Excitement

a sudden wave of keen emotion or excitement – theact of exciting, or the state of being roused into action,or of having increased action; impulsion; agitation

Danger,entertaining,stunned

740 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 11: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

categories (Table 2). After settling on a comprehensive list of discrete emotions,independent coders were trained on the conceptual definitions of the emotion termsand words appearing in LIWC’s affective dictionary were coded into the new emo-tion categories. Intercoder reliability was assessed using Cohen’s Kappa (κ= 0.73;Cohen 1960). Previous research employing LIWC was used as a guide for the analy-sis procedures (see Alpers et al. 2005; Hancock et al. 2007; Toma and Hancock2010). Percentages of the total number of words are used instead of a raw wordcount to control differences in the length of the textual content of each website.Thus, a record was created for each website wherein each category was assigned apercentage of the overall text from the website which pertained to those categories.Mean percentages and standard deviations are reported unless otherwise specified.

Results

General description of the websites

The final sample consisted of 53 websites with over 200 pages of content (contactthe first author for a list of the websites included). LIWC analysis indicated thewebsites varied in the number of words they contained (Min = 1053 Max = 30,415,M= 5891 words, SD= 5989); for the linguistic analyses below, average percentagesare reported unless otherwise noted. Shark diving was the primary focus of themajority of websites (67%). Other websites (33%) included shark diving but alsooffered other activities (e.g. scuba diving and safari). The dive locations included inthe sample were: South Africa (47%), the US–California (15%), Mexico (9%), andAustralia (8%).

Most of the websites (70%) included a photo gallery of sharks or shark diving.Nearly all the websites (96%) provided contact information and included informa-tion about types of dives available including the details of the dive trip (85%). Justover half the websites (51%) included a page about dive staff qualifications andmany had a frequently asked questions page (40%). Thirty-two percent of the web-sites included a conservation page where shark conservation issues were addressed(e.g. ocean pollution, overfishing, and harvesting fins).

Characterizing the extent and nature of messages

The first objective was to characterize the extent and nature of severity and suscep-tibility messages, including whether threats are explicitly described. Our firstresearch question (RQ1) characterized threats to humans portrayed on the websites.Many of the websites included threats to humans associated with the shark divingprocess and were classified as either being specifically from sharks (e.g. potentialinjury) or related to other non-shark threats (e.g. sunburn). The cases in which thethreat of sharks to humans were addressed on the websites were largely implicit.That is, the threat was implied and minimized rather than explicitly stated (e.g.‘frisky animals have turned and nibbled on cages’). Explicit threats (e.g. ‘the mostcommon type of Hawaii shark attack is the so-called “hit and run” assault’) wereless commonly evidenced on the websites. Forty-five percent of the websites statedan implicit threat from sharks while fewer (25%) explicitly stated that sharks pose adirect threat to humans.

In terms of non-shark threats, 40% of the websites stated an explicit threat tohumans with sunburn and seasickness being the most frequent. A number of web-

Journal of Risk Research 741

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 12: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

sites contained implicit non-shark threats (30%) as well. In addition to the contentanalysis, linguistic analysis using LIWC indicated portrayals of threatening wordsin the following categories: anxiety (M= 0.07%, SD= 0.07), death (M= 0.05%,SD= 0.06), and anger (M= 0.21%, SD= 0.16). The mean percentages on the sharkwebsites are relatively low (i.e. significantly below the average emotion portrayal)as compared to other word categories (see Table 3). In sum, both the scientificanalysis and the LIWC indicate that there was little portrayal of explicit threat tohumans from sharks but that implicit threats were more regularly portrayed.

The next research question (RQ2) explored how the severity of the shark risk tohumans was portrayed on the websites. Analysis indicated that the majority of thewebsites (66%) did not textually portray the severity of sharks as a risk to humans.Although most websites did not portray severity, those that did included messageswith varying levels of severity. For example, one website stated ‘Warning: not forthe faint-hearted … a baited tiger shark dive is classified as a high risk dive. Thereason for this is that we are in the water with a shark that is in a feeding situation.’Some websites presented two-sided messages containing severity information cou-pled with a refutation (e.g. ‘… even the elusive oceanic white-tip and tiger sharkslike to come in and raise the adrenaline every so often; sharks often evoke fear (but11 years of feeding without a single incident shows this fear is not justified)’).

The next research question (RQ3) explored how susceptibility was portrayed onthe shark diving websites. Portrayals of susceptibility were largely absent fromthese websites; only 26% of the websites included portrayals of human susceptibil-ity to the risks from sharks. Of those that did, most were largely paired with refuta-tional messages (e.g. ‘humans are not a shark’s first choice for a meal, but why riskbeing mistaken for an elephant seal or giant fish?’). Another website indicated:‘Tiger sharks are responsible for most Hawaii shark attacks on humans. This largeshark species regularly eats human-size prey … Only about three shark attacksoccur per year in Hawaii and few shark attacks are fatal.’ Moreover, susceptibilitymessages were often paired with efficacy messages; a point we will return to later.

The extent and nature of efficacy portrayals

Our second objective focused on assessing the extent and nature of efficacy mes-sages on shark diving websites, specifically the presence and nature of self-efficacy(RQ4) and response efficacy (RQ5). Efficacy can only be considered in the contextof some recommended response. Overwhelming, the recommended response onthese websites was ‘dive with us’ and implied that by diving with a particular divecompany, divers would not have to take steps on their own to reduce the risk butthat they would be ‘safe’ from threats with particular dive operators. That is, mostself-efficacy messages centered primarily on the ease of diving with a particularcompany as opposed to the ease of taking steps to reduce the risk to shark attacks.Indeed, 94% of the websites promoted the ease and simplicity of diving with sharks(e.g. ‘for the first time, the great white shark is well within the reach of the averageperson’ and ‘it’s easy; just add water and we’ll take care of the rest’). Another web-site stated: ‘You don’t have to be qualified or be able to swim. The youngest so farwas 5 and the oldest was 79 and neither cagers [sic] were qualified!’ The websiteslargely did not address efficacy related to threat reduction by individual divers by,for example, suggesting that divers would be able to mitigate threats in the water.

742 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 13: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Table3.

LIW

Canalysisof

emotionwords

presentin

website

content;descriptivestatisticsandt-testsagainstthemeanem

otionpercentage

(0.33%

).

Descriptiv

eStatistics

M(%

)SD

M(w

ords)

SD

tp

Anger

0.0825

0.06961

5.42

6.95

�27.377

<.001

Anxiety

0.0754

0.05992

4.72

5.99

�32.451

<.001

Com

passion

0.3902

0.19951

22.02

22.75

2.371

.021

Disgust

0.0931

0.07193

6.61

8.18

�25.153

<.001

Envy

0.01

0.017

0.29

0.59

�143.172

<.001

Fear

0.0359

0.04796

2.04

3.35

�46.868

<.001

Guilt

0.0042

0.01004

0.27

0.73

�248.266

<.001

Happiness/Joy

1.3724

0.68599

72.79

69.20

11.687

<.001

Hope

0.3297

0.16200

17.63

17.01

.051

.959

Jealousy

0.00

0.003

0.08

0.32

�993.353

<.001

Love

0.3483

0.14511

20.51

20.99

1.044

.301

Pride

1.1334

0.54459

61.34

54.22

11.351

<.001

Relief

0.1903

0.11390

11.65

13.60

�9.323

<.001

Sadness

0.1042

0.09929

7.63

12.08

�17.355

<.001

Shame

0.1005

0.10190

6.99

9.34

�17.192

<.001

Thrill/Excitement

0.9892

0.44207

51.19

42.71

11.478

<.001

Journal of Risk Research 743

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 14: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Response efficacy messages were prevalent on all the websites except one andagain, were focused on the recommended response ‘dive with us’ and addressed thefact that diving with their company would ensure the safety of the dive. Many web-sites provided some basic assurance of safety (e.g. ‘Safety is always a first priority,’and ‘… we have a 100% operational record, not one diver has been injured beyonda sunburn’). Another website stated: ‘(Company Name) is an industry leader in Safeand Sane Shark Diving™. It’s a way of doing shark encounters right, the first timeand every time.’ Messages also dealt with the safety of the boat, cage, and thesafety record of the crew. The websites also promoted heuristics for judging thesafety of the operation such as celebrity appeals (e.g. ‘Used by ESPN and BBC out-doors’ or other named individual celebrities), awards and accreditations, and experi-ence of team (e.g. ‘We are the oldest shark diving operation on the West Coast ofthe US’). In short, the recommended response on most sites was ‘dive with ourcompany,’ the self-efficacy messages promoted the ease of doing so, and theresponse efficacy messages promoted the safety of doing so.

Affective cues and MSV

Our third objective considered the extent to which risk-promoting messages por-trayed emotion and the components of MSV. Research question six asked howaffective cues would be portrayed on the websites and was accomplished largelythrough the LIWC program. The linguistic analysis indicated the average number ofaffect-related words (Min = 48, Max = 1047, M= 291.20, SD= 256.11) varied acrosswebsites. Although both positive and negative affective cues were portrayed on thewebsites, the sites largely focused on positive emotions or mood states. There werea substantive number of words linked with positive emotion (M= 4.59%,SD= 1.32); Table 4 presents data that compare the positive emotion percentagesfound in this study to other sample texts compiled and analyzed by Pennebaker,Francis, and Booth (2007) as a point of comparison. Specifically, dimensions ofhappiness (M= 1.37%, SD= 0.69), thrill/excitement (M = 0.99%, SD= 0.44), andpride (M = 1.13%, SD= 0.54) were most frequently represented on the websites.

Words displaying negative emotions were less common (M= 0.54%, SD= 0.32;Table 4). Specifically, sadness (M= 0.10%, SD= 0.10), shame (M= 0.10%,SD= 0.10), and disgust (M= 0.09%, SD= 0.07) were the most often portrayed nega-tive emotions although these occurred at lower rates than words portraying positiveemotions. These negative emotions were not portrayed on all websites as the rangewas between 0% and 0.82% on any individual site. Interestingly, words associatedwith fear were rarely present on websites (M= 0.04%, SD= 0.05). The mean per-centage for each individual emotion was tested against the overall mean percentageacross emotions to indicate whether each individual emotion was portrayed more orless often than the average (Table 3). All positive emotions (e.g. happiness andlove) were present at levels above the mean and all negative emotions (e.g. angerand sadness) were present at levels below the mean; most differences are statisti-cally significant (see Table 3). The data suggest positive emotions are portrayedmuch more frequently than negative emotions and specifically that fear was rarelyportrayed on the shark diving websites.

A final research question (RQ7) asked what indicators of MSV would be presenton shark diving websites. The content analysis indicated arousal cues on websitesincluded pictures with both intense (66% of sites) and passive (72%) images. The

744 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 15: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

focus of these pictures was most often of sharks breaching (i.e. jumping out of thewater), sharks with open mouths and with teeth exposed (both coded as intense),and sharks swimming with a closed mouth (coded as passive). Most websites(79%) included pictures of people in the process of cage diving. Additionally, 75%of the websites contained pictures of people co-located with sharks, but the majorityof the visual images on the websites (53%) did not focus on humans. Beyond pho-tographs, the sensory cues on the websites were rather limited. Only three of thewebsites played music and 28% of the websites provided video. The indicator ofnovelty in the websites, novel or surprising information below the initial screen ofthe webpage (i.e. second-half punch) was only evidenced on one website.

LIWC linguistic analysis of the extent to which sensory cues were representedthrough textual content on the websites indicated words dealing with sight(M= 1.68%, SD= 0.74), hearing M= .10%, SD= 0.09), and touch (M = .36%,SD= 0.18) were relatively infrequent but not inconsistent with other texts (Table 4).The predominance of indicators of sight does confirm the photograph-rich findingsfrom the content analysis. LIWC was also used to determine the existence ofintense language on the websites. Intense language was measured in LIWC throughthe presence of profanity, verbs, other motion words, and adverbs. Although profan-ity was not often present on websites (M= 0.004%, SD= 0.01), verbs (M= 8.09%,SD= 1.72), adverbs (M = 3.54%, SD= 0.67), and other motion words (M= 2.30%,SD= 0.57) were often present on the websites analyzed. In sum, this analysis indi-cates that the MSV cues were largely present in the form of visual images of sharksand use of verbs and adverbs as a form of intense language.

Discussion

Risk-promoting messages with emotive dimensions are an important, but understud-ied, component of risk communication because risk communicators are oftencharged with designing messages to inhibit or decrease risk-taking behaviors. Thisresearch has implications beyond this study because research suggests it may beprofitable to promote certain risky behaviors to help people avoid others (Donohew,Lorch, and Palmgreen 1991); moreover, understanding risk promotion can informthe design of risk-prevention messages. Data herein contribute unique insight into

Table 4. Mean emotion and MSV presence on shark websites in this study as compared toother types of writing as analyzed by LIWC (see Pennebaker et al. 2007).

Positiveemotion

Negativeemotion Sight Hearing Touch Profanity Verbs Adverbs

Sharkwebsites

4.59 0.54 1.68 0.10 0.36 0.004 8.09 3.54

LIWC textaverage

2.74 1.63 0.86 0.56 0.61 0.15 13.93 4.59

Emotionalwriting

3.28 2.67 0.53 0.44 0.96 0.11 17.44 6.29

Controlwriting

1.83 0.71 0.83 0.35 0.62 0.03 13.59 4.48

Sciencearticles

1.33 0.84 0.65 0.06 0.24 0.00 4.98 1.35

Blogs 3.72 2.07 0.87 0.65 0.60 0.33 14.61 5.46

Journal of Risk Research 745

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 16: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

risk-promoting messages with emotive content and thus more effective risk commu-nication in several key ways.

One noteworthy finding from this study is that risks from sharks to humans areinfrequently explicitly addressed on study websites; when risks were mentioned,they were framed as mild and slight (e.g. sharks nibbling on cages), not severe (e.g.death and dismemberment). The focus on implicit as opposed to explicit risks isone that is not commonly seen in the risk communication literature on risk reduc-tion. It is also striking given the high levels of threat attached to and from sharks(Philpott 2002). Are website designers capitalizing on the apparent evolutionaryhuman fear response to sharks? Are they relying on cultural stereotypes of sharksas threatening animals rather than address threats through explicit threat portrayalsor severity messages? These are ultimately empirical questions, but answers mayrest upon the relative and experiential basis of risk perceptions (Weinstein 1989)and how these bases can be accounted for in risk messages.

Perhaps, site designers are attempting to mediate behavioral responses to getclose to but not cross a ‘tipping point,’ such is the case for fish consumptionadvisories that warn consumers about health risks from overconsumption of certainfish species while concomitantly advising about health benefits from general fishconsumption (Connelly and Knuth 1998). This notion of a risk message tippingpoint seems particularly salient to this case study given the business dimension ofshark dive operations. Dive operators cannot take too much of the risk out of sharkdiving or else they will alienate sensation seekers (one potential market), however,they do need to get as many people as possible to dive with them in order to beprofitable.

Our data show website designers navigated this tipping point by using fewseverity or susceptibility messages; when severity messages were used, they weretwo-sided refutational messages (Allen 1991) that briefly acknowledged severityand then immediately refuted the scary things about sharks. Study websites illus-trated a pattern of refuting the severity associated with shark diving by offeringother information about risk (e.g. people consider sharks the most feared animal ofthe sea, but really they are largely docile creatures). Following the assumption thata messenger cannot or should not message to scare the already scared, messagedesigners appear to be capitalizing on the inherent fear evoked by sharks and alsomoderating fear through messages about efficacy and positive emotions.

An additional noteworthy point from this research is that susceptibility to riskwas rarely portrayed. Similar to severity messages, when susceptibility was por-trayed it was commonly accompanied by response efficacy messages or refutationalmessages (i.e. sharks sometimes attack humans, but we have an excellent safetyrecord). It is reasonable that a degree of risk susceptibility is necessary to completeengagement in the risk-approach behavior for sensation seekers. Risk susceptibilitymessages and perceptions are tempered for non-sensation seekers through provisionof efficacy information. Our analysis indicates there may be a ‘susceptibility thresh-old’ for risk-promoting messages, whereby messages portray a behavior as thrilling,but not life threatening. Future research should further examine this notion.

Self-and response efficacy messages considered in the light of the most commonrecommended response on the websites ‘dive with us,’ were clearly stated in thewebsite content for nearly every site. None of the sites exhibited messages aboutwhat individual divers could do to reduce the threats posed from sharks. The effi-cacy messages in general were centered around things that the shark dive outfitter

746 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 17: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

has done or will do to make the dive easy and safe. This is different from efficacymessages seen in risk-communication campaigns in which the onus of responsibilityis placed largely on the person who is at risk (see Witte et al. 1998) and is verymuch in line with the service orientation of these websites. It can also provide a dif-ferent mechanism for promoting efficacy for risk reduction; in some cases, risk mes-sages which place the responsibility for safety outside of the individual may makesense (for example, with environmental risks).

Results indicated the clear use of affective appeals across websites; theseappeals were largely centered on positive as opposed to negative emotions. Themost common appeals were content that promoted happiness, thrill, pride (in doingsomething unique), or excitement associated with the dive experience. Study web-sites appear to be designed to appeal to the excitement of a human-apex predatorinteraction. The nature of these portrayals capitalizes on inherent qualities of therisk to appeal to receivers and it is consistent with conceptualizations of sensation-seeking (e.g. Zuckerman 1979, 1994). What was noteworthy was the broad appealfor love or compassion; for example, several of the websites suggested divers wouldexperience ‘bonding’ with the sharks and enhanced this through anthropomorphiz-ing (e.g. assigning human names) sharks, providing people with information on theindividual ‘personality traits’ of individual animals, or encounter histories. LIWCrevealed that positive emotions like happiness and love were portrayed much morefrequently than negative emotions such as anger and fear; although in general, therelative number of words for any emotion was low.

The use of positive emotion on the websites adds to the body of literature whichindicates emotions are an important part of the risk communication processalthough most current risk message theories focus on the effects of negative emo-tions (e.g. fear; Nabi 2002a). Investigating how risk communication informationand messages can capitalize on positive emotion associated with both risk avoid-ance and approach behaviors may contribute to the improved efficiency, and refine-ment of prevention campaigns.

Finally, the third objective of this study was undertaken with the assumptionthat people who decide to shark dive are likely to be high-sensation seekers(Palmgreen et al. 1995) and this drove us to explore the extent to which MSV isportrayed on the sites. Results indicate that the most well-represented componentsof MSV were portrayed through imagery of sharks and the use of intense lan-guage including verbs and adverbs. To the extent that affective words yield arou-sal, these words were well represented on the sites. The other components ofMSV (e.g. intense music, video, and second-half punch) were not well representedon the websites. This lack of message sensation cues present on shark diving web-sites may represent an intuitive appeal to a larger, more general audience asopposed to only sensation-seekers. It may be a function of the implicit reliance onthe arousal likely to be elicited by the images people have of sharks – somethingthat is easily cued by photographs. Importantly, there is limited information aboutthe role of visual images in driving risk perception and decision-making. Giventhat this study has characterized the nature of the content on these sites, additionalexperimental research might examine the extent to which these messages appeal tosensation-seekers can be linked to risky decisions particularly for sensationseekers.

Results and implications herein should be prudently interpreted. The nature ofour content analysis procedure precluded measures of intercoder reliability

Journal of Risk Research 747

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 18: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

(Wimmer and Dominick 2003) and this should be considered a caveat to theinterpretation of the findings. The method of analysis was chosen strategicallybecause of the contextualized nature of the research questions making unitization ofcontent reductionistic and not useful. LIWC proved to be an important tool for tri-angulating findings from the social scientific content analysis, however, the limitedexisting LIWC dictionary required the post hoc addition of the supplemental emo-tion dictionary. The supplemental dictionary required complex synthesis of theexisting emotion conceptual schemes and resulted relatively modest estimates ofintercoder reliability. This study has done much of the ‘heavy lifting’ for researcherswishing to use LIWC by creating an emotion word dictionary for coding emotionalmessage content; additional research can work to refine this approach. We agreewith Shaver et al. (1987) about the utility of linguistic analysis for communicationresearch yet acknowledge the method can be further improved to minimize threatsto validity and reliability. Thus, although results should not be generalized beyondthe study sample, they do offer a baseline upon which researchers may sourcefuture research refining, revising, and improving the knowledge base about risk-pro-moting messages with an emotive dimension. The other study results have implica-tions for risk message design and MSV beyond the scope of the context of sharkdiving. Other risk-taking, which involves higher levels of actual risk might be con-sidered in light of this study. It may be that promotion of truly high-risk activitiessuch as mountaineering or sky-diving must present an accurate accounting of therisks on their websites for liability purposes, but the use of positive emotion andperhaps sensation-seeking cues may be similar to those seen in shark dive sites.

Positive outcomes associated with risk-taking receive considerably less attentionfrom the risk communication community than do negative outcomes with risk-tak-ing, perhaps because of a lack of prosocial implications. Improved messaging aboutrisk-promoting behavior with emotive dimensions may help contribute to moreeffective decision-making about risks. This empirical investigation of risk messagedesign, MSV, and emotional appeals in shark diving helps to advance a sparseknowledge base. With additional inquiry, risk scholars and practitioners may cometo better understand the complex relationship between positive emotion and riskperception.

Notes1. The International Shark Attack file notes that the number of shark attacks in any given

year or region is highly influenced by the number of people entering the water. The oddsof an unprovoked shark attack is estimated to be 1 in 11.5 million; the odds of anunprovoked shark attack fatality is 0 is 264.1 million. Shark diving may constitute aprovoked encounter depending on how the dive operator chooses to facilitate thehuman-shark interaction. http://www.flmnh.ufl.edu/fish/sharks/ISAF/ISAF.htm (Accessed10/22/2012).

2. The sample was limited to the American version of the Google search engine becauseof the tourism focus of the research. This is appropriate because the current study isfocused on individuals who are using the websites to aid in their decision making aboutpotential shark diving. Redoing the sampling procedures in October 2011 revealed 49 ofthe 53 sites were still active.

3. Actual search terms were: shark AND ‘tiger OR white OR bull’ AND ‘dive OR divingOR swim OR cage dive OR cage diving OR tour OR trip OR vacation OR operator ORtourism OR excursion OR holiday’.

748 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 19: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

ReferencesAllen, M. 1991. Meta-analysis comparing effectiveness of one and two-sided messages. Wes-

tern Journal of Speech Communication 55: 390–404.Alpers, G.W., A.J. Winzelberg, C. Classen, H. Roberts, P. Dev, C. Koopman, and C.B. Tay-

lor. 2005. Evaluation of computerized text analysis in an Internet breast cancer supportgroup. Computers in Human Behavior 21: 361–76.

Breckler, S.J. 1993. Emotion and attitude change. In Handbook of Emotions, ed. M. Lewisand J.M. Haviland, 461–73. New York, NY: Guilford Press.

Choi, Y., G.T. Cameron, G. Leshner, and M.T. Stephenson. 2004. Sensation seeking target-ing and fear appeal of anti-smoking public service announcement messages for youngadults. Paper presented at the annual meeting of the Association for Education in Jour-nalism and Mass Communication, August 4–7, in Toronto, Canada.

Cohen, J. 1960. A coefficient of agreement for nominal scales. Education and PsychologicalMeasurement 20, no. 1: 37–46.

Connelly, N.A., and B.A. Knuth. 1998. Evaluating risk communication: Examining targetaudience perceptions about four presentation formats for fish consumption health advi-sory information. Risk Analysis 18, no. 5: 649–59.

Davey, G.L., A.S. McDonald, U. Hirisave, G.G. Prabhu, S. Iwawaki, C.I. Jim, H. Merckel-bach, P.J. de Jong, P.W. Lejny, and L. Reimann. 1998. A cross-national study of animalfears. Behaviour Research and Therapy 36: 735–50.

Dobson, J. 2008. Shark! A new frontier in tourist demand for marine wildlife. In MarineWildlife and Tourism, ed. J. Hingham and M. Luck, 49–65. Cambridge, MA: CABInternational.

Donohew, L., E. Lorch, and P. Palmgreen. 1991. Sensation seeking and targeting of televisedanti-drug PSAs. In Persuasive communication and drug abuse prevention, ed. L.Donohew, H.E. Sypher, and W.J. Bukoski, 209–26. Hillsdale, NJ: Lawrence Erlbaum.

Ekman, P. 1992. An argument for basic emotions. Cognition and Emotion 6: 169–200.Everett, M.W., and P. Palmgreen. 1995. Influences of sensation seeking, message sensation

value, and program context on effectiveness of anticocaine public serviceannouncements. Health Communication 7, no. 3: 225–48.

Fico, F.G., S. Lacy, and D. Riffe. 2008. A content analysis guide for media economicsscholars. Journal of Media Economics 21, no. 2: 114–30.

Gore, M.L., B.A. Knuth, P.D. Curtis, and J.E. Shanahan. 2007. Factors influencing risk per-ception associated with human-black bear conflict. Human Dimensions of Wildlife 12,no. 2: 133–6.

Griffin, R.J., S. Dunwoody, and K. Neuwirth. 1999. Proposed model of the relationship ofrisk information seeking and processing to the development of preventive behaviors.Environmental Research 80, no. 2: S230–245.

Hancock, J.T., L.E. Curry, S. Goorha, and M. Woodworth. 2007. On lying and being lied to:A linguistic analysis of deception in computer-mediated communication. DiscourseProcesses 45: 1–23.

Igartua, J.J., L. Cheng, and O. Lopes. 2003. To think or not to think: Two pathways towardpersuasion by short films on AIDS prevention. Journal of Health Communication 8:513–28.

International Shark Attack File. 2009. http://www.flmnh.ufl.edu/fish/sharks/isaf/isaf.htm.Kinnick, K.N., D.M. Krugman, and G.T. Cameron. 1996. Compassion fatigue: Communica-

tion and burnout toward social problems. Journalism and Mass CommunicationQuarterly 73: 687–707.

Lapinski, M.K. 2006. StarvingforPerfect.com: A theoretically-based content analysis of pro-eating disorder websites. Health Communication 20, no. 3: 243–53.

Lazarus, R.S. 1991. Emotion and adaptation. New York, NY: Oxford Press.Lazarus, R.S. 1998. Emotion and adaptation. In Human emotions: A reader, ed. J. Jenkins,

K. Oatley, and N.L. Stein, 38–44. Hoboken, NJ: Wiley-Blackwell.Lerner, J.S., and D. Keltner. 2000. Beyond valence: Toward a model of emotion-specific

influences on judgment and choice. Cognition and Emotion 14: 473–93.Lerner, J.S., and D. Keltner. 2001. Fear, anger, and risk. Journal of Personality and Social

Psychology 81, no. 1: 146–59.

Journal of Risk Research 749

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 20: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

March, J.G., and Z. Shapira. 1987. Managerial perspectives on risk and risk taking.Management Science 33, no. 11: 1404–18.

McComas, K.A. Defining moments in risk communication research: 1996–2005. Journal ofHealth Communication 11, no. 1: 75–91.

McMillian, S.J. 2000. The microscope and the moving target: The challenge of applyingcontent analysis to the world wide web. Journalism & Mass Communication Quarterly77, no. 1: 80–98.

Morgan, S.E., P. Palmgreen, M.T. Stephenson, R.H. Hoyle, and E.P. Lorch. 2003. Associa-tions between message features and subjective evaluations of the sensation value of anti-drug public service announcements. Journal of Communication 53, no. 3: 512–26.

Muter, B.A., M.L. Gore, and S.J. Riley. 2009. From victim to perpetrator: Evolution of riskframes related to human-cormorant conflict in the Great Lakes. Human Dimensions ofWildlife 14, no. 5: 366–79.

Muthusamy, N., T.R. Levine, and R. Weber. 2009. Scaring the already scared: Some prob-lems with HIV/AIDS fear appeals in Namibia. Journal of Communication 59, no. 2:317–44.

Nabi, R.L. 1999. A Cognitive-Functional Model for the effects of discrete negative emotionson information processing, attitude change, and recall. Communication Theory 9, no. 3:292–320.

Nabi, R.L. 2002a. Anger, fear, uncertainty, and attitudes: A test of the cognitive-functionalmodel. Communication Monographs 69, no. 3: 204–16.

Nabi, R.L. 2002b. Discrete emotions and persuasion. In The persuasion handbook, ed. J.P.Dillard and M. Pfau, 289–308. Thousand Oaks, CA: Sage.

Nicholson, N., E. Soane, M. Fenton-O’Creevy, and P. Willman. 2005. Personality anddomain specific risk taking. Journal of Risk Research 8, no. 2: 157–76.

Palmgreen, P., L. Donohew, E.P. Lorch, M. Rogus, D. Helm, and N. Grant. 1991. Sensationseeking, message sensation value, and drug use as mediators of PSA effectiveness.Health Communication 3, no. 4: 217–27.

Palmgreen, P., E.P. Lorch, L. Donohew, N.G. Harrington, M. D’Silva, and D. Helm. 1995.Reaching at-risk populations in a mass media drug abuse prevention campaign: Sensa-tion seeking as a targeting variable. Drugs and Society 8: 29–45.

Pennebaker, J.W., M.E. Francis, and R.J. Booth. 2007. Linguistic Inquiry and Word Count(LIWC): LIWC2007. Mahwah, NJ: Lawrence Erlbaum Associates.

Peters, E.M., B. Burraston, and C.K. Mertz. 2004. An emotion-based model of risk percep-tion and stigma susceptibility: Cognitive appraisals of emotion, affective reactivity,worldviews, and risk perceptions in the generation of technological stigma. Risk Analysis24: 1349–67.

Philpott, R. 2002. Why sharks may have nothing to fear than fear itself: An analysis of theeffect of human attitudes on the conservation of the white shark. Colorado Journal ofInternational Environmental Law and Policy 13, no. 2: 445–72.

Pringle, L. 2001. Sharks: Strange and wonderful. Honesdale, PA: Boyds Mills Press.Shaver, P., J. Schwartz, D. Kirson, and C. O’Connor. 1987. Emotion knowledge: Further

exploration of a prototype approach. Journal of Personality and Social Psychology 52,no. 6: 1061–86.

Shimp, T.A., and E.W. Stuart. 2004. The role of disgust as an emotional mediator of adver-tising effects. Journal of Advertising 33: 43–53.

Slovic, P. 1987. Perception of risk. Science 236, no. 4799: 280–5.Slovic, P., and E. Peters. 2006. Risk perception and affect. Current Directions in Psychologi-

cal Science 15: 322–5.Slovic, P., E. Peters, M.L. Finucane, and D.G. MacGregor. 2005. Affect, risk, and decision

making. Health Psychology 24: S35–40.Stephenson, M.T., and P. Palmgreen. 2001. Sensation seeking, perceived message sensation

value, personal involvement, and processing of anti-marijuana PSAs. CommunicationMonographs 68: 49–71.

Tashakkori, A., and C. Teddlie. 1998. Mixed methodology: Combining qualitative and quan-titative approaches. Thousand Oaks, CA: Sage Publications.

750 M.K. Lapinski et al.

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014

Page 21: On: 12 March 2014, At: 10:18websites Publisher: Routledge ... · Shark bytes: message sensation value and emotional appeals in shark diving websites Maria Knight Lapinskia*, Lindsay

Toma, C.L., and J.T. Hancock. 2010. Reading between the lines: Linguistic cues to deceptionin online dating profiles. Paper presented at the annual meeting of Computer SupportedCooperative Work, February 6–10, in Savannah, Georgia.

Topelko, K., and P. Dearden. 2005. The shark watching industry and its potential contribu-tion to shark conservation. Journal of Ecotourism 4, no. 2: 108–28.

Trochim, W.M.K. 2001. The research methods knowledge base. Ithaca, NY: Atomic Dog.Turner, M.M. 2007. Using emotion in risk communication: The anger-activism model. Pub-

lic Relations Review 33: 114–9.Weinstein, N.D. 1989. Effects of personal experience on self-protective behavior. Psycholog-

ical Bulletin 105, no. 1: 31–50.Williams, C. 2003. New technologies in self-help: Another effective way to get better?

European Eating Disorders Review 11, no. 3: 170–82.Wimmer, R.D., and J.R. Dominick. 2003. Mass media research: An instruction manual.

Belmont, CA: Wadsworth/Thompson Learning.Witte, K. 1992. Putting the fear back into fear appeals: The extended parallel process model.

Communication Monographs 59, no. 4: 329–49.Witte, K., and M. Allen. 2000. A meta-analysis of fear appeals: Implications for effective

public health campaigns. Health Education & Behavior 27, no. 5: 591–615.Witte, K., K.A. Cameron, M.K. Lapinski, and S. Nzyuko. 1998. A theoretically-based evalu-

ation of HIV/AIDS prevention campaigns along the Trans-Africa highway in Kenya.Journal of Health Communication 3: 345–63.

Witte, K., and K. Morrison. 1995. The use of scare tactics in AIDS prevention: The case ofjuvenile detention and high school youth. Journal of Applied Communication Research23: 128–42.

Zaleskiewicz, T. 2001. Beyond risk seeking and risk aversion: Personality and the dual nat-ure of economic risk taking. European Journal of Personality 15, no. 1: S105–122.

Zuckerman, M. 1979. Sensation seeking: Beyond the optimal level of arousal. Mahwah, NJ:Lawrence Erlbaum Associates.

Zuckerman, M. 1994. Behavioral expression and biosocial bases of sensation seeking. NewYork, NY: Cambridge University Press.

Journal of Risk Research 751

Dow

nloa

ded

by [

Mic

higa

n St

ate

Uni

vers

ity]

at 1

0:18

12

Mar

ch 2

014