• American Meteorological Society, 2017: Ad hoc committee on naming winter storms. Accessed 1 June 2018, https://www.ametsoc.org/cwwce/index.cfm/committees/ad-hoc-committee-on-naming-winter-storms/.

  • Anderson, J., and Coauthors, 2016: Far far away in Far Rockaway: Responses to risks and impacts during Hurricane Sandy through first-person social media narratives. Proc. 13th Int. Conf. on Information Systems for Crisis Response and Management, Rio de Janeiro, Brazil, ISCRAM, 16 pp., http://idl.iscram.org/files/jenningsanderson/2016/1388_JenningsAnderson_etal2016.pdf.

  • Baker, E. J., K. Broad, J. Czajkowski, R. Meyer, and B. Orlove, 2012: Risk perceptions and preparedness among mid-Atlantic residents in advance of Hurricane Sandy: Preliminary report. Risk Management and Decision Processes Center, Wharton School, University of Pennsylvania Working Paper 2012-18, 42 pp., http://opim.wharton.upenn.edu/risk/library/WP2012-18_EJB-etal_RiskPerceptions-Sandy.pdf.

  • Barthel, M., and E. Shearer, 2015: How do Americans use Twitter for news? Pew Research Center, accessed 1 December 2015, http://www.pewresearch.org/fact-tank/2015/08/19/how-do-americans-use-Twitter-for-news/.

  • Burke, J. A., P. R. Spence, and K. A. Lachlan, 2009: Gender and age differences in use and perceptions of emergency messages during Hurricane Ike. La. Commun. J., 11, 95109.

    • Search Google Scholar
    • Export Citation
  • Burke, J. A., P. R. Spence, and K. A. Lachlan, 2010: Crisis preparation, media use, and information seeking during Hurricane Ike: Lessons learned for emergency communication. J. Emerg. Manage., 8, 2737, https://doi.org/10.5055/jem.2010.0030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DHS, 2013: Lessons learned: Social media and Hurricane Sandy. Virtual Social Media Working Group and DHS First Responders Group Rep., Department of Homeland Security, 39 pp., www.dhs.gov/publication/lessons-learned-social-media-hurricane-sandy.

  • Dixit, V., C. Wilmot, and B. Wolshon, 2012: Modeling risk attitudes in evacuation departure choices. Transp. Res. Rec., 2312, 159163, https://doi.org/10.3141/2312-17.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Earle, P., 2010: Earthquake Twitter. Nat. Geosci., 3, 221222, https://doi.org/10.1038/ngeo832.

  • Edwards, C., P. R. Spence, C. J. Gentile, A. Edwards, and A. Edwards, 2013: How much Klout do you have…A test of system generated cues on source credibility. Comput. Human Behav., 29, A12A16, https://doi.org/10.1016/j.chb.2012.12.034.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fishbein, M., and I. Ajzen, 2010: Predicting and Changing Behavior: The Reasoned Action Approach. Psychology Press, 538 pp.

    • Crossref
    • Export Citation
  • Fritz, A., 2015: Weather Channel unveils 2015–2016 winter storm names, including Mars, Regis and Yolo. Washington Post, 13 October, https://www.washingtonpost.com/news/capital-weather-gang/wp/2015/10/13/weather-channel-announces-new-winter-storm-names-including-mars-regis-and-yolo/.

  • Gibson, R., and D. Zillmann, 1998: Effects of citation in exemplifying testimony on issue perception. J. Mass Commun. Quart., 75, 167176, https://doi.org/10.1177/107769909807500116.

    • Search Google Scholar
    • Export Citation
  • Gibson, R., and D. Zillmann, 2000: Reading between the photographs: The influence of incidental pictorial information on issue perception. J. Mass Commun. Quart., 77, 355366, https://doi.org/10.1177/107769900007700209.

    • Search Google Scholar
    • Export Citation
  • Gibson, R., C. Callison, and D. Zillmann, 2011: Quantitative literacy and affective reactivity in processing statistical information and case histories in the news. Media Psychol., 14, 96120, https://doi.org/10.1080/15213269.2010.547830.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Goble, H., A. J. Beattie, and C. Edwards, 2016: The impact of twitterbot race on interpersonal impressions. Iowa J. Commun., 48, 2335.

  • Goddard, A., P. R. Spence, and K. A. Lachlan, 2012: Black beaches and bayous: The BP Deep Water Horizon spill disaster. Hazard, Outrage and Locality: An Analysis of Two Oil Spills, L. A. Eargle and A. M. Esmail, Eds., University Press of America, 249–261.

  • Graham, M. W., E. J. Avery, and S. Park, 2015: The role of social media in local government crisis communications. Public Relat. Rev., 41, 386394, https://doi.org/10.1016/j.pubrev.2015.02.001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hastall, M. R., and S. Knobloch-Westerwick, 2013: Severity, efficacy, and evidence type as determinants of health message exposure. Health Commun., 28, 378388, https://doi.org/10.1080/10410236.2012.690175.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kietzmann, J. H., K. Hermkens, I. P. McCarthy, and B. S. Silvestre, 2011: Social media? Get serious! Understanding the functional building blocks of social media. Bus. Horiz., 54, 241251, https://doi.org/10.1016/j.bushor.2011.01.005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2007: Hazard and outrage: Developing a psychometric instrument in the aftermath of Katrina. J. Appl. Commun. Res., 35, 109123, https://doi.org/10.1080/00909880601065847.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2009: Emergency communication: A framework for planning and targeting messages. J. Emerg. Manag., 7, 6972, https://doi.org/10.5055/jem.2009.0033.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2010: Communicating risks: Examining hazard and outrage in multiple contexts. Risk Anal., 30, 18721886, https://doi.org/10.1111/j.1539-6924.2010.01511.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, K. M. Najarian, and M. Del Greco, 2014a: Twitter use during a weather event: Comparing content associated with localized and nonlocalized hashtags. Commun. Stud., 65, 519534, https://doi.org/10.1080/10510974.2014.956940.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, and M. Del Greco, 2014b: Screaming into the wind: Examining the volume and content of tweets associated with Hurricane Sandy. Commun. Stud., 65, 500518, https://doi.org/10.1080/10510974.2014.956941.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, K. Najarian, and M. Del Greco, 2016: Social media and crisis management: CERC, search strategies, and Twitter content. Comput. Human Behav., 54, 647652, https://doi.org/10.1016/j.chb.2015.05.027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, and X. Lin, 2017: Natural disasters, Twitter, and stakeholder communication: What we know and directions for future research. Social Media and Crisis Communication, L. Austin and J. Yan, Eds., Routledge, 296–305.

    • Crossref
    • Export Citation
  • Lin, X., K. A. Lachlan, and P. R. Spence, 2016a: Exploring extreme events on social media: A comparison of user reposting/retweeting behaviors on Twitter and Weibo. Comput. Human Behav., 65, 576581, https://doi.org/10.1016/j.chb.2016.04.032.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, X., P. R. Spence, T. L. Sellnow, and K. A. Lachlan, 2016b: Crisis communication, learning and responding: Best practices in social media. Comput. Human Behav., 65, 601605, https://doi.org/10.1016/j.chb.2016.05.080.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, X., P. R. Spence, and K. A. Lachlan, 2016c: Social media and credibility indicators: The effect of influence cues. Comput. Human Behav., 63, 264271, https://doi.org/10.1016/j.chb.2016.05.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Martin, I. M., H. Bender, and C. Raish, 2007: What motivates individuals to protect themselves from risks: The case of wildland fires. Risk Anal., 27, 887900, https://doi.org/10.1111/j.1539-6924.2007.00930.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mileti, D., 1999: Disasters by Design: A Reassessment of Natural Hazards in the United States. Joseph Henry Press, 371 pp., https://doi.org/10.17226/5782.

    • Search Google Scholar
    • Export Citation
  • Miller, C. H., B. J. Adame, and S. D. Moore, 2013: Vested interest theory and disaster preparedness. Disasters, 37, 127, https://doi.org/10.1111/j.1467-7717.2012.01290.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morss, R. E., and Coauthors, 2017: Hazardous weather prediction and communication in the modern information environment. Bull. Amer. Meteor. Soc., 98, 26532674, https://doi.org/10.1175/BAMS-D-16-0058.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • NOAA, 2017: Tropical cyclone names. NOAA/National Hurricane Center, accessed 30 May 2017, http://www.nhc.noaa.gov/aboutnames.shtml.

  • Palmer, R., 2013: What’s in a storm name? Weather Channel policy draws critics, but catches on. International Business Times, 8 February, http://www.ibtimes.com/whats-storm-name-weather-channel-policy-draws-critics-catches-1073016.

  • Rainear, A. M., K. A. Lachlan, and C. A. Lin, 2017: What’s in a #name? An experimental study examining perceived credibility and impact of winter storm names. Wea. Climate Soc., 9, 815822, https://doi.org/10.1175/WCAS-D-16-0037.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reynolds, B., and M. W. Seeger, 2005: Crisis and emergency risk communication as an integrative model. J. Health Commun., 10, 4355, https://doi.org/10.1080/10810730590904571.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sandman, P. M., 2003: Four kinds of risk communication. The Synergist, April 2003, American Industrial Hygiene Association, 26–27.

  • Sawilowsky, S. S., 2007: ANCOVA and quasi-experimental design: The legacy of Camp- bell and Stanley. Real Data Analysis, S. S. Sawilowsky, Ed., Information Age, 213–238.

  • Seeger, M. W., 2006: Best practices in crisis communication: An expert panel process. J. Appl. Commun. Res., 34, 232244, https://doi.org/10.1080/00909880600769944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sorensen, J. H., and D. S. Mileti, 1988: Warning and evacuation: Answering some basic questions. Ind. Crisis Quart., 2, 195209, https://doi.org/10.1177/108602668800200302.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., and K. A. Lachlan, 2010: Disasters, crises, and unique populations: Suggestions for survey research. New Dir. Eval., 2010, 95106, https://doi.org/10.1002/ev.332.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. A. Lachlan, X. Lin, and M. del Greco, 2015: Variability in Twitter content across the stages of a natural disaster: Implications for crisis communication. Commun. Quart., 63, 171186, https://doi.org/10.1080/01463373.2015.1012219.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., D. D. Sellnow-Richmond, T. L. Sellnow, and K. A. Lachlan, 2016a: Social media and corporate reputation during crises: The viability of video-sharing websites for providing counter-messages to traditional broadcast news. J. Appl. Commun. Res., 44, 199215, https://doi.org/10.1080/00909882.2016.1192289.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. A. Lachlan, and A. M. Rainear, 2016b: Social media and crisis research: Data collection and directions. Comput. Human Behav., 54, 667672, https://doi.org/10.1016/j.chb.2015.08.045.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. Lachlan, T. Sellnow, R. G. Rice, and H. Seeger, 2017a: That is so gross and I have to post about it: Exemplification effects and user comments on a news story. South. Commun. J., 82, 2737, https://doi.org/10.1080/1041794X.2016.1265578.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. Lachlan, D. Westerman, X. Lin, C. J. Harris, T. L. Sellnow, and D. D. Sellnow-Richmond, 2017b: Exemplification effects: Responses to perceptions of risk. J. Risk Res., 20, 590610, https://doi.org/10.1080/13669877.2015.1100658.

    • Search Google Scholar
    • Export Citation
  • Spence, P. R., D. Westerman, and R. G. Rice, 2017c: Exemplification theory in health and risk messaging. Oxford Research Encyclopedia of Communication, http://communication.oxfordre.com/view/10.1093/acrefore/9780190228613.001.0001/acrefore-9780190228613-e-526.

    • Crossref
    • Export Citation
  • Starbird, K., and L. Palen, 2011: “Voluntweeters”: Self-organizing by digital volunteers in times of crisis. Proc. SIGCHI Conf. on Human Factors in Computing Systems, Vancouver, BC, Canada, ACM, 1071–1080, https://doi.org/10.1145/1978942.1979102.

    • Crossref
    • Export Citation
  • Sutton, J., E. S. Spiro, B. Johnson, S. Fithigh, B. Gibson, and C. Butts, 2014: Warning tweets: Serial transmission of warning messages during a disaster event. Inf. Commun. Soc., 17, 765787, https://doi.org/10.1080/1369118X.2013.862561.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Villegas, J., C. Matyas, S. Srinivansan, I. Cahyanto, B. Thapa, and L. Pennington-Gray, 2013: Cognitive and affective responses of Florida tourists after exposure to hurricane warning messages. Nat. Hazards, 66, 97116, https://doi.org/10.1007/s11069-012-0119-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and K. A. Lachlan, 2009: Telepresence and the exemplification effects of disaster news. Commun. Stud., 60, 542557, https://doi.org/10.1080/10510970903260376.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and K. A. Lachlan, 2012a: Telepresence and exemplification: Does spatial presence impact sleeper effects? Commun. Res. Rep., 29, 299309, https://doi.org/10.1080/08824096.2012.723272.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and B. Van Der Heide, 2012b: A social network as information: The effect of system generated reports of connectedness on credibility on Twitter. Comput. Human Behav., 28, 199206, https://doi.org/10.1016/j.chb.2011.09.001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and X. Lin, 2015: Telepresence and exemplification in health messages: The relationships among spatial and social presence and exemplars and exemplification effects. Commun. Rep., 28, 92102, https://doi.org/10.1080/08934215.2014.971838.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., 1999: Exemplification theory: Judging the whole by the sum of its parts. Media Psychol., 1, 6994, https://doi.org/10.1207/s1532785xmep0101_5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., 2002: Exemplification theory of media influence. Media Effects: Advances in Theory and Research, 2nd ed. J. Bryant and D. Zillmann, Eds., Lawrence Erlbaum Associates Publishers, 213–245.

  • Zillmann, D., 2006: Exemplification effects in the promotion of health and safety. J. Commun., 56, S221S237, https://doi.org/10.1111/j.1460-2466.2006.00291.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., and S. Gan, 1996: Effects of threatening images in news programs on the perception of risk to others and self. Medienpsychol. Z. Individ. Massenkommun., 8, 168185, 239–240.

    • Search Google Scholar
    • Export Citation
  • Zillmann, D., and H. B. Brosius, 2000: Exemplification in Communication: The Influence of Case Reports on the Perception of Issues. Routledge, 168 pp.

  • Zillmann, D., R. Gibson, and S. L. Sargent, 1999: Effects of photographs in news-magazine reports on issue perception. Media Psychol., 1, 207228, https://doi.org/10.1207/s1532785xmep0103_2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Screenshot of mock NOAA Twitter feed utilized in no-name condition.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 185 146 19
PDF Downloads 114 103 21

Don’t Sleep on It: An Examination of Storm Naming and Potential Heuristic Effects on Twitter

View More View Less
  • 1 The Pennsylvania State University Scranton, Dunmore, Pennsylvania
  • | 2 Department of Communication, University of Connecticut, Storrs, Connecticut
  • | 3 Nicholson School of Communication and Media, University of Central Florida, Orlando, Florida
  • | 4 Department of Communication, University of Connecticut, Storrs, Connecticut
© Get Permissions
Full access

Abstract

Humans often prefer representations that are cognitively easier to store, and these representations are easier to retrieve later to make judgments about events. Exemplification theory draws on evolutionary logic and argues that simple, iconic, concrete, and emotionally arousing depictions of events (exemplars) are favored and thus more likely to be stored and used than are abstract, inconsequential depictions or representations. This study examined exemplified aspects of storm warnings in a Twitter feed. A three-condition study was completed, and variables examined included storm severity, susceptibility, hazard, outrage, and willingness to change or engage in specific behaviors. Results suggest the possibility of a sleeper effect impacting perceptions of severity. Results are discussed in theoretical and practical applications along with the consideration of other theories to be applied to future research.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Adam M. Rainear, adam.rainear@uconn.edu

Abstract

Humans often prefer representations that are cognitively easier to store, and these representations are easier to retrieve later to make judgments about events. Exemplification theory draws on evolutionary logic and argues that simple, iconic, concrete, and emotionally arousing depictions of events (exemplars) are favored and thus more likely to be stored and used than are abstract, inconsequential depictions or representations. This study examined exemplified aspects of storm warnings in a Twitter feed. A three-condition study was completed, and variables examined included storm severity, susceptibility, hazard, outrage, and willingness to change or engage in specific behaviors. Results suggest the possibility of a sleeper effect impacting perceptions of severity. Results are discussed in theoretical and practical applications along with the consideration of other theories to be applied to future research.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Adam M. Rainear, adam.rainear@uconn.edu

1. Introduction

Warning the public about potential risks and motivating the public to take protective actions is an area that calls for continued scholarly inquiry (Lachlan and Spence 2007, 2009). Various identification strategies and theoretically based approaches have been used to persuade and motivate the public, particularity in the event of weather-related emergencies (Fishbein and Ajzen 2010; Dixit et al. 2012; Miller et al. 2013; Martin et al. 2007). One common feature of large-scale storms that make them distinct from tornadoes, earthquakes, and other natural disasters is their identification by a name. For example, a tropical system that develops in the Atlantic Ocean and achieves a wind speed stronger than 39 miles per hour (~17.4 m s−1) is given a name from one of six annual rotating lists. These lists contain 21 predetermined names [managed by the World Meteorological Organization (WMO)] and are assigned when a system reaches tropical storm status (thus meeting the 39 miles per hour threshold). If more than 21 storms reach the threshold of being named within in a single hurricane season, names from the Greek alphabet are then used.

The naming of such storms may also provide the opportunity for individuals to use heuristic processing to understand the severity and threat of a storm. Consequently, it is possible that a more menacing or iconic name may motivate people differently that a more subdued or unemotional name. Rainear et al. (2017) were one of the first to put forth research investigating if naming storms has any impact on perceptions of the event, and the American Meteorological Society even formed an ad hoc committee on the process at their 2017 annual meeting (American Meteorological Society 2017). With this focus in mind, the current study is an initial examination into the perceptions of threats and the actions of public as a result of a name given to a storm, specifically when viewed on social media. In the following section, the history of storm naming is reviewed, as there has been much recent debate in the field of meteorology on the subject (see Fritz 2015; Palmer 2013). Then, exemplification theory is presented as a means of explaining the potential of named storms to influence perceptions, followed by an explanation of how social media is used for information seeking and communication during weather-related events. The results of an experiment examining this feature are presented, followed by discussion and implications for future research.

2. Storm naming history

In the field of meteorology, storm naming is not a novel concept. The exercise of naming Atlantic tropical cyclones has been in practice for more than 60 years. In the 1950s, the Institute for Meteorology of Free University Berlin began using female names for low pressure systems and male names for high pressure systems, in order to track large-scale weather systems without the necessity of naming storms after awkward longitude and latitude data. The current North Atlantic and eastern Pacific naming lists are managed by the WMO, and the names are assigned by the National Hurricane Center, once a storm reaches the magnitude of a tropical storm. Storms are named in alphabetical order, starting with the letter A (five letters are excluded) until either the season ends, or the 21 names are used in succession and a secondary list is utilized (NOAA 2017).

There has been discussion about the most appropriate naming strategies for large-scale weather events, after The Weather Channel (TWC) began naming (North American) winter storms in 2012. In 2015, similar strategies were adopted by both the Met Office and Ireland’s Met Eireann in naming respective large-scale winter season cyclones or wind storms. The official weather warning agency in the United States, the National Weather Service (NWS), has refrained from using or acknowledging winter storm names because of the difficulties in risks and impacts of a winter storm dependent on geographic location (Palmer 2013). Similarly, there has been a call for more synergy among the field of meteorologists about the anecdotal and trivial nature of the actual storm names, which differ from more common Western names utilized in the WMO lists (Fritz 2015).

Naming storms can have various implications for how the information might be sought, received, or utilized. If a primary goal of weather-related risk and crisis communication is to prevent or limit harm during some event, and consistent and timely warning messages help to achieve that goal and save lives (Mileti 1999; Reynolds and Seeger 2005), then naming storms in a simple and consistent fashion should help individuals uncover important behavior. Additionally, discovering information in the precrisis stage of an extreme event (e.g., before an event occurs) about appropriate behaviors is often difficult (Spence et al. 2015). Sorensen and Mileti (1988) posit that the media plays a useful role when individuals are preparing for events with long lead times, but when events occur quickly or at a more rapid pace, then individuals rely on a wider range of sources to make decisions. Individuals are unlikely to respond to a weather risk unless they understand the forecast, believe it is personally relevant to them, and believe it puts them at risk (Baker et al. 2012; Villegas et al. 2013). Social media may be especially useful during times of crisis—especially in a longer-lead-time event—given its capacity to provide real-time updates and that it is fairly resistant to infrastructure failure and physical threats (Spence et al. 2015). Although it is not known if the original decision to name storms was theoretically based, exemplification theory may provide enough explanatory and predictive power to explain the advantages or disadvantages to this practice (Spence et al. 2017c).

3. Exemplification theory

Exemplification theory (Zillmann 1999, 2002; Zillmann and Brosius 2000) suggests that events portrayed in an emotional and vivid manner will have a stronger influence on receivers of that message than events that are constructed with more general descriptions or base-rate information. Exemplification theory is a theory of media influence that translates well to social media and risk events. The theory revolves around the use and effects of media representations, or exemplars. Exemplars exist on a continuum from how accurately or inaccurately a portrayal represents the larger occurrence; they are portrayals that have a high likelihood to drive judgments of the public. The theory draws on evolutionary principles along with three cognitive mechanisms (quantification, representativeness, and availability heuristics) to explain and predict that exemplars that are concrete, iconic, and emotionally arousing influence issue perceptions more than portrayals that are abstract, symbolic, and emotionally inconsequential exemplars (Zillmann 2002; Spence et al. 2016a). The representations that humans attentionally favor, find cognitively easier to store, and retrieve from memory are more likely to be used to make judgments of the social world than abstract, base-rate information. Base-rate information is simply another way to provide information, or it can be explained as another way to tell a story. Base-rate information often consists of statistics and detailed nonemotional descriptions of events. Often humans consider base-rate information inconsequential and not necessary, and therefore it is discarded and not acted upon. Thus, in providing information about any event (for purpose of this article, a weather event), a media outlet, government agency, emergency manager, or social media feed manager can explain the facts about the storm in a measured, systematic, and unemotional style, or the same weather even can be portrayed with exemplars, such as images of fierce winds and destruction, emotional quotes about the weather event, and graphic descriptions. Zillmann (1999) notes that in a highly fact-focused story, the presence of one exemplar can cause the public to perceive the event as congruent with the exemplified portrayal and thus forget the base-rate information. Thus, a storm with a name that is concrete and iconic may drive impressions of the severity of the threat and have more influence on motivations of the public than a storm with a name that is considered inconsequential or unemotional or no name at all, even when only base-rate information is provided.

Exemplars embedded within mediated accounts can have an influence on the public’s risk perceptions and judgments (Hastall and Knobloch-Westerwick 2013; Zillmann Gibson and Sargent 1999; Westerman et al. 2015). This ability to influence the perceptions is powerful even when base-rate information is offered in the same portrayal (Gibson et al. 2011; Zillmann and Brosius 2000). Research on exemplars and risk information is still in the early stages of investigation, but the ability of an exemplar to motivate the public to avoid a risk has been well established (Westerman et al. 2009; Spence et al. 2015, 2016a). Moreover, these same findings have translated well and have been replicated in research examining exemplification processes in social media (Spence et al. 2017a). However, the exemplified portrayal of weather forecasts and warnings may be important areas to investigate as it is a risk that, in one way or another, people have similar susceptibility to. Although people still use legacy media for weather consumption, new media is a popular and well-used supplemental source. Because one iconic or emotionally driven image or phrase has been shown to influence message receivers in a style consistent with the distribution of the exemplified portrayal (Zillmann 1999), social media may be a prime tool to examine exemplars and weather forecasts.

4. Social media and weather-related events

Given the value of media use in managing crises and other cataclysmic events, it becomes vital to examine the factors that may be key in our understanding of how new media functions in this context. Recent studies on natural disasters further indicate social media as an indispensable role to promote dynamic crisis communication (Graham et al. 2015).

Morss et al. (2017) outline the present state of communication in a modern information environment, specifically from the perspective of the field of meteorology. The authors discuss how audience members are no longer passive information consumers but rather active information seekers, who are seeking to participate in the communication to better understand a situation. One example of this is dialogic communication with an agency account or professional whom they may message or request information from. Furthermore, users can relay posted information from accounts to their own social circle, furthering the spread of information about a weather risk or event, or search information via a search bar available on nearly every popular social media platform.

A consideration must be made toward behavior after consuming information about the weather on social media. Twitter data after Hurricane Sandy suggest that people do interact online not only to share information but to also engage in protective decision-making help when necessary (Anderson et al. 2016; Morss et al. 2017). Even after the disaster strikes, social media allows for mobilization of groups and volunteers, both in actual real-world behavior and in connecting resources on the local level (Starbird and Palen 2011; DHS 2013).

Because social media users have the ability to both create and transmit information, it may be the case that messages with particular characteristics may be more likely to be retweeted (Lin et al. 2016c). In an initial exploration of these factors, Sutton et al. (2014) analyzed tweets sent by emergency response agencies during a wildfire in Colorado. They found that tweets that were advisory in nature (e.g., articulating the nature of the risk and its consequences) were more likely to be retweeted than those that were purely instructive. They also found that tweets that contained imperative (as opposed to declarative) statements were more likely to receive serial transmission. In short, there were aspects of the content that led the messages to be retweeted and spread organically throughout the community of those affected, resulting in greater reach. These message aspects were related to the nature of the event, threats proposed, and actions that can be taken. To some degree, the magnitude of the event played a role in this response, and it may be assumed that to a certain degree the induction of fear or negative affect may have promoted this serial transmission. It could be the case that different storm naming strategies also elicit varying levels of fear or negative affect, and that this may lead to similar degrees of serial transmission. However, consideration of the factors driving serial transmission, including negative affect, is largely ignored by organizations using Twitter to manage crises and disasters.

Across several empirical studies, research suggests that Twitter is used by organizations managing disasters and widespread crises in a manner not dissimilar to how legacy media have been used. The aforementioned research by Sutton et al. (2014) adds that the organizations sending tweets were sending messages designed for general audiences, and that the factors driving serial transmission (in their case, phrasing) did not appear to be a consideration. In failing to consider the opportunity for serial transmission, these organizations missed out on an opportunity to engage in dialogic communication with those who were in harm’s way.

Information concerning these crises may have the capacity to inform and motivate, but in the sea of available information it may become lost if treated like a public service announcement (PSA) or Wireless Emergency Alerts (WEA) message (Lachlan et al. 2014b). What makes Twitter and other social media platforms unique under these circumstances is their capacity for generating dialogue (Lin et al. 2016b; Lachlan et al. 2017). Social media offer a chance for emergency management agencies to communicate directly with those adversely affected and to exchange information with those who are at the site of a disaster; this information can then be shared with the organization’s entire timeline and can be used to build relationships and trust with the communities who are affected (Kietzmann et al. 2011).

By creating a dialogue with the public—or at least the impression that one exists—emergency management agencies may be better able to build trust and receive detailed messages concerning mitigation and response to individuals who would be otherwise without access to this information. A sizable body of research in the crisis communication literature argues that dialogic communication between emergency managers and affected communities is critical under conditions of extreme duress, and that the use of these strategies is essential in building trust under high equivocal conditions (see Seeger 2006).

This same literature also argues that consideration of the concerns of those affected is key not only in gaining trust, but also in obtaining the information necessary to make good decisions in terms of emergency relief, mitigation, and evacuation procedures. In addition to motivating individuals to action, Twitter provides a means by which organizations responding to crises can get the information from the public that they need in order to make better decisions, and the capacity to do so in real-time as conditions and concerns change. This kind of real time modification and information management may help emergency managers create immediacy with those at risk, and at the same time develop up-to-minute and effective strategies for response (Lachlan et al. 2014a). For instance, the American Red Cross suggests organizations use their Facebook pages as a “clearing house” during emergencies, providing up-to-date information for local communities. Also, agencies are tracking the content and geographic distributions of online postings as information supplements to provide instrument-based estimates of natural disasters such as earthquake location and magnitude (Earle 2010).

In short, if effective, naming strategies may not only be important in terms of their impact on the individual user, but may drive individuals to share the information in question and engage in critical, dialogic communication with those responding to the event. This leads to the question of what specific message content may drive risk perception and response.

The research outlined in this article has supported the notion that exemplified portrayals can impact perceptions of risk and intentions to change behavior in legacy media and social media environments. However, less is known about exemplified portrayals of weather-related events through social media. Because this is a new area of application of exemplification theory, the following research question is offered: To what extent does the naming of a story have effects on perceptions of 1) storm severity, 2) likelihood of storm damage, 3) intentions to change behavior, and 4) perception of event hazard and outrage?

Demographic variables have also been shown to impact perceptions of risk in weather-related events (Burke et al. 2009, 2010), and the relationship between demographic variables and exemplars warrants further study (see Spence et al. 2017b). Because of the impact demographic variables have played in previous exemplification and disaster research studies, further examination is needed, and the following research question is offered: To what extent do demographic characteristics impact 1) perceptions of storm severity, 2) likelihood of storm damage, 3) intentions to change behavior, and 4) perception of event hazard and outrage?

5. Method

a. Procedure

A valid total of 211 participants were recruited from introductory communication and advanced business classes housed in separate colleges at a research university in the southeastern United States. The sample was 26% male and 71% female (with 3% choosing not to answer). In terms of ethnicity, 83% self-reported as Caucasian, 9% as African American, and 4% as Asian. Most participants (59%) reported an annual household income greater than $70,000. The mean participant age was 18.69 [standard deviation (SD) = 3.45]. Participants were offered course credit for participation. Instructions were provided for participants to navigate to a website where they were asked to read an informed consent statement and told that by continuing in the study they have agreed to consent. After providing consent, participants were randomly assigned to one of three experimental conditions. The experimental procedure was constructed to allow the participants to see a National Oceanic and Atmospheric Administration (NOAA) Twitter feed with one of the experimental conditions (Fig. 1). The procedure replicated the exact appearance of the NOAA Twitter feed. Participants were made to believe they were viewing/reading an actual (unaltered) Twitter feed taken from a storm. The only changes to the makeup of the Twitter feed involved the experimental manipulations. The continue button that was part of the survey software was not available for participants to navigate past the experimental stimulus until 90 s after the Twitter feed appeared. This produced an incentive to read the Twitter feed and reduced the opportunity for participants simply navigate past the stimulus materials.

Fig. 1.
Fig. 1.

Screenshot of mock NOAA Twitter feed utilized in no-name condition.

Citation: Weather, Climate, and Society 10, 4; 10.1175/WCAS-D-18-0008.1

b. Stimulus materials

The NOAA Twitter feed was chosen because it is an organization that would report on storms and is an organization that provides authority cues (Lin et al. 2016c). Within the Twitter feed there were no references to the year or date of the tropical storm. However, there were system-generated cues to the minutes and hours that passed between each tweet. This was explained in the instructions as being a feed from a previous storm. For this reason, the dates on the Twitter page were removed. Thus, the experimental conditions were set up to create the impression that it was a previous event.

The Twitter feed had all the elements of the verified NOAA Twitter page—it was an exact replication. Thus, information such as number of tweets, followers, feed description, date the account was created, and any other information was identical to the NOAA Twitter page. Participants viewed the seven most recent posts, which included four tweets by NOAA, a retweet of the NOAA Ocean Service account, and two separate retweets of the National Weather Service, one which included a retweet of the NHC Atlantic Ops. The tweets represented a 16-h update and followed the pattern and content of a previous storm feed. The tweets indicated the geographic areas of severe flooding, included maps of the flooding areas. Tweets also indicated where the Federal Emergency Management Agency (FEMA) and other federal agencies were setting up staging areas, including the number of trucks involved and cooperation with media outlets. Other tweets indicated how to contact the Red Cross and how to let loved ones know about issues of safety. Likes, retweet numbers, and all other system and user-generated cues were consistent across conditions and guided by established research (Lin et al. 2016a; Westerman et al. 2012b; Lachlan et al. 2014b). The only difference between the Twitter pages involved the experimental condition, which was the naming of the storm.

Names were pretested in a class the previous semester. Students were asked to indicate perceptions of names, all beginning with the same letter. Then these perceptions were used to determine which names were most iconic and emotionally arousing. The first condition contained a storm that was unnamed. It was simply “tropical storm.” For example, some tweets would begin with “in preparation for the tropical storm, FEMA and other federal agencies.” The second condition used the name Sam, as it was not seen as emotional. Thus, for condition 2, tweets would state “in preparation for the tropical storm Sam, FEMA and other federal agencies.” The third condition used the name Sebastian, as it was perceived as emotional and iconic. In this condition, tweets would indicate “in preparation for the tropical storm Sebastian, FEMA and other federal agencies.” The tropical storm was located off the coast of Texas, geographically distant from the respondents, and a tropical storm was used rather than a hurricane as to ensure participants were believing it was a real previous event, the assumption being that participants may remember a hurricane and therefore determine that the page was manipulated for experimental reasons.

The experimental procedures provide a realistic manipulation concerning how an organization such as NOAA would promote a storm. The tweets were taken from actual NOAA tweets concerning a tropic storm, but modified for the experiment. Organizations such as NOAA not only provide information, but also retweet from other organizations, use images and maps in their Twitter feed, and allow likes and retweets (Lachlan et al. 2016).

c. Measures

After being randomly assigned and viewing one of the conditions, participants were asked to fill out a questionnaire evaluating their responses on a web-based survey. To test perceptions of the severity of the storm, three questions were used on a 10-point scale. The first question asked participants “how severe would you rate this storm?” with anchors of mild and severe. The second question asked “how much damage to physical structures do you think this storm will cause?” with anchors of “no physical damage” to “much physical damage.” The third question asked “how likely is it that people would be hurt due to this storm?” Adequate reliability was detected for the three-item index at α = 0.88.

One question asked about perceived likelihood of storm damage. The question asked respondents on a 10-point scale “how likely is it that flooding caused by storm surge or waves from the storm will cause significant damage to homes or possessions?” with anchors of “no flooding” and “high levels of flooding.”

Six questions were used to gauge behavior intentions. Respondents were asked the question “after reading this NOAA Twitter feed, if you lived in the affected area, how likely do you believe you would do the following” with options such as “purchase supplies for the home, such as food, water and batteries,” and “take furniture or other outside items inside as a precaution” and “evacuate immediately.” Items were measured on a five-point scale from “not likely” to “very likely.” Reliability for the index was found to be α = 0.81.

Also included was the Event Hazard Outrage Scale (Lachlan and Spence 2010; Goddard et al. 2012). This 16-item instrument is designed to capture risk perception and negative effect as two separate but related constructs. Based on theoretical work by Sandman (2003) and others, the scale attempts to capture risk response as two separate but related constructs—one effective (“outrage” in Sandman’s nomenclature) and one cognitive (“hazard”) using a series of seven-point Likert items. These measures were tested with Hurricane Katrina (Lachlan and Spence 2007) and have been proven applicable to weather-related events. Confirmatory factor analysis using analysis of a moment structures (AMOS) software indicated good support for the two-factor model (hazard and outrage), with CMIN/df = 4.25, CFI = 0.94, and RMSEA = 0.09 with two items removed (CMIN/df is minimum discrepancy divided by its degrees of freedom, CFI is comparative fit index, and RMSEA is root-mean-square error of approximation). Coefficient α was detected at α = 0.85 for the hazard subscale and α = 0.93 for outrage.

6. Results

To explore the central research questions, a series of hierarchical regression analyses were performed. For each dependent variable, demographic indicators of age, sex, and income were entered on the first block. On the second block, these variables were included with the addition of experimental condition (effect coded such that −1 = no name, 0 = nonexemplified name, and 1 = exemplified name). Diagnostics did not indicate collinearity for any of the analyses [all variance inflation factor (VIF) < 5, most VIF < 2]. The analyses are discussed in terms of both model fit and the relevance of any demographic covariates.

For perceptions of severity, the initial model failed to account for a significant proportion of the variance, F (3, 199) = 2.28, n.s. (not significant). However, the addition of the experimental condition produced a significant model, where F (4, 198) = 2.72, p < 0.03 and ΔR2 = 0.02. An examination of the standardized regression coefficient for experimental condition indicates β = −0.137 and p < 0.05. While accounting for a small amount of variance, an examination of the descriptive statistics indicates that severity perceptions were strongest in the no-name condition [mean (M) = 7.01, SD = 1.42], followed by the nonexemplified name (M = 6.83, SD = 1.44) and the exemplified name (M = 6.60, SD = 1.52). Age also emerged as a significant predictor, as older respondents indicated lower perceptions of severity, with β = −0.154 and p < 0.03.

Similar results were detected for the single item outcome for likelihood. The initial model failed to account for a significant amount of variance, F (3, 199) = 2.50, n.s. The addition of the experimental condition produced a significant model, F (4, 198) = 2.49, p < 0.05 and ΔR2 = 0.02. The standardized coefficient for experimental condition did not achieve significance, though once again age negatively predicted perceptions of the likelihood of flood damage, where β = −0.179 and p < 0.02.

For hazard, both the initial F (3, 199) = 3.48, p < 0.009 and step 2 model F (4, 198) = 3.31, p < 0.012 reached statistical significance (Table 1). The addition of experimental condition did not, however, improve the model fit, β = −0.079, n.s. Once again, age was the strongest predictor of hazard, with older respondents perceiving less risk than younger ones, β = −0.172, p < 0.03.

Table 1.

Step 2 regression models.

Table 1.

This analysis plan was repeated for the remaining two dependent variables. Neither offered evidence that the message type influenced the outcome. For behavioral intentions, the step 2 model produced F (3, 199) = 0.57, n.s.; for outrage, F (3, 199) = 2.28, n.s.

7. Discussion

The results of this study contribute well to the literature on using naming strategies, heuristics, or exemplars for naming and organizing risk and crisis events. This area of investigation is important as approximately two-thirds of Twitter users use the platform to obtain news information (Barthel and Shearer 2015).

Regarding the experimental naming strategy, a few considerations exist that might explain the overall results. It is possible that the presence of a name for a storm is not enough of a heuristic to induce stronger perceptions or impressions of a storm. Individuals may be primed and familiar with receiving an influx of messages or pieces of information when a truly “severe” storm exists, and thus, they do not fully consider the information unless this happens. Similarly, the sampled population of this study was not from a coastal region and may have less experience preparing for or considering information toward a tropical storm risk. A coastal population or a population more familiar with flood and tropical storm risks may be more motivated by a naming heuristic, since they are already familiar and cognizant of this potential risk. Theoretically, this should not have been the case, but if, as results suggest, the act of placing a name on a storm is not a strong enough heuristic, then this explanation is plausible.

Demographics alone are also likely not as important in explaining severity perceptions toward the storm. On the surface, it appears that severity perceptions of the storm are the only dependent variable influenced by the experimental naming strategy. Individuals perceived the highest severity when in the no-name condition, followed by the nonexemplified name, and the exemplified name. This pattern of means is opposite of that which would be anticipated, since in the United States, a tropical system being named generally indicates it is stronger (and thus potentially more severe) than an unnamed system. Following arguments posited by risk theories and frameworks (e.g., Health Belief Model or Protection Motivation Theory), only inducing or being able to induce higher levels of threat or severity—without providing information of efficacious strategies for behavior—can also be counterintuitive to gaining compliance or getting a population to engage in proper behavior. Similarly, age predicted perceptions of severity, with older individuals holding lower perceptions of severity, suggesting that being older influences perceptions of severity as well. Older individuals have likely lived through and been exposed to more tropical systems over the course of their life, and these individuals may be less likely to find a storm more severe with a brief exposure message. Age also played a similar role in explaining negative likelihood of flood damage and lower perceptions of hazard. It should also be noted that older users utilize Twitter less compared to younger groups of the population.

With those explanations in mind, the following alternative explanation is offered, that the sleeper effect (although not directly measured in this study) was in effect and may explain why perceptions of severity were highest in the condition where the storm did not have a name. A component of exemplification theory outlined in previous research (Westerman et al. 2012a) is the sleeper effect. It has been noted that when measured immediately after exposure to an experimental stimulus with varying exemplars, relatively equal impact on perceptions of the receiver were found (Zillmann and Gan 1996, as discussed in Zillmann 2006; Spence et al. 2017c). However, when the outcomes were measured two weeks later, those exposed to more threatening exemplified portrayals reported higher threat perceptions and indicated more intention to change behavior. Thus, the impact of the exemplar was dormant and was only detected later. Zillmann and Brosius (2000) explain that the sleeper effect is likely due to the increased memory created by emotional and threatening news depictions. Depictions with exemplars are less easily forgotten and are therefore more likely to be accessible and to drive impressions.

Regarding the current study, in the condition without a name given to the storm and the absence of other information from which to put the storm in context create a circumstance where people may think back to what they remember most about storms. Even in accurate portrayals, weather is often sensationalized for entertainment value. Storms are often portrayed as dangerous and intense in the media. Although base-rate information may be provided, the iconic and concrete depictions of storms are what are remembered. Therefore, this is what is available from memory (the past exemplars) and used to make this judgment because other information is absent. This could be called the “Weather Channel heuristic,” because in the absence of any other information, or the absence of a prime, people might remember, recall, and apply these types of portrayals to the unnamed storm. If this is the case, it explains the distributions of the means for perceptions of severity.

The sleeper effect in exemplification research and in the area of named storms calls for further inquiry; it is possible that after a period of time, the effects of the naming of the storm may emerge in the original anticipated direction, or, as explained, even though it was not measured, the sleeper effect was invoked. If, after a period of time, the effects of naming emerge in the anticipated direction, this has some implications for practice and could also explain why some populations delay in taking storm-related precautions after receiving information. More research is needed on the sleeper effect to determine the amount of time needed after exposure to a stimulus for the effect to emerge. Also, the sleeper effect needs to be examined in a social media environment. If, in the case of named storms, the sleeper effect only needs minutes or hours to evoke heuristic processing, this would provide support for the naming of storms using exemplars and provide new areas of research with exemplification and weather events through social media. These suggestions are speculative in the absence of further research.

It is also possible that exemplars or exemplification theory may not be the best framework for examining how individuals consider and process storm information on Twitter. One such possible explanation for this might be the fact that the storm was categorized as a “tropical storm” for purposes of this study, rather than a hurricane. Because there exists only one quantified level of tropical storms, but five potential categories of hurricanes (based on storm strength and size), individuals may perceive tropical storms as less hazardous or less severe generally. Additionally, it is possible that a user anticipates seeing additional information on Twitter, which can cause central processing, thus rendering a heuristic less important for message consideration. The idea that the naming of a storm is an exemplar rests on heuristic processing; if the additional information on the Twitter feed motivated central processing, the anticipated effects of an exemplar would be diminished. Future research could further tease this out by investigating the impact of naming strategies and exemplars on hurricanes of different strengths. Similarly, other important factors for exemplars include using a direct quote or a visceral image in conjunction with the naming strategy. Incorporating these aspects into the experimental design would also help to better understand what heuristics are most important in individual decision-making and impression formation in the context of a weather risk. The only exemplar used in this study was the name of the storm, which was the central focus. All other tweets contained base-rate information. Future research may seek to examine the use of quotes and images within tweets to determine their effectiveness in impacting issues of threat, severity, and behavioral intentions. In previous exemplification research, Gibson and Zillmann (1998) found that people are more influenced by direct quotations than paraphrased ones, highlighting the influence a direct quote has on an audience. Similar support has been found with the use of images (Gibson and Zillmann 2000; Westerman et al. 2015). Both direct quotes and images lend themselves well to future research in social media and weather-related events.

Other variables examined included perceptions of hazard and outrage in addition to intentions to change behavior. These were not impacted by the experimental condition. Age did serve as a predictor for hazard, outrage, and likelihood of damage, but was a negative predictor. Older individuals were more likely to report lower perceptions of severity, hazard, outrage, and likelihood of damage, suggesting that these individuals may be relying on lived experience to make their outcome judgements. Generally speaking, older individuals are likely to have lived through more events, and unless they were extremely adversely affected, these individuals are likely to feel more comfortable with a weather risk. On the reverse, younger individuals may not have as many of these experiences to draw from, and thus have higher uncertainty about how they will be affected. Motivating people to take protective actions for a storm is a well-examined area of research, and many models have outlined this difficulty in consistently achieving results. Although several motivating factors have been identified throughout the literature, our findings are consistent with at least one study (Rainear et al. 2017) in suggesting that storm naming conventions may be relatively unimportant, and that other demographic and psychographic considerations are more central to the conversation.

a. Limitations

Limitations exist within this study that should be improved upon for future research. This study took place in an experimental setting. Although disaster and risk research often needs to be conducted in an experimental setting (see Spence and Lachlan 2010; Spence et al. 2016b), the limitations of this approach are documented (see Sawilowsky 2007). Although storms are a risk that people have various degrees of susceptibility to, the current sample was predominantly college-aged and lived in an area not frequented by tropical storms. Many arguments could be made covering the acceptability of this sample, but it does also constitute a limitation.

Moreover, this manipulation used a static Twitter feed. Although similar experimental designs have been used in previous studies (Lin et al. 2016c; Edwards et al. 2013; Goble et al. 2016), the use of a live Twitter feed that updates while the respondent is reading may add to the ecological validity of the study.

b. Conclusions

Although the complex dynamics of risk perceptions and exemplification effects did not emerge consistent with theory, these findings open several possibilities to the examination of perceptions of the risk of various weather-related events communicated through social media environments. Much research is needed to better understand how to motivate and inform the public in weather-related events. Given the ability of social media to allow the public to have a larger voice in the conversation, new challenges and opportunities emerge. The science of naming storms, how to communication weather-related risks, and the best practices to motivate audiences is ripe for continued research.

Acknowledgments

The authors would like to note that an earlier version of this paper received the Em Griffen Top Paper Award at the 2018 Central States Communication Association Convention. The authors would also like to thank the anonymous reviewers for their constructive feedback on earlier versions of the manuscript.

REFERENCES

  • American Meteorological Society, 2017: Ad hoc committee on naming winter storms. Accessed 1 June 2018, https://www.ametsoc.org/cwwce/index.cfm/committees/ad-hoc-committee-on-naming-winter-storms/.

  • Anderson, J., and Coauthors, 2016: Far far away in Far Rockaway: Responses to risks and impacts during Hurricane Sandy through first-person social media narratives. Proc. 13th Int. Conf. on Information Systems for Crisis Response and Management, Rio de Janeiro, Brazil, ISCRAM, 16 pp., http://idl.iscram.org/files/jenningsanderson/2016/1388_JenningsAnderson_etal2016.pdf.

  • Baker, E. J., K. Broad, J. Czajkowski, R. Meyer, and B. Orlove, 2012: Risk perceptions and preparedness among mid-Atlantic residents in advance of Hurricane Sandy: Preliminary report. Risk Management and Decision Processes Center, Wharton School, University of Pennsylvania Working Paper 2012-18, 42 pp., http://opim.wharton.upenn.edu/risk/library/WP2012-18_EJB-etal_RiskPerceptions-Sandy.pdf.

  • Barthel, M., and E. Shearer, 2015: How do Americans use Twitter for news? Pew Research Center, accessed 1 December 2015, http://www.pewresearch.org/fact-tank/2015/08/19/how-do-americans-use-Twitter-for-news/.

  • Burke, J. A., P. R. Spence, and K. A. Lachlan, 2009: Gender and age differences in use and perceptions of emergency messages during Hurricane Ike. La. Commun. J., 11, 95109.

    • Search Google Scholar
    • Export Citation
  • Burke, J. A., P. R. Spence, and K. A. Lachlan, 2010: Crisis preparation, media use, and information seeking during Hurricane Ike: Lessons learned for emergency communication. J. Emerg. Manage., 8, 2737, https://doi.org/10.5055/jem.2010.0030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DHS, 2013: Lessons learned: Social media and Hurricane Sandy. Virtual Social Media Working Group and DHS First Responders Group Rep., Department of Homeland Security, 39 pp., www.dhs.gov/publication/lessons-learned-social-media-hurricane-sandy.

  • Dixit, V., C. Wilmot, and B. Wolshon, 2012: Modeling risk attitudes in evacuation departure choices. Transp. Res. Rec., 2312, 159163, https://doi.org/10.3141/2312-17.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Earle, P., 2010: Earthquake Twitter. Nat. Geosci., 3, 221222, https://doi.org/10.1038/ngeo832.

  • Edwards, C., P. R. Spence, C. J. Gentile, A. Edwards, and A. Edwards, 2013: How much Klout do you have…A test of system generated cues on source credibility. Comput. Human Behav., 29, A12A16, https://doi.org/10.1016/j.chb.2012.12.034.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fishbein, M., and I. Ajzen, 2010: Predicting and Changing Behavior: The Reasoned Action Approach. Psychology Press, 538 pp.

    • Crossref
    • Export Citation
  • Fritz, A., 2015: Weather Channel unveils 2015–2016 winter storm names, including Mars, Regis and Yolo. Washington Post, 13 October, https://www.washingtonpost.com/news/capital-weather-gang/wp/2015/10/13/weather-channel-announces-new-winter-storm-names-including-mars-regis-and-yolo/.

  • Gibson, R., and D. Zillmann, 1998: Effects of citation in exemplifying testimony on issue perception. J. Mass Commun. Quart., 75, 167176, https://doi.org/10.1177/107769909807500116.

    • Search Google Scholar
    • Export Citation
  • Gibson, R., and D. Zillmann, 2000: Reading between the photographs: The influence of incidental pictorial information on issue perception. J. Mass Commun. Quart., 77, 355366, https://doi.org/10.1177/107769900007700209.

    • Search Google Scholar
    • Export Citation
  • Gibson, R., C. Callison, and D. Zillmann, 2011: Quantitative literacy and affective reactivity in processing statistical information and case histories in the news. Media Psychol., 14, 96120, https://doi.org/10.1080/15213269.2010.547830.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Goble, H., A. J. Beattie, and C. Edwards, 2016: The impact of twitterbot race on interpersonal impressions. Iowa J. Commun., 48, 2335.

  • Goddard, A., P. R. Spence, and K. A. Lachlan, 2012: Black beaches and bayous: The BP Deep Water Horizon spill disaster. Hazard, Outrage and Locality: An Analysis of Two Oil Spills, L. A. Eargle and A. M. Esmail, Eds., University Press of America, 249–261.

  • Graham, M. W., E. J. Avery, and S. Park, 2015: The role of social media in local government crisis communications. Public Relat. Rev., 41, 386394, https://doi.org/10.1016/j.pubrev.2015.02.001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hastall, M. R., and S. Knobloch-Westerwick, 2013: Severity, efficacy, and evidence type as determinants of health message exposure. Health Commun., 28, 378388, https://doi.org/10.1080/10410236.2012.690175.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kietzmann, J. H., K. Hermkens, I. P. McCarthy, and B. S. Silvestre, 2011: Social media? Get serious! Understanding the functional building blocks of social media. Bus. Horiz., 54, 241251, https://doi.org/10.1016/j.bushor.2011.01.005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2007: Hazard and outrage: Developing a psychometric instrument in the aftermath of Katrina. J. Appl. Commun. Res., 35, 109123, https://doi.org/10.1080/00909880601065847.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2009: Emergency communication: A framework for planning and targeting messages. J. Emerg. Manag., 7, 6972, https://doi.org/10.5055/jem.2009.0033.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., and P. R. Spence, 2010: Communicating risks: Examining hazard and outrage in multiple contexts. Risk Anal., 30, 18721886, https://doi.org/10.1111/j.1539-6924.2010.01511.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, K. M. Najarian, and M. Del Greco, 2014a: Twitter use during a weather event: Comparing content associated with localized and nonlocalized hashtags. Commun. Stud., 65, 519534, https://doi.org/10.1080/10510974.2014.956940.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, and M. Del Greco, 2014b: Screaming into the wind: Examining the volume and content of tweets associated with Hurricane Sandy. Commun. Stud., 65, 500518, https://doi.org/10.1080/10510974.2014.956941.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, X. Lin, K. Najarian, and M. Del Greco, 2016: Social media and crisis management: CERC, search strategies, and Twitter content. Comput. Human Behav., 54, 647652, https://doi.org/10.1016/j.chb.2015.05.027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lachlan, K. A., P. R. Spence, and X. Lin, 2017: Natural disasters, Twitter, and stakeholder communication: What we know and directions for future research. Social Media and Crisis Communication, L. Austin and J. Yan, Eds., Routledge, 296–305.

    • Crossref
    • Export Citation
  • Lin, X., K. A. Lachlan, and P. R. Spence, 2016a: Exploring extreme events on social media: A comparison of user reposting/retweeting behaviors on Twitter and Weibo. Comput. Human Behav., 65, 576581, https://doi.org/10.1016/j.chb.2016.04.032.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, X., P. R. Spence, T. L. Sellnow, and K. A. Lachlan, 2016b: Crisis communication, learning and responding: Best practices in social media. Comput. Human Behav., 65, 601605, https://doi.org/10.1016/j.chb.2016.05.080.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, X., P. R. Spence, and K. A. Lachlan, 2016c: Social media and credibility indicators: The effect of influence cues. Comput. Human Behav., 63, 264271, https://doi.org/10.1016/j.chb.2016.05.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Martin, I. M., H. Bender, and C. Raish, 2007: What motivates individuals to protect themselves from risks: The case of wildland fires. Risk Anal., 27, 887900, https://doi.org/10.1111/j.1539-6924.2007.00930.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mileti, D., 1999: Disasters by Design: A Reassessment of Natural Hazards in the United States. Joseph Henry Press, 371 pp., https://doi.org/10.17226/5782.

    • Search Google Scholar
    • Export Citation
  • Miller, C. H., B. J. Adame, and S. D. Moore, 2013: Vested interest theory and disaster preparedness. Disasters, 37, 127, https://doi.org/10.1111/j.1467-7717.2012.01290.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morss, R. E., and Coauthors, 2017: Hazardous weather prediction and communication in the modern information environment. Bull. Amer. Meteor. Soc., 98, 26532674, https://doi.org/10.1175/BAMS-D-16-0058.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • NOAA, 2017: Tropical cyclone names. NOAA/National Hurricane Center, accessed 30 May 2017, http://www.nhc.noaa.gov/aboutnames.shtml.

  • Palmer, R., 2013: What’s in a storm name? Weather Channel policy draws critics, but catches on. International Business Times, 8 February, http://www.ibtimes.com/whats-storm-name-weather-channel-policy-draws-critics-catches-1073016.

  • Rainear, A. M., K. A. Lachlan, and C. A. Lin, 2017: What’s in a #name? An experimental study examining perceived credibility and impact of winter storm names. Wea. Climate Soc., 9, 815822, https://doi.org/10.1175/WCAS-D-16-0037.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reynolds, B., and M. W. Seeger, 2005: Crisis and emergency risk communication as an integrative model. J. Health Commun., 10, 4355, https://doi.org/10.1080/10810730590904571.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sandman, P. M., 2003: Four kinds of risk communication. The Synergist, April 2003, American Industrial Hygiene Association, 26–27.

  • Sawilowsky, S. S., 2007: ANCOVA and quasi-experimental design: The legacy of Camp- bell and Stanley. Real Data Analysis, S. S. Sawilowsky, Ed., Information Age, 213–238.

  • Seeger, M. W., 2006: Best practices in crisis communication: An expert panel process. J. Appl. Commun. Res., 34, 232244, https://doi.org/10.1080/00909880600769944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sorensen, J. H., and D. S. Mileti, 1988: Warning and evacuation: Answering some basic questions. Ind. Crisis Quart., 2, 195209, https://doi.org/10.1177/108602668800200302.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., and K. A. Lachlan, 2010: Disasters, crises, and unique populations: Suggestions for survey research. New Dir. Eval., 2010, 95106, https://doi.org/10.1002/ev.332.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. A. Lachlan, X. Lin, and M. del Greco, 2015: Variability in Twitter content across the stages of a natural disaster: Implications for crisis communication. Commun. Quart., 63, 171186, https://doi.org/10.1080/01463373.2015.1012219.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., D. D. Sellnow-Richmond, T. L. Sellnow, and K. A. Lachlan, 2016a: Social media and corporate reputation during crises: The viability of video-sharing websites for providing counter-messages to traditional broadcast news. J. Appl. Commun. Res., 44, 199215, https://doi.org/10.1080/00909882.2016.1192289.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. A. Lachlan, and A. M. Rainear, 2016b: Social media and crisis research: Data collection and directions. Comput. Human Behav., 54, 667672, https://doi.org/10.1016/j.chb.2015.08.045.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. Lachlan, T. Sellnow, R. G. Rice, and H. Seeger, 2017a: That is so gross and I have to post about it: Exemplification effects and user comments on a news story. South. Commun. J., 82, 2737, https://doi.org/10.1080/1041794X.2016.1265578.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spence, P. R., K. Lachlan, D. Westerman, X. Lin, C. J. Harris, T. L. Sellnow, and D. D. Sellnow-Richmond, 2017b: Exemplification effects: Responses to perceptions of risk. J. Risk Res., 20, 590610, https://doi.org/10.1080/13669877.2015.1100658.

    • Search Google Scholar
    • Export Citation
  • Spence, P. R., D. Westerman, and R. G. Rice, 2017c: Exemplification theory in health and risk messaging. Oxford Research Encyclopedia of Communication, http://communication.oxfordre.com/view/10.1093/acrefore/9780190228613.001.0001/acrefore-9780190228613-e-526.

    • Crossref
    • Export Citation
  • Starbird, K., and L. Palen, 2011: “Voluntweeters”: Self-organizing by digital volunteers in times of crisis. Proc. SIGCHI Conf. on Human Factors in Computing Systems, Vancouver, BC, Canada, ACM, 1071–1080, https://doi.org/10.1145/1978942.1979102.

    • Crossref
    • Export Citation
  • Sutton, J., E. S. Spiro, B. Johnson, S. Fithigh, B. Gibson, and C. Butts, 2014: Warning tweets: Serial transmission of warning messages during a disaster event. Inf. Commun. Soc., 17, 765787, https://doi.org/10.1080/1369118X.2013.862561.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Villegas, J., C. Matyas, S. Srinivansan, I. Cahyanto, B. Thapa, and L. Pennington-Gray, 2013: Cognitive and affective responses of Florida tourists after exposure to hurricane warning messages. Nat. Hazards, 66, 97116, https://doi.org/10.1007/s11069-012-0119-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and K. A. Lachlan, 2009: Telepresence and the exemplification effects of disaster news. Commun. Stud., 60, 542557, https://doi.org/10.1080/10510970903260376.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and K. A. Lachlan, 2012a: Telepresence and exemplification: Does spatial presence impact sleeper effects? Commun. Res. Rep., 29, 299309, https://doi.org/10.1080/08824096.2012.723272.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and B. Van Der Heide, 2012b: A social network as information: The effect of system generated reports of connectedness on credibility on Twitter. Comput. Human Behav., 28, 199206, https://doi.org/10.1016/j.chb.2011.09.001.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Westerman, D., P. R. Spence, and X. Lin, 2015: Telepresence and exemplification in health messages: The relationships among spatial and social presence and exemplars and exemplification effects. Commun. Rep., 28, 92102, https://doi.org/10.1080/08934215.2014.971838.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., 1999: Exemplification theory: Judging the whole by the sum of its parts. Media Psychol., 1, 6994, https://doi.org/10.1207/s1532785xmep0101_5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., 2002: Exemplification theory of media influence. Media Effects: Advances in Theory and Research, 2nd ed. J. Bryant and D. Zillmann, Eds., Lawrence Erlbaum Associates Publishers, 213–245.

  • Zillmann, D., 2006: Exemplification effects in the promotion of health and safety. J. Commun., 56, S221S237, https://doi.org/10.1111/j.1460-2466.2006.00291.x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zillmann, D., and S. Gan, 1996: Effects of threatening images in news programs on the perception of risk to others and self. Medienpsychol. Z. Individ. Massenkommun., 8, 168185, 239–240.

    • Search Google Scholar
    • Export Citation
  • Zillmann, D., and H. B. Brosius, 2000: Exemplification in Communication: The Influence of Case Reports on the Perception of Issues. Routledge, 168 pp.

  • Zillmann, D., R. Gibson, and S. L. Sargent, 1999: Effects of photographs in news-magazine reports on issue perception. Media Psychol., 1, 207228, https://doi.org/10.1207/s1532785xmep0103_2.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save