1. Kenneth Wilson, Catherine Smith, Donna Kain and Amanda Drozdowski East Carolina University The Coastal Society June 2010

2. When hazardous storms threaten coastal communities, people need information to decide how to respond to this potential emergency. NOAA and NC Sea Grant are funding a two-year project.

3. To learn how residents, government officials, businesses and other organizations are informed and use information regarding hurricane and tropical storms.

4. to generate a model of risk and emergency communication that foregrounds the ways that different sectors of public seek and respond to information processes and products related to hurricanes and tropical storms.

5. Update a 1999 demographic survey of households and businesses in 8 North Carolina coastal counties that includes information about evacuation decision- making and behavior. Extend current pilot study of risk and emergency communication in Dare County. Conduct document evaluation study.

6. The effectiveness of storm emergency information is limited by public perception, awareness of risks and hazards, and characteristics of messages that can enhance or inhibit communication. The primary claim driving our perspective is that these limitations can be mitigated in practice by applying a robust concept of communication that accounts not only for sending and receiving messages but also for sense-making by diverse publics as they are influenced by location, situation, knowledge, and social and cultural contexts.

7. The questions that inform our study frame the investigation of socio-cultural and cognitive influences on the ways in which various publics access, interpret, and use information about risk. In this paper we will report our findings concerning how people access information about risk in emergency situations?

8. To gain deeper levels of information and to better frame the question in the survey instrument, we conducted more in-depth face to face interviews with a snowball sample of 120 residents (20 in each of 6 counties). We also interviewed 24 local government officials (4 in each of 6 coastal and coastal-area counties, deliberate sample, face-to-face interview).

9. Initial contact was made with each respondent through a postcard mailed to their home address. This card informed them of the purpose of the survey and gave them the option to complete it as a web survey. Most respondents did not opt for the web survey and interviewers began calling them a week after the postcard arrived. After seven weeks of calling, a sample of 1079 randomly selected residents in 20 coastal and coastal-area counties had been completed.

10. While a third of the respondents knew that they would ride out a hurricane and other 7% knew that they would evacuate, most respondents (59.8%) reported that their initial response to a threatening hurricane is to gather information to decide if they should evacuate or ride out the storm.

11. In order to explore the sources of information that respondents used, each respondent was asked about nine possible sources of information. The nine information sources are TV, commercial or public radio, newspapers, NOAA Weather Radio, Internet website, social networks, alert services, local officials or state/national officials.

12. The nine information sources are TV, commercial or public radio, newspapers, NOAA Weather Radio, Internet website, social networks, alert services, local officials or state/national officials.

13. “When a hurricane is threatening this county, how often do you get information from ________?” The answer ranged from Never (1), Daily (2), Several times a day (3), Every hour (4) and More than once an hour (5).

14. NMean Std. Deviation Television 10073.71.2 Social network 9792.21.1 Commercial/public radio 9602.21.4 Alert Service 9372.11.4 Internet Websites 9771.91.2 NOAA Weather Radio 10331.81.2 Local officials 9431.40.9 Newspapers 9791.30.5 State or national officials 9431.20.6 Valid N (listwise) 778

15. Respondents were also asked to rate the quality of the information they received from each source. Ratings ranged from Excellent (5) to Poor (1). Notice that the number of people rating the information varies substantially because only respondents who indicated that they used a source of information were asked to rate it. The average respondent used 4 different sources of information.

16. N MeanStd. Deviation Television 949 3.13.865 Internet Websites 460 3.01.964 NOAA Weather Radio 368 3.01.910 Alert Service 475 2.94.952 Commercial/public radio 532 2.75.929 Local officials 276 2.65 1.096 Social network 667 2.61 1.037 State or national officials 168 2.28 1.110 Newspapers 356 2.08 1.166 Valid N (listwise) 6

17. Factor 1: Alert Alert Service TV Radio Factor 2: Network Social Network Local Officials State/National Officials. The other 3 sources (NOAA, Internet Websites and Newspapers) are analyzed as individual items.

18. Correlations AlertNetworkNOAAInternetNewspaper Coastal Correlation.018-.023-.016.119 ** -.096 ** Sig..596.495.618.000.003 N887 1033 977979 Married -.024-.044.004.098 ** -.041 Sig. (2-tailed).481.193.904.002.206 N856863992 944 Male Correlation-.018 -.077 *.047.061 -.064 * Sig. (2-tailed).602.022.140.061.049 N867 872 1004955 956 Education Correlation.017-.027.029.184 ** -.009 Sig. (2-tailed).626.433.357.000.795 N845852979 931 932 White Correlation -.074 * -.066.014.085 ** -.070 * Sig. (2-tailed).031.052.669.009.032 N 846 855982 932933 Over65 Correlation-.038-.006.056 -.219 **.186 ** Sig. (2-tailed).273.872.078.000 N850856983 937936 Income1 Correlation-.006-.044-.012.229 ** -.045 Sig. (2-tailed).871.205.716.000.168 N844851977 929

19. Correlations AlertNetworkNOAAInternetNewspaper EVACOrder Correlation.111 **.059.082 **.067 * -.029 Sig. (2-ailed).001.081.009.038.362 N 873 877 1007961 neighbors Correlation.009.054-.011-.027-.010 Sig. (2-ailed).799.148.755.449.774 N 728732815790789 island Correlation.002.078 *.000.103 ** -.073 * Sig. (2-ailed).957.022.990.001.024 N 860 867 994 946950 SurgeZone Correlation.019.047.060.095 **.028 Sig. (2-ailed).602.189.071.005.417 N 786795897 862 864 floodrisk Correlation -.012-.029.038.092 **.008 Sig. (2-ailed).723.396.238.005.794 N 860870991 950 949 windrisk Correlation.033-.022-.032.057.026 Sig. (2-ailed).349.519.329.085.440 N 832837957914915 MobleHome Correlation.070 *.005-.010 -.087 ** -.071 * Sig. (2-ailed).040.892.747.008.030 N 863 868999 950951

20. Correlations Alert NOAA Internet NetworkNewspaper seekinfoPearson Correlation.170 **.023.110 **.006.023 Sig. (2-tailed).000.460.001.861.470 N 874 1010 964 882965

21. Most people use multiple sources of information but few people use them all. It should be somewhat reassuring to learn that the most frequently used source of information (television) is also the source with the highest quality rating. It is also a source that is very available to local emergency managers who need to get important information to the residents of their communities. Television is also used roughly equally by people in all our different social and demographic groups.

22. It is interesting that while alert systems are relatively new, they are relied on almost as much as radio or the Internet and more than NOAA Weather Radio. Their use does not vary by education or income and minority respondents report more frequent use than white respondents. They are used most often by residents of ocean front counties, those who know their homes are covered by evacuation orders and those people living in mobile homes.

23. The Internet is more likely to be used by people who are aware not only that the relevance of evacuation orders but also of specific risks. Living on an island Living in a storm surge zone Living in a flood zone Alert Services and Internet Websites are also the information sources most like to be used by people seeking information to make an evacuation decision.

24. Kenneth Wilson, Catherine Smith, Donna Kain and Amanda Drozdowski East Carolina University The Coastal Society June 2010