This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Running Head: SOCIAL MEDIA AND EMERGENCY WARNINGS
OPTIMIZING THE USE OF SOCIAL MEDIA
FOR THE DISSEMINATION OF EMERGENCY WARNINGS
By
JAMES EMBLETON-FORREST
A Major Research Project submitted in partial fulfilment of the requirements for the degree of
MASTER OF ARTS
In
DISASTER AND EMERGENCY MANAGEMENT Accepted and Recommended:
Abstract The purpose of this study was to determine how emergency managers in Vancouver could
optimize the use of social media for the formal dissemination of warnings. Specifically, this
study asked what are the most utilized forms of social media in Vancouver, how these
applications are perceived by users in terms of credibility, and whether there is a relationship
between the source of an emergency warning message and the assessment of credibility by the
recipient. Finally, the appetite of Vancouver residents to receive emergency warnings via social
media was assessed. A mixed-methodology was employed using an Internet-based survey
instrument with follow-up interviews and exposure to a series of simulated warning messages.
The study concludes that there is a strong linkage between the popularity of social media
applications and the assessment of credibility by users. Message recipients prefer to know the
originating source of a message to aid in their own assessment of credibility. While there is an
appetite for emergency warnings via social media, there is trepidation that it may lead to
information overload. Emergency managers in Vancouver should consider email, linkedin,
Facebook, Twitter, You Tube and Google+ a priority for warning promulgation via social media.
Emergency managers should be transparent about their reasons for using social media with
recipients and establish an online identity prior to any crisis. Also, message recipients should be
encouraged to forward emergency warnings onwards to their own social networks by email as
email from a known source is highly regarded by respondents in terms of credibility.
SOCIAL MEDIA AND EMERGENCY WARNINGS
3
Table of Contents Abstract 2 Table of Contents 3 List of Figures 5 List of Tables 6 List of Appendices 7 Chapter One: Study Introduction 8 Introduction 8 Purpose 9 Outcome 10 Chapter Two: Literature Review 11 Literature Review 11 Search Procedures 12 Defining Social Media 12 What is Social Media? 13 Key Principles 14 Web 2.0 16 Social Media: A Working Definition 18 Paradigm Shifts 19 Research Trends and Deficiencies 23 Summary 25 Chapter Three: Methodology 27 Purpose 27 Outcome 28 Methodological Approach 28 Research Design 29 Research Methods 29 Conduct of Survey 29 Conduct of Interviews and Simulated Warning Messages Assessment 30 Survey Group Composition 31 Timeframe of Study 33 Data Analysis 33 Presentation of Results 35 Limitations of the Study 36 Delimitations of the Study 36 Assumptions 36 Ethical Considerations 36 Chapter Four: Study Findings 38 Respondent Demographics 38 Question Three to Seven: Internet Usage 40 Questions Eight to Ten: Social Networking Applications and Credibility 45 Questions Eleven to Fourteen: Source vs. Medium and Credibility 50
SOCIAL MEDIA AND EMERGENCY WARNINGS
4
Subsequent Interviews and Simulated Message Assessment 51 Coding of Qualitative Interview Data 52 Summary 54 Chapter Five: Discussion, Conclusions and Recommendations 56 Research Aim 56 Demographics and Internet Usage 56 Social Media Utilization 60 Social Media Credibility 63 Source versus Credibility 65 Receptiveness 70 Conclusions 71 Recommendations 72 Implications 74 Chapter Six: Reflective Learning 76 Researching Skills 76 Thinking Skills 78 Conclusion 81 References 82 Appendices 88
SOCIAL MEDIA AND EMERGENCY WARNINGS
5
List of Figures
Fig 3.1. Methods of Assessment 34 Fig 4.1. Respondents by Age and Gender 40 Fig 4.2. Internet Access at Work 41 Fig 4.3. Internet Access at Work Regulated by Employer 42 Fig 4.4. Use of Mobile Internet Device 43 Fig 4.5. Use of Push Notifications 44 Fig 4.6. Social Media Applications Used 46 Fig 4.7. Assessment of Credibility of Each Social Media Application Used 47 Fig 4.8. Credibility Rating by Social Media Application 48 Fig 4.9. Likelihood of Behavioral Change by Social Media Application 49 Fig 4.10. Simulated Messages Ratings. 52
SOCIAL MEDIA AND EMERGENCY WARNINGS
6
List of Tables
Table 2.1. Paradigm Shifts in Emergency Management 21 Table 3.1. Methods of Triangulation 35 Table 4.1. Frequency Distribution by Age Group and by Gender 38 Table 4.2. Vancouver Population by Age and Gender 39 Table 4.3. Summary of Analytical Coding by Age and Gender 53
SOCIAL MEDIA AND EMERGENCY WARNINGS
7
List of Appendices Social Media and Emergency Management Survey Appendix A Consolidated Response Data for Survey Appendix B Simulated Emergency Management Messages Appendix C Simulated Message Assessment Sheet Consolidated Responses Appendix D Coded Interview Data Appendix E
SOCIAL MEDIA AND EMERGENCY WARNINGS
8
Chapter One: Study Introduction
Introduction
One of the consistent trends in emergency management is that disasters are becoming more
frequent and affecting more people (ICRC, 2005). An increase in the interdependence of
individuals, communities, communication networks, industry, and governments is creating
complex disasters that can rapidly traverse geographic, social, and political boundaries (Canada,
2011; Murphy & Etkin, 2011; Tierney, 2001). Part of the role of the emergency manager in this
increasingly complex environment is to initiate protective measures against an event such as
extreme weather, seismic activity, a health pandemic, a major infrastructure failure, or a terrorist
incident.
More traditional methods of warning dissemination include word of mouth and mass media
(Sutton, Palen & Shklovski, 2008). However, it is important to recognize that conventional
television and radio, while potentially very influential, are not the only forms of high-speed mass
communication (Scanlon, 2007), and indeed in the future may not be the most frequently used.
The proliferation of mobile electronic devices, the Internet and social media has created an
almost instantaneous flow of information that the public is increasingly using for a wide variety
of purposes, including emergency warnings. Yet research into more traditional dissemination
methods shows that the first thing a recipient usually does upon receipt of an emergency warning
is verifies it with another source (Cutter & Smith, 2009; Perry & Greene as cited in Scanlon,
2007; Sorensen & Sorensen, 2007). Therefore the emergency manager, working in a world that
now encompasses both traditional and new forms of high speed mass communication, must
consider the most credible means by which to disseminate a warning in order to maximize the
SOCIAL MEDIA AND EMERGENCY WARNINGS
9
probability that the message will be perceived as reliable, understood by the target audience, and
if necessary, complied with by those affected.
Social media offers a rapid and versatile means of communication with significant reach.
Over one third of adults who use the Internet are using web logs (blogs), social media, and text
messaging on their portable devices (Veil, 2011). The volume of traffic via social media is
staggering: Twitter generates an estimated 55 million tweets a day, Flickr amasses more than
6000 photos each minute, YouTube accumulates over 24 hours of video a minute and takes up
more than 10% of all internet traffic, and Facebook has more than 400 million active users,
making it the most visited site on the Internet in the USA (Kavanaugh, 2012)
However, while much research has been conducted into threat warnings, crises
communication, perception of threat, and warning compliance, there has been less work in the
use of social media to convey a warning message prior to an incident. While traditional print and
electronic mass media (television and radio) is subject to the legal and ethical vetting that one
would expect of a professional media organization, the multi-directional configuration and mass
participation of social media is not (Fearn-Banks, 2011). While this makes the social media
information landscape more fluid and current, it also potentially impacts upon its perceived
credibility (Seeger, 2006) and therefore raises questions as to its potential for use by emergency
managers.
Purpose
The purpose of this study is to determine how emergency managers in Vancouver can
optimize the use of social media for the formal dissemination of warnings. In addition, the
following sub-questions will be addressed:
a. What are the most utilized forms of social media within Vancouver for personal use? The
purpose of this question is to determine what forms of social media Vancouver residents use for
SOCIAL MEDIA AND EMERGENCY WARNINGS
10
general social networking and everyday use. This will provide an idea of the social media
‘channels’ available to the emergency manager through which to promulgate a warning.
b. How are those forms of social media perceived in terms of credibility by the residents of
Vancouver? Different social media sites may mean different things to different people. Some
sites may be highly regarded in terms of credibility, whereas others may be perceived as lighter
entertainment and not a worthwhile source of warning information. Understanding the
relationship between the forms of social media used and the assessment of credibility by the
target audience may make positive behavioral change and warning compliance more likely. The
more credible the site is perceived to be, the higher priority that site should be for use as a tool
for the dissemination of warnings by an emergency manager.
c. Is there a relationship between the source of the message and the perception of
credibility? Social media can serve as a rapid and widespread distribution method for
information. However, often the originating source for that piece of information can be lost in
the flurry of ‘re-tweets’, ‘posts’ or ‘likes’. How important is it to a social media user to know the
originating source for a piece of information? In relation to the prior sub-question, is knowledge
of the originating source more important than the perceived credibility of the form of social
media upon which that information is currently being viewed?
d. Would Vancouver residents wish to receive emergency warning information via their
social media? This question will hopefully identify if there is an appetite for using social media
for emergency warning messages in Vancouver.
Outcome
The hope is to provide data that will aid emergency managers in Vancouver in selecting the
optimal type of social media via which to promulgate a warning message to ensure speed of
delivery and to enhance the probability of compliance, thereby contributing to public safety.
SOCIAL MEDIA AND EMERGENCY WARNINGS
11
Chapter Two: Literature Review
Literature Review
A great deal of research has been conducted into the field of emergency warning messages
and compliance. However the majority of this material is centered on conventional mainstream
mass media, social networks and word of mouth. Although this research provides a good starting
point, this Major Research Project (MRP) will focus exclusively on social media as the means of
message promulgation. Prior research shows that to be effective, an emergency warning must be
timely, accurate, credible, and transmitted via as many sources as possible (Auf de Heide, 1989;
Mileti & Peek, 2000). Current research into social media suggests that these factors continue to
be important and shape target audience perception (Veil, 2011; White, 2012). However,
developments over the last decade in social media suggest that people are even less likely now to
rely on single sources for official information, in part because the Internet provides instantaneous
access to a wider variety of sources to consult (Mersham, 2010). The purpose of this MRP will
be to examine message promulgation via social media, specifically in Vancouver. Social media
provides a means to rapidly repeat and exponentially transfer a message on numerous sources
simultaneously. The utility of social media for promulgating emergency warning messages
seems clear. However the relationship between the forms of social media used, the originating
source of the message and the resultant impact on credibility by the viewer is less understood and
represents a shortfall in the current research. Furthermore, there is very limited research into
understanding how social media can be used to assist the public in making decisions that result in
recommended actions (O’Brien, 2012). Understanding the characteristics of social media, how
social media has changed emergency management warnings, and where the deficiencies in
SOCIAL MEDIA AND EMERGENCY WARNINGS
12
research currently exist will be key in framing the context of this MRP. Consequently, this
literature review will focus on three main areas:
a. Defining social media.
b. Paradigm shifts in emergency management related to social media.
c. Research trends and deficiencies.
Search Procedures
The primary means used for research for this literature review was the Internet. Google
Scholar, Summon and the Royal Roads University library database were used in conjunction with
the following search terms:
a. Social media + emergency management.
b. Social networks + emergency warnings.
c. Crisis communication + social media.
d. Social media information flow.
e. Social media + credibility.
f. Social media + government.
g. Trust + Internet.
h. Emergency warning compliance.
i. Emergency warning + social media.
j. Web 2.0.
k. Mass media + emergency management.
Defining Social Media
It is difficult to imagine a world without the Internet and social media. Social media has been
likened to the agricultural or industrial revolutions in the sense that it fundamentally alters the
way we function and in a sense, the balance of power in terms of information (Crowe, 2012).
SOCIAL MEDIA AND EMERGENCY WARNINGS
13
When compared to other communication mediums, the pace at which social media has become so
influential is staggering. It took 38 years for radio to reach 50 million listeners while terrestrial
television took only 13-years to do the same. However the website Facebook had over 500
million users within only 4-years of being developed (Walaski, 2013). With such a short history
and a rapid and ongoing evolution, the breadth of research into social media is limited but
increasing and there are a number of key themes emerging that are relevant to this study.
What is Social Media?
A definition often centers on something’s purpose or reason for being. Drawing together a
working definition for social media from across the current research is difficult as new
applications and ways of using social media are developing every day. “Social media is an
umbrella term that covers all the various electronic tools, technologies, and applications that
facilitate interactive communication and content exchange, enabling the user to move back and
forth easily between the roles of audience and content producers” (Spicer, 2013). This notion of
the user being both the audience and the producer of content is key and impacts on information
credibility as information is perhaps less likely to come from a subject matter expert or person in
authority, but instead from a bystander or layperson. Social media stresses openness and
participation where the consumers are the contributors, and the contributors are the consumers
(Veil, 2011). Credibility remains important because although social media may serve to
eliminate the “middleman (i.e. the Media)” in the passage of information, the information still
must be believed to be relevant (White, 2012). White further defines social media as a form of
“…electronic communication through which users create online communities to share
information, ideas, personal messages and content” (White). Social media is further
characterized by “…dynamic interaction in a networked format” (Crowe, 2012). Social media
SOCIAL MEDIA AND EMERGENCY WARNINGS
14
can also be defined as “…Internet-based applications that enable people to communicate and
share resources and information” (Lindsay, 2011).
Key Principles
Although the definitions are varied and are developing at the same explosive pace as the
Internet itself, a few key principles that define social media resonate across the existing literature.
Firstly, social media is Internet-based and electronic in nature. Although this may seem obvious,
it does provide some basic parameters. To use social media, a person must have access to an
electric power source and a reliable communication system, they must have access to the Internet
(be it wireless or via a ‘hard-wire’), and they must have a computer device like a laptop, desktop
computer or smart phone to view content (Shankar, 2008). Devices capable of viewing social
media content are becoming cheaper to produce and highly accessible. About 40% of current
Facebook users access the site via a mobile device and mobile users are twice as active on the site
as those who use the site from a non-mobile device such as a desktop computer (White, 2012).
Both electric power and an Internet signal are increasingly being considered as ‘core’ utilities to
be provided to a community. In fact the United Nations has ruled that Internet access is a basic
human right to be guaranteed and protected by every state (United Nations General Assembly,
2012).
Next, social media is about, and for, people. It is a way of interacting that transcends
geographic space. It may seem natural to stereotype social media to being predominantly the
remit of the younger generation, but interestingly, the 50-64 year old age group is the fastest
growing demographic amongst social media users (Crowe, 2012). Although commonality of
language can be a barrier to some, English appears to be the predominant means of
communication over social media globally. However, popular applications like Facebook are
available in over 70 different translations (White, 2012). Also, the capability to exchange and
SOCIAL MEDIA AND EMERGENCY WARNINGS
15
view images and video often transcends a common language. Although the quality of images
uploaded over social media may be less polished than those produced by a professional media
outlet, Crowe suggests that raw, grainy, hand-held footage or unedited audio files may seem
more compelling and believable to some viewers, potentially enhancing credibility (2012).
Although the Internet is subject to regulation by government, the viral nature of information
transfer makes policing the Internet highly challenging. Even if disseminated information is
subsequently revealed to be inaccurate, once that information is on the Internet or being
exchanged between social media sites or users, it is hard to remove. Social media is ever
changing, transparent and cost effective in that it continues to rapidly morph to meet new
requirements limited in a sense only by creativity (Crowe, 2012). Whereas uploading footage
from a remote location once required physical transfer of the written word on paper or expensive
and cumbersome satellite upload technology, the same can be achieved almost instantaneously
with a palm-sized mobile device and a social media account. Despite social media seemingly
being by and for the people, one must not under-estimate the commercial power of the Internet
and social media. Over half of Twitter users in the USA are currently following a product brand
or a corporation (Crowe). So although the Internet and social media are very much in the hands
of the general public, that does not preclude government and industry from using social media for
marketing, public relations, and crisis communication purposes. The originating source of a
message may have an impact on message credibility, but determining whom the originating
source is can be difficult when a message has passed through a number of different social media
sites and users. Social media can potentially be used as a tool for disinformation as well, by
releasing privileged or ‘leaked’ information or deliberately malicious or erroneous material.
Social media stresses pro-activeness and projection. The speed of information transfer has
created very short cycles in which to receive, review, analyze, decide upon and respond to
SOCIAL MEDIA AND EMERGENCY WARNINGS
16
information. During the 2013 crash of Asiana flight OZ214 in San Francisco, the first image of
the impact had been uploaded to Twitter within 30 seconds by a Google employee boarding an
unrelated flight. This was over a minute before the emergency slides on the crashed aircraft had
even deployed to aid in evacuation. The same Google employee then received numerous direct
messages from news agencies via Twitter requesting interviews and more details (Simpliflying,
2013). By news agencies and interested parties automatically monitoring social media for key
terms like ‘plane crash’ or ‘terrorist explosion’, the speed at which interested parties can identify
information and then promulgate that information to a wider audience is increasingly rapid.
Currently however, in terms of emergency management, government uses social media somewhat
passively to disseminate information, relying on the target audience seeking information from
websites and electronic bulletin boards rather than information being ‘pushed’ towards a targeted
audience (Lindsay, 2011). This view is supported by research conducted by Spicer who suggests
that formalized emergency management as a discipline under utilizes the capability of social
media (2013). It is perhaps this ‘thirst’ for information amongst Internet and social media users
and the seemingly limited provision of information by government that has helped the Internet
and social media evolve into more of a two-way communication tool, forcing government to re-
evaluate how they use social media.
Web 2.0
The early days of the Internet were generally somewhat one-way in nature. Those with
Internet access uploaded information onto the Internet for viewing by others with Internet access.
Once that content was uploaded, the sender could conceivably forget about it, and simply let
others passively view the material. There may be some limited facility for dialogue with the
ability to leave comments or ‘post’ feedback, but the relationship was somewhat one-way and
linear in nature. There was less of a linkage between feedback and content. The Internet and
SOCIAL MEDIA AND EMERGENCY WARNINGS
17
social media have now developed to enable a greater degree of collaboration and communication
and have become known in characteristically computer-centric speech as Web 2.0 (White, 2012).
Applications like “Wikis” and shared websites allow for cooperative construction of social media
content with ‘real-time’ dialogue and feedback. Simple applications such as the ability to ‘like’
something on Facebook are being used by television newscasts to permit real-time viewer
feedback on news stories and comment sharing via social media polling. Educational institutions
are placing an increasing number of courses on-line providing students with the ability to attend
lectures and discuss course content ‘virtually’. Students who have never actually met in person
can create online learning communities using social media, enabling the completion of group
projects online. A person’s social circle is perhaps as likely to contain people that they have not
met or whom they seldom see in person, as people they see everyday. As a consequence, in an
emergency management context, relying on a member of one’s social network for advice,
assistance or an opinion may mean relying on someone who is not physically co-located with the
user or even subject to the same impending threat.
A survey by the Canadian Red Cross suggests that 63% of Canadians think that first
responders should be prepared to respond to a request for assistance posted by people on social
media (2012). There is an expectation now by Canadians that emergency management
authorities not only use social media but actively monitor it to help drive response activity.
However only about one third of Canadians polled believe that such a request posted on social
media would currently translate to aid (Red Cross). This perhaps indicates a capability gap
between the Canadian public’s appetite for the use of social media as an emergency management
tool and the capacity of first responders to meet that expectation. In part this is perhaps due to
simple resource limitations: including a lack of capacity on the part of emergency agencies to
SOCIAL MEDIA AND EMERGENCY WARNINGS
18
devote staff to monitor social media and the subsequent linking of monitoring to the operational
deployment of resources.
The monitoring of social media websites by government and non-governmental agencies can
be a good source of situational awareness and trend analysis. Palen (2008) has suggested a tighter
linkage is required between emergency management and social media research proposing that
organizational processes are crucial to leveraging publicly disseminated information (i.e. social
media monitoring). Mersham (2010) suggests that there is a cultural mindset amongst public
officials for only using ‘credible’ sources of information and that public officials feel that social
media is not a readily verifiable source. This comes from an understandable fear of promulgating
false or unaccredited information and a moral responsibility to ‘get it right’ and not spread undue
panic. Emergency management organizations and government may be held accountable for their
decision to issue or not issue a warning and the lack of attribution to some viral warning
messages makes that decision to promulgate more complex (Sene, 2008). This should not detract
from government’s ability to monitor social media for information and situational awareness
however. Nonetheless the volume of traffic over social media and the difficulty of discerning the
true volume due to re-posting, repeating etc. makes social media monitoring complex and
potentially resource intensive (Imran, 2013).
Social Media: A working definition
Based on the literature reviewed and in light of the purpose of this MRP, the author proposes
the following definition of social media:
“Social media refers to Internet-based applications that enable the free and rapid exchange of
information and opinion within a virtual community.”
SOCIAL MEDIA AND EMERGENCY WARNINGS
19
Paradigms Shifts
In relation to emergency warning dissemination, social media represents a potentially radical
change in conventional warning models. Firstly, social media has enabled a shift from
unidirectional warnings to multi-directional warnings. Traditional warning method research
stressed the need for multiple channels but there was little crossover between those channels
creating ‘stove-piped’ and linear pathways of information (Botterell, 2008; Mileti, 2007; Scanlon,
2007). Although there was a degree of ‘cross-fertilization’ of information amongst a person’s
actual social network, the majority of the information routes into that social network were
singular. Social media allows information to follow numerous divergent and often intersecting
channels, creating more of a potential for dialogue between emergency managers and a
population (Fearn-Bank, 2011). However, predicting how and where that information may travel
via the Internet is extremely difficult. White suggests that controlling the flow of information on
social media is fruitless (2012). However, by contrast Hui et al. suggest that in times of crisis,
information sharing on mediums like Twitter reverts back to being one-way in nature, with a high
proportion of information posted being passed on to the public via social media but not
necessarily replied to or questioned by the recipient (2012). This would suggest that in times of
crisis, people prefer to receive specific and clear instructions that are consistent across multiple
channels, a principle well established in research into more traditional emergency warning
methods (Auf de Heide, 1989). What social media does is provide a significantly greater number
of channels by which to pass that information quickly and the facility to engage in dialogue for
those amongst the target audience who wish to do so.
Historically, emergency management structures have generally been hierarchical and based on
traditional military and government command and control models (Quarantelli, 2000).
Information flows downwards from a central control node, usually a government official or
SOCIAL MEDIA AND EMERGENCY WARNINGS
20
department. The speed at which that government department can be alerted to an impending
crisis, means the decision to disseminate a warning, and then promulgate that information
through traditional channels, is often slow and mired by bureaucracy (Sorenson, 2000). Social
media however is readily available to the wider population or the ‘citizen journalist’. Current
information or even images captured on a cell phone at the scene of an incident as it unfolds,
possibly out of context, can instantaneously be uploaded and disseminated onto social media
before formal emergency management authorities are aware the incident has even occurred and
decisions to act can be made. This to an extent puts the power of information at the lowest levels
and de-links the power to affect change from official sources and government authorities.
Information flow has switched from the impersonal to the personal with an emphasis on ‘back-
channels’ facilitated by social media (Mersham, 2010; Sorensen, 2000). However, with
information less likely to originate with an official source, will this in turn impact on credibility?
Social media permits online communities to form based on geographic proximity or a shared
jeopardy in relation to an impending emergency. However there has been limited research into
how proximity to jeopardy impacts upon a social media user’s perception of credibility. For
example, what will someone who faces an impending disaster find more credible as the source of
a warning over social media: a distant, removed government expert commenting on an impending
disaster or an average citizen who lives only a few minutes away and is fundamentally at risk as
well? A person can have links with both of those entities via social media, but it is unclear,
which would seem more compelling and credible to the viewer. Credibility may be further
impacted by social media often not permitting the traditional cues people rely upon for
determining truthfulness like posture, intonation, eye contact etc. (Castillo, 2011). So while prior
research indicates that warning messages from authority figures and respected traditional media
sources still have a level of credibility, there is limited research into how the same messages
SOCIAL MEDIA AND EMERGENCY WARNINGS
21
promulgated over social media are regarded. While some researchers suggest that word of mouth
is perceived as more trusted than mainstream media, existing research makes this less easy to
support when that word of mouth is a 140-character message on Twitter received from a person
you have never physically met and cannot see (Imran, 2013).
With information more likely to be transmitted over social media by a ‘non-professional’ in
the field of crisis communication, the information is more likely to be personalized, subjective
and localized (Veil, 2011). Traditional emergency management warnings are issued by official
or media organizations that have specific criteria, standards and thresholds to facilitate the
decision to promulgate (Sene, 2008). News agencies are bound by legal and ethical codes of
conduct that regulate the release of information and warnings. A participant in social media
however, may not consider the consequences of dissemination, possibly leading to the
exaggeration of severity, false warnings, and confusion (Kasperson & Kasperson, 1996).
However, what emergency managers need to realize is the sum total of the situational awareness
within any social network is likely to exceed the knowledge of any one individual emergency
management practitioner (Crowe, 2012). As a consequence, emergency managers exclude social
media from their warning strategy at their peril.
So in short, current research suggests that the major paradigm shifts in the field of emergency
management warnings related to the evolution of social media can be summarized as follows:
Table 2.1
Paradigm Shifts in Emergency Management
Emergency Management: Pre-social media Emergency Management: Post-social media
Unidirectional information flow: -From official sources to the public via the traditional mass media. -Slower and linear in nature.
Multi-directional information flow: -From both unofficial and official sources via a complex web of multi-channel linkages. -Distance and time no longer an issue.
SOCIAL MEDIA AND EMERGENCY WARNINGS
22
-Can be virtually instantaneous.
Lack of Information: -Timetable of information flow controlled by sender, often a central authority. -Hesitancy to release partial information picture, leading to slower flow of information or absence of information.
Information Overload: -Information from multiple sources. -Greater tempo of information shared and repeated exponentially. -Desire by central authority such as government to keep ahead of the ‘information cycle’. -Can create an overload of information and comment of varying credibility/utility. -Information fusion and assessment can be difficult for the individual.
Top-down in terms of authority: -Official source vested with the authority and responsibility to issue warnings.
Bottom-down in terms of power: -Unofficial sources lacking formal authority but who possess the power to affect change through influence and social connectivity.
Official sources: -Government, first responders, news media, emergency management practitioners. -Attributed sources. -Accountability.
Unofficial sources: -Eyewitnesses, bloggers, social media users, non-professionals, community opinion formers. -Less attributed. -Limited accountability.
Control of information and process: -Formal thought process and decision-making cycle to disseminate warnings based around legislation, reporting thresholds, and codes of ethical conduct.
Transparency: -Honesty, timeliness, candor, and a lack of process and formalization. -Limited ‘ownership’ of information posted to the Internet.
Impersonal: -Objective, focused on the wider community as opposed to individuals. -Emotionally detached and pragmatic. -Regional versus individual in terms of focus.
Personal: -Often highly subjective, personalized, individual and anecdotal. -Authentic human voice to which audience can relate possibly enhancing credibility. -Geographically relevant and localized.
Mass media as the source: -Messaging via mass media most common.
Social media as the source: -Mass media now using social media to deliver
SOCIAL MEDIA AND EMERGENCY WARNINGS
23
-Attribution to formal sources impacting upon credibility. -News agencies work hard to cultivate a reputation for honest, accurate, and responsible reporting.
their message. -Originating source of information often difficult to discern. -Social media channel being used may impact on credibility of message.
Empowerment of Authorities: -Knowledge is power. -Control of that flow of knowledge exacerbates the hierarchical relationship of the emergency management authority and the community.
Empowerment of Community and Individuals: -Ordinary citizen often has access to the most up-to-date information and the means to promulgate via social media. -Validation of actions through shared opinion and support by on-line community.
Research Trends and Deficiencies
Although the advent of the Internet and social media has created much discussion within
emergency management, the majority of this research focuses on crisis communication (Coombs
& Holladay, 2009). In other words how to handle a crisis as it emerges using social media as
almost a public relations tool. Much of the existing research into social media also seems to be
focused on citizen response to a disaster (i.e. online information sharing, resiliency, and
coordinating response and recovery activities) with less emphasis on the promulgation of initial
warnings using social media before an event. Schultz et al. have looked at message credibility
and reputation, but again with a focus more on crisis communication and crisis containment
strategy (2011). Schmierbach and Oeldorf-Hirsch (2012) have conducted extensive research into
the issue of credibility for specific areas of social media like Twitter. But they acknowledge that
there has been little comparative research of how the same information from the same source
disseminated on different social media channels is perceived in terms of credibility.
Much of the research into disaster warnings has been somewhat restricted to singular
communication channels, which is not as relevant to the multi-directional and multi-channel
SOCIAL MEDIA AND EMERGENCY WARNINGS
24
distribution that social media can achieve (Burkhart, 1990). Historically, there has been research
into how social networks impact upon a person’s decision to comply with warning information.
However, much of this research involves face-to-face contact and ‘actual’ social networks as
opposed to the virtual social networks that social media can provide. Widener has suggested a
need to examine how geographically distant members of a person’s social network, joined only
by social media, may impact upon a decision to comply with a localized warning (2013). He
further suggests that from an emergency management perspective, determining whom to target
within social networks in order to maximize the chance of message dissemination and enhance
credibility should be a focus for future study (2013), a research need also supported by Crowe
(2012).
The question of the timescale in terms of threat is also an area that is not well understood,
specifically slow versus rapid onset threats. The speed of transmission and dissemination on
social media lends itself to rapid-onset warnings like seismic activity, extreme weather or a
terrorist incident. However the collaborative and conversational nature of social media also
provides a mechanism for debate over slower-onset events like climate change or a seasonal
pandemic threat. What research has been conducted into the credibility of warning information
via social media has not differentiated slower versus rapid onset emergencies. Potentially, the
lack of time to process information and make a decision to act in rapid onset emergencies makes
related information received over social media more credible out of necessity. In other words, if
there is limited time to make a decision, a person may rely on the most readily available opinion
or source of information, often found on the Internet via social media. At this point in time there
is limited research available to fully examine this issue. This MRP will focus exclusively on
rapid onset emergencies in relation to warning promulgation.
SOCIAL MEDIA AND EMERGENCY WARNINGS
25
I have identified the relationship between social media and credibility with regard to
emergency warnings as the most critical research gap resulting from this literature review.
Traditional ‘rules’ for warning messages suggest that an authority figure providing a serious yet
empathetic message will have the best effect on a population (Auf de Heide, 1989; Seeger, 2006).
However, what needs to be considered is the effect on credibility that the same messages will
have when viewed on social media. Different social media platforms may impact on the
perceived credibility of the message. Furthermore, this perceived level of credibility for official
messaging must be contrasted with messages received from unofficial sources and ‘ordinary’
citizens via social media. Determining through further research, the perception of credibility of
different social media forms and methods of dissemination over social media, will provide
guidance of how best to disseminate a warning when time is short.
Summary
Traditional disaster warning methodology is based around a unidirectional flow of information
from official source to target population. However disaster and emergency management scholars
increasingly recognize “...that technological and social developments over the last decade means
the public no longer relies on a single source of official information” (Mersham, 2010, p. 139).
Social media has brought about a shift from official information and warning distribution from
practitioners to the public via mass media being the sole source of information, to an unrestricted
and multi-directional sharing of knowledge from any and all sources. While credibility in
conventional mass media has been extensively studied, less is known about the credibility of
using social media for disseminating these rapid onset emergency warnings. What is unclear is
what the effect of where the message originates and the social media form used will have upon
credibility amongst the target audience. The aim of this project will be to examine these factors
SOCIAL MEDIA AND EMERGENCY WARNINGS
26
in a small study in Vancouver, Canada in order to suggest ways to maximize the effectiveness of
issued warnings via social media.
SOCIAL MEDIA AND EMERGENCY WARNINGS
27
Chapter Three: Methodology
Purpose
The purpose of this study is to determine how emergency managers in Vancouver can
optimize the use of social media for the formal dissemination of warnings. Such agencies could
include municipal, regional, provincial and federal government, emergency services and non-
governmental aid agencies. In addition, the following sub-questions will be addressed:
a. What are the most utilized forms of social media within Vancouver for personal use? The
purpose of this question is to determine what forms of social media Vancouver residents use for
general social networking and everyday use. This will provide an idea of the social media
‘channels’ available to the emergency manager through which to promulgate an emergency
warning.
b. How are those forms of social media perceived in terms of credibility by the residents of
Vancouver? Different social media sites may mean different things to different people. Some
sites may be highly regarded in terms of credibility, whereas others may be perceived as lighter
entertainment and not a worthwhile source of warning information. Understanding the
relationship between the forms of social media used and the assessment of credibility by the
target audience may make positive behavioral change and warning compliance more likely. The
more credible the site is perceived to be, the higher priority that site should be for use as a tool
for the dissemination of warnings by an emergency manager.
c. Is there a relationship between the source of the message and the perception of
credibility? Social media can serve as a rapid and widespread distribution method for
information. However, often the originating source for that piece of information can be lost in
the flurry of ‘re-tweets’, ‘posts’ or ‘likes’. How important is it to a social media user to know the
SOCIAL MEDIA AND EMERGENCY WARNINGS
28
originating source for a piece of information? In relation to the prior sub-question, is knowledge
of the originating source more important than the perceived credibility of the form of social
media upon which that information is currently being viewed?
d. Would Vancouver residents wish to receive emergency warning information via their
social media? This question will hopefully identify if there is an appetite for using social media
for emergency warning messages in Vancouver.
Outcome
The hope is to provide data that will aid emergency managers in Vancouver in selecting
the optimal type of social media via which to promulgate a warning message to ensure speed of
delivery and to enhance the probability of compliance, thereby contributing to public safety.
Methodological Approach
The key to establishing a research methodology is to understand how best to answer the
research questions in an objective way (O’Leary, 2010). The research questions in this project
lend themselves to both quantitative and qualitative examination and thus this project used a
mixed-methods approach to determine what forms of social media the survey group uses, how
credible each form is considered and how best to employ social media for the promulgation of
emergency warnings. The mixed-methods approach was also used in part to overcome the
“…dominance of case-study-based predictions in the field of crisis communication” (Coombs &
Holladay as cited in Schultz, Utz & Goritz, 2011, p. 25). Sub-question A addressed the social
media sites used by Vancouver residents and can be reflected as a list of social media sites linked
to the number of survey respondents who use those sites. The respondents were also asked to
give qualitative comment as to why they choose to use those sites specifically. This question was
therefore addressed using both qualitative and quantitative methods. Sub-question B asked for an
assessment by respondents of credibility and was assessed by not just a numerical scale
SOCIAL MEDIA AND EMERGENCY WARNINGS
29
(quantitative) but by respondent opinion and anecdotal evidence (qualitative). Sub-question C is
based around how important respondents feel knowing the originating source of a message is,
and its impact on credibility. While this question was in part answered using a simple scale of
preference (quantitative), understanding the linkage between source and credibility is perhaps
better understood via respondent descriptions and qualitative reporting and is also the subject of
future research proposed by Castillo, Mendoza and Poblete (2011). Finally, sub-question D
measured the appetite for social media use for emergency management in Vancouver and was
examined quantitatively along a numerical scale of preference. Widener, Horner, and Metcalf
contend that future research should “…investigate empirical forms of online social networks as
well as how much weight is placed upon information garnered from social media “ (2012, p.
207). This research need will in part be satisfied by the qualitative assessments outlined above. It
is hoped that the combination of quantitative and qualitative methods employed in this MRP will
produce some recommended best practice for use of social media by emergency managers in
Vancouver.
Research Design
Research Methods
The following research methods were used:
a. Online self-administered survey using survey software.
b. One-on-one interviews.
c. Administered written simulated warning message assessment.
Conduct of Survey
The initial intention was to distribute both an identical online and hardcopy survey to capture
the widest possible audience. However, it was correctly anticipated that the greatest return was
for the version of the survey that was electronically produced and distributed using an Internet-
SOCIAL MEDIA AND EMERGENCY WARNINGS
30
based survey platform. The written survey and the Internet-based survey were identical. Part of
the survey asked about social media usage patterns (types of social media used, degree of access
at home and at work, usage of automatic updates, and why the respondent chooses the social
media platforms that they do). This section also included relative assessments by the user of the
degree of perceived general credibility for each of the types of social media they use. Much of
this section was represented numerically with responses to either yes or no questions or a 5-point
scale of preference, with some qualitative comment from the respondent. The survey continued
with an opportunity for comment regarding how the credibility of these forms of social media
could be improved and credibility enhanced regarding dissemination of emergency warning
information. The commentary section provided informants the opportunity to amplify and
provide further detail on their quantitative and qualitative responses. Respondents were asked
how important they feel knowing the originating source of a message found on social media is
and the impact on credibility of knowing the source. Respondents answered this question on both
a numerical scale of preference and through qualitative commentary. A complete copy of the
survey can be found at Appendix A. A consolidated list of survey respondent data can be found
at Appendix B.
Conduct of Interviews and Simulated Warning Messages Assessment
Six respondents (or just over 5% of the survey group) were also randomly selected for a
subsequent follow-up interview and for exposure to a series of four simulated emergency
warning messages on different social media applications. Interviews were conducted in person
and recorded. The basis of each interview was the completed survey with encouragement for
interviewees to expand on their quantitative assessments and survey comments in order to gather
a greater depth of qualitative data. Interviewees were then exposed to four simulated emergency
messages from a municipal emergency management organization. The four simulated test
SOCIAL MEDIA AND EMERGENCY WARNINGS
31
messages can be found at Appendix C and a copy of the Simulated Message Assessment Sheet
with consolidated responses can be found at Appendix D. The forms employed for the four
simulated warning messages included a link to a municipal website posted via text message on a
cell phone, a Facebook post to a municipal website, a Twitter feed warning, and a warning
message distributed by email. All of these messages were simulated and this was explained in
advance to respondents. Respondents were then invited to quantify the degree of perceived
credibility for each message using a 5-point scale of preference and their likelihood of complying
with or changing their behavior in response to that warning. Respondents were then invited to
provide their opinion of how these methods could be made more credible and the likelihood they
would pass that message along to their own social network.
Survey Group Composition
The survey group was drawn exclusively from residents of the City of Vancouver, British
Columbia, Canada. The study was only conducted in English and therefore only involved those
able to read and write fluently in English. Survey participants were selected predominantly
based on availability and convenience. The aspiration was to have a survey group of a minimum
of ten subjects from each of four different age groups (18-25 years, 26-40 years, 40-65 years, and
over 65 years). This was to gather a spread of opinion from across the adult population and to
provide sufficient representation to ensure statistical rigor. It was hoped that this aspiration
would be met and even exceeded by the distribution of the survey instrument via the Internet.
The survey respondents consisted of a mix of males and females within each age group.
However, the over 65 years of age group was under-represented. A comparison of the current
Vancouver census by age group against respondent age group for this survey can be found in
Chapter 4. Online surveys were distributed via email and it was hoped that a snowball strategy
would enable the widest possible audience within Vancouver. The link to the survey was
SOCIAL MEDIA AND EMERGENCY WARNINGS
32
distributed via email through friends, family, work colleagues, and the social media website
linkedin. The primary caveat was that adult residents of Vancouver only complete the survey.
The initial study design included an approach the City of Vancouver Emergency Management
Department to request permission to attend the emergency management workshops run for
Vancouver residents by the Department (2-3 per month). These workshops were normally
conducted on weekday evenings in community centers and schools and attended voluntarily by
residents of the City of Vancouver hoping to broaden their knowledge of emergency
preparedness. It was hoped that the City of Vancouver would permit the distribution of written
surveys to attendees of these workshops or alternatively, copies of the electronic link to the
online survey for workshop attendees who were unwilling or unable to complete the survey in
writing while at the workshop but willing to do so subsequently. Also the City of Vancouver’s
permission was to be sought to conduct one-on-one interviews amongst workshop attendees. The
plan was to draw no more than half the survey participants from the City of Vancouver
workshops. The rationale for this is that attendance at the City of Vancouver workshop indicates
a possible interest in emergency management. This would have meant that some percentage of
survey participants would have had a level of personal interest in emergency management and
potentially some knowledge of how emergency warnings work and where they come from.
However, at least 50% of the respondents were to be drawn from participants not attending the
City of Vancouver workshops and without any kind of obvious emergency management interest
to yield a cross-section of opinion based upon differing levels of emergency management
knowledge and interest. The aspiration was to conduct the one-on-one interviews at the
sponsored workshops. The aim would be to not detract from the flow of the City of Vancouver
workshops and interviews were to be conducted after the workshops had concluded or at alternate
venues at a mutually agreed upon time for those willing participants unable to spare additional
SOCIAL MEDIA AND EMERGENCY WARNINGS
33
time for interview immediately after the workshop. Regrettably, the City of Vancouver
discontinued its program of emergency preparedness workshops a few weeks before the request
for assistance could be made by the researcher. The City of Vancouver now only runs periodic
training sessions for business and community groups by request, and the program was only just
being established during the period when this MRP survey was in the field. Consequently, the
workshops were not used and the focus of the survey became entirely online. Respondents to the
survey were also asked if they would be willing to partake in interviews and the simulated test
message assessment phase. Respondents were randomly selected for the interview and simulated
warning message assessment phase based upon willingness, convenience, and availability.
Timeframe of Study
The electronic survey was available online between 14 November 2013 and 02 January 2014,
or a total of 50 days. The follow-up interviews and test message assessments overlapped with the
final few weeks of the online survey and occurred between 22 Dec 2013 and 10 January 2014.
Participants for the test message assessments were selected at random as outlined above. The
initial research plan called for the researcher to attend a series of City of Vancouver emergency
management workshops between 04 Nov and 13 December 2013, but the last minute cancellation
of the workshop program prevented this. Instead, in addition to the online survey, one-on-one
interviews were conducted in person at a mutually acceptable venue (researcher’s home, coffee
shop). Each interview lasted approximately 20-30 minutes.
Data Analysis
Part of this project examined simple preference with regard to the use of social media. This
was summarized using quantitative methods and analysis. As this project is limited in scope,
quantitative analysis will allow some conclusions to be drawn about the wider Vancouver
population through the use of statistics and probability (O’Leary, 2010). The remaining part of
SOCIAL MEDIA AND EMERGENCY WARNINGS
34
this project goes beyond statistical preference and examines informant descriptions of how social
media can be made more credible and appealing as an emergency management tool. Qualitative
assessment of both the descriptive responses to surveys and interviews are best suited to this part
of the project, as informants were asked open-ended questions and their opinions to build upon
their ratings and preferences. Qualitative data was then coded using descriptive, topical and
analytical coding. Several of the questions were examined using both qualitative and quantitative
means. This allowed for some triangulation of results to aid in the verification of result validity
(Richards, 2009). Survey data was also triangulated against respondent reactions to the simulated
warning messages contained within the survey. This aided in validating one part of the survey
against the other. Although beyond the scope of this project, ideally, the qualitative conclusions
drawn from the survey could be operationalized in a further quantitative study to verify their
validity and produce a grounded theory (Richards). Finally, the focused interviews conducted
with an element of the survey group served to expand upon responses and verify the validity of
survey answers through further triangulation. Interviews were coded as outlined above. These
multiple levels of triangulation ensured a rigorous process that hopefully produced trustworthy
results yielding pertinent recommendations for emergency managers in Vancouver. The methods
of assessment and triangulation are represented below in figure 3.1 and table 3.1 respectively:
Figure 3.1
Methods of Assessment
Sub-‐question A: Social media used
Sub-‐question B: Credibility
Quantitative reflection of social media used by Vancouver residents. Comment on why that type of social media is used.
Quantitative assessment of the credibility of each form. Qualitative reflection on what makes each social media platform credible or lacking in credibility as applicable.
SOCIAL MEDIA AND EMERGENCY WARNINGS
35
Table 3.1
Methods of Triangulation Data Gathered: To be Triangulated against: Sub-question A: Usage Sub-question B, simulated messages and interviews. Sub-question B: Credibility Simulated messages and interviews. Sub-question C: Source Sub-question A & B, simulated messages and interviews. Sub-question D: Receptiveness Simulated messages, interviews and sub-question A. Simulated Warning Messages Sub-questions A, B, C and D. One-on-one interviews Sub-questions A, B, C and D and simulated messages. Presentation of Results
Results were summarized in the form of a statistical analysis of social media usage by age
group and gender. Also a quantitative and qualitative analysis was conducted to ascertain the
existence of any demographic trends in terms of credibility perception regarding dissemination of
emergency warning information. This was followed by any trends noted in the manner in which
social media can be enhanced when employed to disseminate emergency warning messages to
Vancouver residents. The objective was to note any clear trends in order to produce
recommendations for social media usage by emergency managers, a research need supported by
Hui et al. in their study of information flow on social media (2012).
Sub-‐question D: Receptiveness
Sub-‐question C: Knowledge of source
Quantitative rating of how important knowing the source is to respondents. Qualitative comment on why knowing source is important.
Quantitative rating of how amenable respondents would be to receiving emergency warnings via their social media.
Simulated Warning Messages
Quantitative rating of message credibility. Qualitative description of why respondents rate their responses the way they do.
One-‐on-‐one Interviews
Further qualitative description of responses to sub-‐questions A, B, C and D and responses to simulated messages.
SOCIAL MEDIA AND EMERGENCY WARNINGS
36
Limitations of the Study
a. Convenience sampling was used.
b. Limited time period for study in accordance with Royal Roads University (RRU)
assignment schedule.
c. Ethnic background was not considered for recipients.
d. Socio-economic background and education level was not considered for recipients.
e. Unethical to create the level of anxiety in respondents that accompanies a warning
message, which could effect accurate reaction to simulated warning messages.
Delimitations of Study
a. Only English-speaking survey participants were used.
b. Only Vancouver residents (not those who work in Vancouver and reside outside).
c. Emergency management workshop schedule cancelled at last minute by City of
Vancouver causing a reliance on online survey.
d. Online survey relies on respondents with Internet access.
Assumptions
a. Participants use social media and have Internet and computer access.
b. Participants have a rudimentary knowledge of the human and environmental hazards they
face as residents of Vancouver.
c. In-line with existing research, credibility is a key driver in considering any emergency
management warning.
Ethical considerations
As this MRP involved human subjects, it complied with both RRU and the Tri-council
Policy Statement for ethical standards for research. Each participant was provided a consent
form and had the option to terminate their involvement in the survey at any time. The consent
SOCIAL MEDIA AND EMERGENCY WARNINGS
37
form explained the aim and format of the survey and how the information would be controlled
and presented. Participants were coded to protect anonymity and this final report does not
contain any information that would suggest the identity of informants. Lastly, the simulated
emergency messages used were clearly explained prior as being simulations only, and that no real
emergency threat was imminent.
SOCIAL MEDIA AND EMERGENCY WARNINGS
38
Chapter Four: Study Findings
Respondent Demographics
There were a total of 102 respondents to the survey. Frequency distribution for respondent
age, both by gender and combined, can be found in table 4.1 below:
Table 4.1
Frequency Distribution by Age Group and by Gender
Age Group % Male % Female % of Sample Population 18-35 5.88 11.76 17.64 36-50 28.43 38.23 66.66 51-65 5.88 8.82 14.70
Over 65 0 0.98 0.98 Total 40.2 59.8 100
Questions one and two of the survey asked respondents to provide gender and age group
information. In terms of gender, 61 females and 41 males responded to the survey. Although
respondents represented all of the four prescribed age groups (18-35, 36-50, 51-65, and over 65
years of age), the majority of respondents fell into the group 36-50 years of age (two thirds of the
sample population). This proportion seems quite high and is possibly a consequence of the fact
that I distributed the survey via email to friends and colleagues, many of whom are of a similar
age group to myself. However, this will be discussed more fully in Chapter 5. The low
representation of the over 65 age group was initially disappointing but is reflective of the smaller
size of the over 65 age group in Vancouver. Based on the potential vulnerabilities faced by the
elderly in terms of emergency management, use of social media by this demographic is a worthy
area for future study. From discussion during the interview stage, the low uptake of Internet
usage by the over 65 age group appears to be more one of trust in social media and ease of use,
and less a factor of physical access to the Internet. To understand how representative the
SOCIAL MEDIA AND EMERGENCY WARNINGS
39
composition of the survey group is in relation to the population of Vancouver, the above data can
be compared with the 2011 Statistics Canada Census for the Vancouver census metropolitan area
(CMA). In this study, the total population (both male and female) was determined to be
2,313,330. Although Statistics Canada used a slightly different division of age groups, the
breakdown of population (both male, female and combined) by age is shown below (Statistics
As can be seen in the Vancouver census, the combined number of males and females as a
percentage of the overall population is approximately 21% each for the 0-19, 20-34 and 50-64
years of age groups. The 35-49 years of age group is slightly larger at just over 23%. The
smallest group in the Vancouver Census is the over 65 years of age group at 13.53%, though the
MRP survey respondent group in this age category was considerably smaller at only 0.98%. The
MRP survey group also records the 35-49 years of age group as the largest, but by a much greater
degree. In the MRP survey, the 35-49 group was almost 66% of the overall survey group. Figure
4.1 below represents the breakdown of survey respondents by gender and age graphically.
SOCIAL MEDIA AND EMERGENCY WARNINGS
40
Figure 4.1. Respondents by Age and Gender
Questions Three to Seven: Internet Usage
Question three of the survey asked whether respondents have access to the Internet at home.
All 102 respondents answered with a unanimous “yes”, unsurprising as the survey was
distributed via the Internet. Question four asked respondents if they had access to the Internet at
their place of work. The clear majority (87% or 89 people) does, with only seven respondents
saying they did not and six answering ‘not applicable’. Figure 4.2 below shows the breakdown
of those respondents who have Internet access at their place of work.
18-‐35 years 36-‐50 years 51-‐65 years over 65 years Male 6 29 6 0 Female 12 39 9 1
0
5
10
15
20
25
30
35
40
45
Freq
uenc
y
SOCIAL MEDIA AND EMERGENCY WARNINGS
41
Figure 4.2. Internet Access at Work
In some workplaces, it is common for an employer to electronically restrict Internet access for
employees. This could be to limit liability in terms of employee access to illegal websites, or it
could be to safeguard the company’s servers and electronic information from hacking by an
external source. Access to some social media sites may be restricted to prevent distractions for
employees and ensure productivity. Question five asked those respondents who do have Internet
access at their place of work, whether their employer restricts their access in any way. All
respondents answered this question and over half (or 55 respondents) said their employers do not
restrict their access to the Internet. Figure 4.3 below shows the breakdown of employers who
regulate their employee’s access to the Internet at work.
87.4%
6.8%
5.8%
Yes
No
Not Applicable
SOCIAL MEDIA AND EMERGENCY WARNINGS
42
Figure 4.3. Internet Access at Work Regulated by Employer
Despite restrictions placed on Internet usage by an employer, the availability of
inexpensive portable devices make accessing the Internet on the move easy and almost
instantaneous. Over 90% of survey respondents routinely access the Internet using a mobile
device. In practical terms, this means that 91 of the 102 respondents use a mobile device and have
portable, potentially instantaneous access to social media and the Internet at both work and home
and on the move. Figure 4.4 below represents this graphically.
38.2%
61.8%
Yes
No
SOCIAL MEDIA AND EMERGENCY WARNINGS
43
Figure 4.4. Use of Mobile Internet Device
However, access to the Internet at work, at home or on the move is of limited use unless a
person makes a conscious choice to use that access. A person has to actually seek out a website
or social networking application through logging-on, possibly requiring a password or individual
user identification. Alternatively, an Internet user could pre-arrange access to certain websites
and with relative ease, permit that website to send them information or alerts of relevance as soon
as they meet certain criteria imposed by the user. The user can tailor what types of information
they get, how frequently they receive it and how they are alerted to the presence of that new
information. Automatic-forwarding from one email address to another or from one device to
another can increase the probability of a message getting to a user anywhere and at anytime.
These automatic ‘push’ notifications can be used for everything from breaking news and
91%
9%
Yes
No
SOCIAL MEDIA AND EMERGENCY WARNINGS
44
changing weather, to changes in stock prices or mention of your own name on a certain website
by a third party. Figure 4.5 shows the breakdown of respondents who currently subscribe to push
notifications via the Internet. In practical terms, this means that 52 respondents of the sample
population of 102 actively seek and receive information from the Internet and social media that
they feel is relevant to them via push notifications.
Figure 4.5. Use of Push Notifications
Questions one to seven served to illustrate the general Internet usage characteristics of the
survey group. In the case of this survey, all respondents have access to the Internet at home,
most have access at their place of work, and most have access through a mobile device.
However, even with all this opportunity to access the Internet, half choose to control their own
access ‘manually’. There does not appear to be a wealth of information or study on the appetite
51%
49% Yes
No
SOCIAL MEDIA AND EMERGENCY WARNINGS
45
for and usage of push notifications in social media at present. Although it is difficult to
determine respondent motivation conclusively, the survey group appears to be equally split
between those who want to control their own access to information by consciously seeking out
what information they need as they need it, and those who prefer to make that decision in
advance and receive information as soon as it is available automatically via push notifications.
Ease of use, degree of control, confidence in the accuracy of information and information
overload may all be factors that affect a person’s choice to use push notifications. This is another
potential area for study, examining motivations for the utilization of push notifications and how
to increase uptake.
Questions Eight to Ten: Social Networking Applications and Credibility
Question eight asked respondents what social networking applications they currently use. There
was no limit to the number of social networking applications a respondent could select and
respondents had the option of adding other sites that were not listed. Figure 4.6 shows the
breakdown. Email from a known contact is in the clear majority (97 respondents or 95%)
followed closely by Facebook (81 respondents or 79%). YouTube (77 respondents or 75%) and
linkedin (71 respondents or 69%) are similarly popular and were generally part of a pair for any
given respondent. In other words, if a person uses linkedin, they seemed likely to use YouTube
and vice versa. Twitter (38 respondents or 37%), email from an unknown source (45 respondents
or 44%) and Google+ (35 respondents or 34%) were the next most popular. The remaining social
networking applications were of limited popularity and were mostly used by 1-2 respondents or
1-2% of respondents only. Responses can be seen below in Figure 4.6.
SOCIAL MEDIA AND EMERGENCY WARNINGS
46
Figure 4.6. Social Media Applications Used
Question nine asked respondents how they rate each of the social networking applications they
currently use in terms of credibility. A five-point scale of preference was used to assess whether
each application was credible or lacking in credibility in the opinion of the user. Email from a
known contact scored the highest with 92 respondents or over 95% of those who used it stating
they found it to be ‘very credible’ (58 respondents or 60%) or ‘somewhat credible’ (34
respondents or 35%). By contrast, just below 17 respondents or 28% of those who used it, found
email from an unknown source to be ‘very credible’ (3 respondents or 5%) or ‘somewhat
credible’ (14 respondents or 23%). Less than 59% or 48 respondents who use Facebook would
rate it as ‘very credible’ (8 respondents or 10%) or ‘somewhat credible’ (40 respondents or 49%)
0
20
40
60
80
100
120
Freq
uenc
y
Social Media Applications Used
SOCIAL MEDIA AND EMERGENCY WARNINGS
47
and 53% or 26 respondents who use Twitter regard it as ‘very credible’ (7 respondents or 14%) or
‘somewhat credible’ (19 respondents or 39%). By contrast over 83% or 55 respondents who
currently use linkedin rated it as ‘very credible’ (12 respondents or 18%) or ‘somewhat credible’
(43 respondents or 65%). This is perhaps a reflection of linkedin being designed and advertised
as a professional network as opposed to a social network. Figure 4.7 below represents this
graphically.
Figure 4.7. Assessment of Credibility of Each Social Media Application Used.
Figure 4.8 shows the data retrieved from question nine in another way. In this graph, only
the top two selections for each social networking application (‘very credible’ and ‘somewhat
credible’) are considered. This combined total is used as an overall credibility rating. That
0
20
40
60
80
100
120
Freq
uenc
y
Social Media Applications Used
Very Lacking in Credibility Lacking in Credibility Neutral Somewhat Credible Very Credible
SOCIAL MEDIA AND EMERGENCY WARNINGS
48
credibility rating is then divided by the size of the survey group to suggest an overall percentage
ranking of credibility for each social networking application used. The applications are then
arranged in descending order of perceived credibility.
Figure 4.8. Credibility Rating by Social Media Application.
From this graph, there appear to be four main groupings:
1. Email from a known sender: 90+% credibility rating
2. linkedin: 55%, and Facebook: 48% credibility rating
3. Twitter, YouTube, Google+ and email from an unknown sender: between 25%
and 15% credibility rating
4. Remaining applications used: less than 9% credibility rating
0 10 20 30 40 50 60 70 80 90 100
Email (known sender)
linkedin
Facebook
Twitter
YouTube
Google+
Email (unknown sender)
Flickr
Instagram
Tumblr
Wikispaces
Vimeo
Vine
Four Square
HeyTell
Skype
Individual Blogs
Flip Board
Google Circles
Snap Chat
% C
redi
bilit
y R
atin
g
Social Media Application Used
Credibility
SOCIAL MEDIA AND EMERGENCY WARNINGS
49
Question ten asked respondents how likely they would be to alter their personal behavior in
response to receiving an emergency message from each application. Over 97% of those who
used email from a known sender stated that they were ‘very likely’ (46 respondents or 70%) or
‘somewhat likely’ (18 respondents or 27%) to change behavior. Interestingly, this represents a
5% rise over the number who rated email from a known sender as being ‘very credible’ or
‘somewhat credible’. Similarly, 67% of respondents who use Facebook stated they were ‘very
likely’ (9 respondents or 15%) or ‘somewhat likely’ (31 respondents or 52%) to change behavior,
an increase of 7% of those who scored Facebook as ‘very credible’ or ‘credible’.
Figure 4.9. Likelihood of Behavioral Change by Social Media Application.
Generally, if a person rated a social networking application as credible in question nine, that
translated into an increased likelihood that the respondent would alter their behavior in response
0
10
20
30
40
50
60
70
Facebook
Twitter
Email (known sender)
Email (unknown
linkedin
YouTube
Flickr
Instagram
Tumblr
Google+
Vimeo
Vine
Wiki Spaces
Snap Chat
Foursquare
HeyTell
Skype
Individual Blogs
Flip Board
Google Circles
Freq
uenc
y
Social Media Application
Very unlikely
Somewhat unlikely
Neither likely nor unlikely
Somewhat likely
Very likely
SOCIAL MEDIA AND EMERGENCY WARNINGS
50
to an emergency warning message promulgated via that medium. Figure 4.9 above shows the
breakdown of responses for each application. This linkage between credibility and behavioral
change extends back further to question eight (social networking application usage). The pattern
emerging consistently across all age groups and both genders of the survey group is that the most
used social networking applications are regarded as the most credible and in turn are the most
likely to cause behavioral change by the user in response to an emergency warning posted on that
application.
Questions Eleven to Fourteen: Source versus Medium and Credibility
Question eleven asked respondents how important they feel that knowing the originating
source of an emergency warning message is when using a social networking application. Almost
90% felt that it was ‘very important’ (91 respondents) and the remaining 10% (11 respondents)
felt that it was ‘somewhat important’ (a combined total of all 102 respondents). Question twelve
asked respondents if knowing the originating source of the message made them more likely to
alter their behavior. Almost unanimously, the survey group agreed that knowing the originating
source made it more likely that they would alter their behavior in response with only 3
respondents or 2.94% stating the opposite. Question 13 asked respondents what in their opinion
contributes more to social media credibility: knowing the originating source of a message or the
social media application upon which the message is viewed. Virtually the same response was
received as with question 12 with 98 respondents or 96% stating that knowing the originating
source of the message was the more important factor. Lastly, in question 14, respondents were
asked it they would like to receive emergency warning information from local, provincial or
federal government via their social networking applications. A total of 94 respondents or over
92% stated that they would with only 8 respondents or less than 8% stating that they would not.
SOCIAL MEDIA AND EMERGENCY WARNINGS
51
However this is a significant increase over the number of respondents who currently subscribe to
push notifications (a little over 50% as determined in question 7).
Subsequent Interviews and Simulated Message Assessment
Six respondents were randomly selected for further interview and for exposure to four
simulated emergency warning messages. The demographic profile of respondents for this phase
included four females 36-50 years of age, one male 36-50 years of age, and one male over 65
years of age. The simulated messages can be found at Appendix C. Each respondent was asked
to expand upon their responses to the original survey and were then exposed to the four simulated
emergency warning messages. Respondents were asked to rate each simulated message in terms
of credibility, likelihood of altering behavior and likelihood of passing the information along.
Respondents were also asked how they were likely to pass the information along if they indeed
chose to do so. The consolidated responses can be found at Appendix D. Figure 4.10 below
represents the results of the test message assessments graphically. In this graph, the blue line
represents test subjects who found each message ‘very credible’ or ‘somewhat credible’. This
line represents an assessment of credibility. The red line represents those test subjects who felt
themselves ‘very likely’ or ‘somewhat likely’ to alter their behavior in response to each message.
This represents an assessment of behavioral change. The green line represents those test subject
who felt themselves ‘very likely’ or ‘somewhat likely’ to promulgate the message further via his
or her own social networks. This represents an assessment of promulgation.
SOCIAL MEDIA AND EMERGENCY WARNINGS
52
Figure 4.10. Simulated Messages Ratings.
Email remains the most highly rated in terms of credibility, likelihood of behavioral change
and likelihood of further promulgation. Twitter and Facebook are broadly equivalent and come
next in order of precedence in terms of credibility, likelihood of causing behavioral change, and
likelihood of further promulgation. Text messages however are seen as the least credible, but
interestingly equal to Facebook and Twitter in terms of likelihood of causing behavioral change.
Coding of Qualitative Interview Data
Commentary provided by respondents during the online survey and interviews was subjected
to descriptive coding in terms of the primary two demographic descriptors in the survey: age
group and gender. Those responses were further coded by subject (topical coding), and finally by
0
1
2
3
4
5
6
7
Email Twitter Facebook Text Message
Freq
uenc
y
Message Medium
Credibility
Behavioral Change
Promulgation
SOCIAL MEDIA AND EMERGENCY WARNINGS
53
theme (analytical coding). The key elements for the topical coding were derived from specific
comments made by respondents. For example, in the male 18-35 years of age group, several
respondents suggested a preference for warnings to be automated. This topic was repeated
several times by different respondents within this group using similar language. As such, it
became a heading under topical coding. A second topic mentioned in the same age and gender
group was ease of ability to authenticate a message and a sender. These two also became
headings under topical coding. Next, during the process of analytical coding and using the
examples above, the prevailing themes that underline automation and the facility to authenticate a
message and sender were identified. One of the prevailing themes chosen that seems to underpin
the above two topics is simplicity. So in essence, the respondents determine the topics that are
coded through their specific comments during the survey and in interview, and the researcher
analyzes the common themes and principles that underpin those topics to establish the headings
for analytical coding. The objective is to identify those common and recurrent themes that
emerge. A detailed breakdown of the descriptive, topical, and analytical coding can be found at
Appendix E. The table below provides a summary of the key themes by gender and age.
Table 4.3
Summary of Analytical Coding by Age and Gender
Descriptive Coding Analytical Coding Age group 18-35
Male Simplicity Ease of use Legitimacy Female Legitimacy
Age group 36-50
Male Simplicity Ease of use Legitimacy Engagement Utility Lack of expense Trust
Female
Legitimacy Trust Verifiable Engagement Message saturation Ease of use
Age group 51-65
Male Legitimacy Message saturation Simplicity Female No comments made
Age group over 65
Male No comments made Female Legitimacy Simplicity Engagement
SOCIAL MEDIA AND EMERGENCY WARNINGS
54
Summary
So in summary, the survey, interviews, and simulated messages reveal the following key
findings:
a. The survey group appears to have comprehensive unrestricted access to the Internet at
home, work and when mobile (in other words virtually 24/7).
b. The survey group is almost equally split on their use, or lack of use of automated (push)
notifications.
c. The most utilized social media applications consistently by age and gender include email
(from a known sender), Facebook, YouTube, linkedin, email (from an unknown sender),
Google+, and Twitter.
d. The most utilized social networking applications are viewed by respondents as more
credible and more likely to produce behavioral change in response to a warning message.
e. Knowing the originating source of a message is highly important to the user and likely to
impact on their decision to change their behavior in response.
f. There appears to be an appetite for emergency warning information via social media
amongst the survey group.
g. Preferred characteristics of a warning message via social media include: legitimacy, ease
of use and authentication, no expense, brevity, and prior engagement by emergency
managers. In the 18 to 35 age group, men prefer simplicity of message and ease of use,
whereas females regard message legitimacy as important. As age increases legitimacy
becomes more important across both genders.
h. Respondent views on how to enhance credibility are less to do with the medium and more
to do with the source of the message and include: use of trusted sources like government
or the emergency services, use of conventional media to authenticate messages, as a
SOCIAL MEDIA AND EMERGENCY WARNINGS
55
source or to promulgate further, use of existing social networks, and saturation of the
target audience with the message. These recommendations are generally consistent across
both genders and across all age groups.
SOCIAL MEDIA AND EMERGENCY WARNINGS
56
Chapter Five: Discussion, Conclusions and Recommendations
Research Aim
Prior to discussion and interpretation of the data collected, it is worth re-visiting the
purpose and the sub-questions posed for this MRP:
The purpose of this study is to determine how emergency managers in Vancouver can
optimize the use of social media for the formal dissemination of warnings. In addition, the
following sub-questions will be addressed:
a. What are the most utilized forms of social media within Vancouver for personal use?
b. How are those forms of social media perceived in terms of credibility by the residents of
Vancouver?
c. Is there a relationship between the source of the message and the perception of
credibility?
d. Would Vancouver residents wish to receive emergency warning information via their
social media?
First, conclusions will be drawn regarding the survey group composition. Then each of
the sub-questions will be discussed in the context of the literature reviewed and with regard to
age group and gender. Finally, to answer the main research question posed, recommendations
will be drawn on how to optimize the use of social media in Vancouver for the promulgation of
emergency warnings. Also, some possible implications for future research will be discussed.
Demographics and Internet Usage
Survey questions one to seven dealt with (i) respondent demographics and (ii) Internet usage.
Table 4.1 summarized the breakdown of respondent age by gender. Table 4.2 showed the
breakdown of the Vancouver population by age and gender according to the 2011 Vancouver
SOCIAL MEDIA AND EMERGENCY WARNINGS
57
census. It is interesting to note that almost two thirds of survey respondents came from the age
group 36-50 years. As can be seen in the Vancouver census, the combined number of males and
females as a percentage of the overall population is approximately 21% each for the 0-19, 20-34
and 50-64 years of age groups. The 35-49 years of age group is slightly larger at just over 23%.
The smallest group in the Vancouver Census is the over 65 years of age group at 13.53%. The
MRP survey group also records the 35-49 years of age group as the largest, but by a much greater
degree. In the MRP survey, the 35-49 years of age group was almost 66% of the overall survey
group. This much higher representation of respondents in the 36-50 age group could be a
consequence of it being the age group in which the researcher belongs. Potentially, the initial
distribution of the survey via peers, family and co-workers focused the survey at people in a
similar social network, age group, socio-economic demographic etc. as the researcher. However,
as initial invitees to the survey were encouraged to forward the survey onwards via social media
and email, it is acknowledged that this may have consequently brought the survey to a wider
spread of respondents of varying ages. Also, although the origins of the Internet can be traced
back to the 1960s, the Internet’s explosion into mainstream popularity began in the mid-1990s.
That would make the 36-50 years of age survey group approximately 16-30 years of age at the
time when the Internet’s popularity began to gather significant momentum. So the most
prominently represented part of the survey group is essentially the first generation to have ‘grown
up’ with the Internet. In terms of gender, the Vancouver census is generally equally split with
each age group being composed of 48-52% representation from each gender. In the MRP survey
however, females are consistently the larger part of each of the respondent age groups. It is
unclear why more females chose to respond to the MRP survey than are recorded in the census,
but if this study was to be repeated or expanded upon, a purposive sampling strategy to ensure
that the survey respondents were more equally split along the lines of gender should be
SOCIAL MEDIA AND EMERGENCY WARNINGS
58
considered. This would then more accurately reflect the current gender breakdown in Vancouver
according to the 2011 census.
With regard to Internet usage, as the survey was distributed predominantly via email and the
Internet, this may have artificially created a higher number of respondents who reported having
access to the Internet. In other words, respondents received the survey link via email, and
therefore, by default, would report having access to the Internet. This does represent a constraint
of the survey as identified in the limitations section in Chapter 3. However, regardless of from
where respondents received the survey (at home, at work, verbal request from the researcher, or
via a mobile device), a sizeable proportion of Vancouver residents appear to have access to the
Internet. This is perhaps indicative of the overall popularity of Internet use in British Columbia
where over 85% report having access to the Internet and use the Internet regularly (Statistics
Canada, 2010). The promulgation of emergency warnings to those members of the population
without Internet access is a worthy area of study, but outside of the remit of this MRP. All MRP
respondents reported having access at home, almost all have access to the Internet via a mobile
device and over half have unrestricted access to the Internet at work. The popularity of mobile
devices alone represents a potent tool for the passage of warning messages to individuals (Ghosh,
2012). Over 90% of MRP respondents report having access to the Internet via a mobile device.
Although not examined in this MRP in detail, even cellular telephones without Internet access
may have utility for message promulgation through short message service (SMS) or ‘text
messages’. In terms of this MRP, the conclusion that can be drawn is that the Internet represents
a potentially instantaneous and highly adaptable means of message promulgation between
emergency managers and the majority of the population of Vancouver.
However, in terms of ensuring the rapid passage of warning information, a major stumbling
block for this survey group appears to be their willingness to use automatic push notifications.
SOCIAL MEDIA AND EMERGENCY WARNINGS
59
The survey group is almost exactly split on their use at present, which could impact on the
timeliness of a warning message being received via social media. There also do not seem to be
any significant differences in preference on the use of automatic notifications by age or gender.
This response is reinforced almost exactly by the findings of the Canadian Red Cross survey on
social media that suggests that 49% of all Canadians would sign up for email and Internet alerts if
there was a significant emergency expected or occurring (2012). This figure is based on
responses from both those with frequent and reliable access to the Internet and those without.
For the interview phase, six willing survey respondents were selected at random based on their
willingness to participate and their availability. They were each exposed to a number of
different messages via social media and an SMS message, each containing an emergency warning
of an impending snowstorm. In addition to rating those messages for credibility and the
likelihood of behavioral change as a consequence of reading the message, respondents were
asked to expand upon their responses to the original survey. Four females and two males were
interviewed representing the 36-50 and over 65 years of age groups. Due to constraints of time
and availability, it did not prove possible to obtain respondents representing both genders from
all four age groups. If this survey was to be expanded or repeated, representation from all age
groups and both genders should be sought. Generally during the interview phase, all respondents
suggested that they were weary of unsolicited information. Respondents also reported that they
were wary of a sender’s motivation and the prospect of signing up for a notification system that
passes on their information to a third party for other purposes like marketing or government
research was a significant concern. Respondent comments during interview reinforce the view
from the reviewed literature that information flow can be overwhelmingly (Crowe, 2012).
Internet based ‘scams’ are getting increasingly sophisticated and frequent, making it a challenge
SOCIAL MEDIA AND EMERGENCY WARNINGS
60
for users to sift through a large volume of emails and messages in order to make an informed
decision about whether to act upon information or even open an email and read it to begin with.
Internet users value their ability to maintain their privacy online. Privacy is a key issue that
has become significantly affected by the Internet (Clarke, 1999). According to a study by
O’Neil, privacy is consistently considered more important to Internet users than convenience.
However there appears to be only minor differences in the level of concern regarding privacy
across gender, education, income level and race (2001). The results of this MRP reinforce
O’Neil’s findings regarding gender but suggest that privacy is a greater concern to older
respondents. Quan-Haase and Young suggest that although the younger Internet and social
media user is concerned with privacy issues, they are more likely to use personal strategies to
protect their own privacy including fake names, alternate email addresses, use of privacy settings
and the blocking of certain users from viewing their personal information (2009). In other
words, the younger generation is perhaps more technically savvy and able to mitigate their
privacy concerns through technical means. If respondents are going to be convinced to subscribe
to automated notifications, a considerable amount of effort must be made to build trust amongst
the public and convince them that the service is legitimate, secure, and will be used only in the
event of an emergency likely to affect them adversely. Walaski (2013) reinforces these findings.
He suggests that organizational integration of social media into risk communication requires a
strategy developed well in advance of a crisis event and hinges on developing trust and credibility
with the audience.
Social Media Utilization
Veil suggests that one third of online adults are using blogs, social networking, video, text, and
portable devices (2011). This is reinforced in a survey by Pew Internet who suggests just over
30% of online adults do the same (2009). A survey by the Canadian Red Cross suggests that
SOCIAL MEDIA AND EMERGENCY WARNINGS
61
60% of Canadians use social media and that 94% of that figure use Facebook (Red Cross, 2012).
That would equate to just over 56% of Canadian adults using Facebook. This MRP would
suggest that either this figure might be much higher in Vancouver, or that usage in this particular
Vancouver sample is much higher. In this study over 97% of respondents use email to and from
known contacts and over 80% use Facebook. In the context of this MRP, comments during
interview show that respondents have interpreted email from a known sender as a source that the
respondent is previously familiar with (either a friend or relative, or a news agency or emergency
service that has a track record for credibility with the recipient). It is important to note that only
one respondent reported using only email and no other social media applications. However, this
represents less than 1% of the overall survey group. As such a small part of this survey group
exclusively uses email, this would mean that the other members of that respondent’s online social
network (email contacts) would statistically almost all utilize another social networking
application. Therefore, that first respondent who uses only email, will be indirectly connected to
other more popular social networking applications through their own email contacts. So despite
using only one social networking application, a person will almost certainly have a second order
connection to the more popular social media sites through their email contacts. Also, even if a
person predominantly uses a seemingly less popular application (for example only 34% of
respondents report using Google+), if that person uses even one other social networking
application, the chances of them receiving information indirectly increases. Measuring the exact
degree of influence is challenging due to ever-changing applications, interconnectedness of
individuals, users, and pages within the various social media systems (Crowe, 2012). Mapping
these complex secondary and tertiary social network connections is beyond the remit of this MRP
and has already been suggested as a topic for future research by Shankar (2008). The point
remains that in this survey group, the overwhelming majority of respondents routinely utilize at
SOCIAL MEDIA AND EMERGENCY WARNINGS
62
least one form of popular online social media. Furthermore, not actively using a particular social
media application does not prevent a person from receiving an email link and following it to that
website to see a warning message. If an Internet user receives a link to a social media platform
that they do not usually use, this adds another layer of decision-making into the passage of that
warning message. If a person does not use Facebook but receives a link from a friend to a
Facebook page of warning information, that receiver has the option to follow that link or discard
it, thereby breaking that warning message pathway. That decision will be impacted greatly by the
recipient’s individual assessment of the relative credibility of the person sending that link.
However, issues of credibility will be discussed in greater detail in the next section.
Several conclusions can be drawn from the responses of the survey group. First, there appears
to be ample use of social media applications that would warrant their use for the promulgation of
emergency warning information in Vancouver. Secondly, there are unsurprisingly several
applications that are more popular than others (Email, Facebook, Twitter, linkedin, YouTube,
Google+) and represent an effective link to the majority of the survey group. Email from a
known sender is the most utilized social media platform by the survey group. Therefore
encouraging recipients of a message by email to pass that message along via email or social
media applications is likely to reach the greatest proportion of a target population. Finally, due to
the overlap of applications and the majority of the survey group using more than one application,
greater message saturation will be achieved through posting information on multiple sites to
ensure numerous, possible converging pathways of warnings to the population of Vancouver.
This traditional layered approach to warning systems is commonly accepted in the literature as
being the most effective way to ensure reception of a message and enhance the chances of
understanding and compliance (Auf de Heide, 1989; Scanlon, 2007; Sene, 2008; Sorensen,
2000).
SOCIAL MEDIA AND EMERGENCY WARNINGS
63
Social Media Credibility
Sorensen suggests that there are six steps in the process of warning response: Hearing the
warning, understanding the message content, believing the warning is credible, personalizing the
warning to one’s own situation, confirming the warning is true and if others are taking heed, and
finally, responding with protective action (2000). The question of social media utilization,
encapsulated in Sub-question A, has in part looked at steps one and two, hearing and considering
the warning, by examining social media usage among the survey group. Sub-question B
examines the third step in the process, specifically, the issue of credibility with regard to social
media.
In terms of usage, email from a known contact, Facebook, YouTube and linkedin are the most
utilized with over 70% of respondents using them routinely. For some of these applications like
Facebook and linkedin, what may differentiate them from the remainder used by the survey
group, is that they enable a user to build a personality online. Facebook users build an online
profile and linkedin users build an online resume. With email from a known sender, to an extent,
the user is potentially already familiar with the identity, characteristics and qualities of the sender
in advance. Therefore, on some level, using these particular applications is akin to having a
conversation with a person you are familiar with. Although this is less the case in the event of a
purely business relationship conducted via email when one is corresponding with a customer
service representative, a bank manager or someone in a purely professional capacity, for
example. With an online profile however, even if you have not met that person physically, you
may have been exposed to a picture of that person online and engaged in online conversations
and exchanges of points of view. Therefore, a person using a social media application will have
an opinion (either conscious or sub-conscious) about the person from whom they receive a
message and the degree to which they trust that person. Dwyer, Hiltz and Passerini suggest that
SOCIAL MEDIA AND EMERGENCY WARNINGS
64
trust is a vital component to any social exchange and the degree to which you trust a person
determines the likelihood that you will enter into an exchange (2007). The creation of an online
profile as in Facebook and linkedin provides a user a more personalized rapport akin to meeting
someone in person. Of course a detailed online persona may not always be an accurate reflection
of the person with whom you are communicating. The earlier mentioned study by Quan-Haase
and Young suggested that some social media users mitigate their privacy concerns by using false
profile information, fake email addresses, and privacy settings (2009). Regardless, these social
networking applications have the potential to personalize warning messages and information as if
it is a single trusted source speaking directly to you (White, 2012). A person is more likely to
consider information and more thoroughly assess that information for credibility if the
information is personalized or relevant to them (i.e. geographically or professionally relevant or
likely to have some personal or financial impact on the recipient). Even if you have never met
the person with whom you are speaking via social media, applications like Facebook and linkedin
allow you to get to know that person ‘virtually’ prior to the receipt of any advice or warning
message. That in turn may create trust so that a message received is less impersonal and more
like ‘word of mouth’ that you would receive from a valued member of your social network (Veil,
2011). In their study of Twitter usage prior to and during an emergency, Hui at al. found that
usage is dominated by repetition of warning information with a high percentage of web addresses
and web links posted (information sharing) and a lower-percentage of replies and discussion
(2012). In other words, users are less inclined to engage in discussion on the impending
emergency and more inclined to absorb and then swiftly pass along warning information that they
find useful and relevant to others who may benefit from being informed. If Sorensen’s six-stage
warning model were utilized, this would suggest that prior to or during an emergency, social
SOCIAL MEDIA AND EMERGENCY WARNINGS
65
media users would have received information (stage one), understood it (stage two), assessed it as
credible (stage three) and repeated it to their own social network (stage six).
Question nine of the survey asked respondents how credible they found each of the social
media applications they used. The applications rated as most credible, were also the applications
rated as being the most utilized. So in summary, the responses to the first two sub-questions are
related in that the most utilized social media applications (sub-question A) were perceived as the
most credible (sub-question B). Email from a known sender, linkedin, Facebook, Twitter and
YouTube remain the most popular social media applications and are perceived as being the most
credible by the survey group.
Source versus Credibility
What is emerging is a link between credibility and popularity. The difficulty is determining
what came first: are these applications popular because they are credible or credible because they
are popular? And does one factor come first for all respondents or does it vary by gender or age.
As suggested above, what differentiates these more popular sites in part is the level of
personalization and the manner in which these popular applications facilitate ‘knowing’ the
person with whom you are communicating. This would suggest that knowing the source of the
message and being in a position to make a value judgment about how credible that person or
source is will guide the respondent’s attitude toward the overall assessment of the level of
credibility of that social media application. Both Facebook and linkedin are essentially social
networking applications, but arguably Facebook is designed more for fun and less for a
professional purpose like linkedin. The social networking application YouTube was rated by
only one third of users as being very credible or somewhat credible. This is interesting as
YouTube is essentially a social networking application to share video images. The popularity
and availability of hand-held high-resolution cameras in cell phones combined with mobile
SOCIAL MEDIA AND EMERGENCY WARNINGS
66
Internet access can allow virtually instantaneous uploading of footage from the scene of an
emergency to a platform such as YouTube, thus making that footage available for all to see.
However, a recipient has to be alerted to the presence of that footage being uploaded, whether it
is through a specific search conducted by a user or through receipt of a link from a contact, or
from an automatic push notification. Crowe suggests that unedited and raw footage of an
emergency situation or incident uploaded by a bystander onto an application like YouTube is
often considered highly convincing and authentic (2012). Yet the results of this survey suggest
otherwise. What may impact upon the credibility rating of YouTube is that in addition to
legitimate footage of actual events, the Internet and YouTube has become a haven for fake
footage, pranks, staged incidents and material that may be difficult to regard as credible. The
ease with which information can be posted to the Internet has made it a potent medium for social
protest and so called “guerilla communication” or just users seeking notoriety or having fun
(Schonberger, 2005, p. 286). In some cases, parody websites have been developed that mislead
web users into believing they are official sites (Johnson & Kaye, 1998). Some footage, even
legitimate footage, may seem so incredible that a viewer simply does not believe it to be true.
This is perhaps compounded by the fact that it is unlikely you will receive a link to some
YouTube footage captured by someone you actually know and perceive as credible, or of a
location that is known to you or geographically proximate and that is experiencing a disaster.
Instead, you may receive a link from a trusted friend to some YouTube footage filmed by
someone unknown to you of somewhere you have never been and of an incident that does not
directly affect you. This depersonalizes the process somewhat. Now, instead of receiving a
personalized warning message from a trusted contact, you are receiving an unverified piece of
footage that may or may not seem relevant or believable from a trusted contact. The absence of
a known source for the footage, even if passed on by a trusted contact, makes the footage seem
SOCIAL MEDIA AND EMERGENCY WARNINGS
67
less credible. As for the remaining applications assessed by the survey group, those social
networking applications with less than 10% usage were generally regarded as lacking in
credibility. This would further indicate a connection between popularity and perception of
credibility by a user. Perhaps, if a person sees that a great number of people are using a particular
application, they will sub-consciously consider it more credible on the basis of a simple majority.
If a person sees their online network using an application that they do not currently use, they may
consider using that application based on the trust and confidence they have in their contacts. But
it must also be considered that some applications may just be considered popular based
predominantly on their entertainment value to the user. An application may have very limited
credibility to the user but be highly utilized just for fun. It could be suggested that the usage of
these entertainment applications would perhaps drop significantly in times of an emergency as
the user’s appetite for credible and useful information increases.
Survey question ten asked respondents how likely they were to alter their behavior in response
to receiving a warning message via social media. The responses reinforce the trend emerging in
questions eight and nine of a connection between popularity and credibility. Responses to
question ten can be triangulated against responses received during the interview phase. During
interview, 66% (or 4 out of 6 interviewees) expressed a preference to have some way of verifying
the validity of emergency warning information viewed on social media. Suggestions included the
use of a follow-up call or email or a verification that only the source could provide. Others
suggested that push notifications be used from sources with established credibility like a news
agency or from the emergency services. This would mirror the statistical results that respondents
found email from a known sender to be the most credible in question nine of the survey. In the
context of the survey and one-on-one interviews, a known sender has been interpreted by
respondents as a source that the respondent is previously familiar with (either a friend or relative,
SOCIAL MEDIA AND EMERGENCY WARNINGS
68
or a news agency or emergency service that has a track record for credibility with the recipient).
It does not seem to have been interpreted by all respondents as a singular entity that is known to
the respondent personally. One interviewee described their preferred source for a warning
message as a “recognized authority” and another suggested “a known political figure”. In either
case, respondent data confirms that the credibility of the source is of more importance to the
respondent than the medium through which that message is transmitted. Mersham contends that
social media and sites like Facebook are challenging the notion of official sources and that
community forums are increasingly seen as reliable and authoritative (2010). This MRP supports
this in part, but interview data, although limited, still demonstrates the importance and impact of
a recognized and credible authority figure in affecting behavior and gaining compliance.
Some respondents both in the survey and the interview phase commented that it was important
to establish a line of communication prior to any emergency event and build that trust or rapport
before an emergency is imminent. This is a theme that can be found across the current literature
on emergency warnings and the Internet (O’Brien, 2012; Veil, 2011; Walaski, 2013). One survey
respondent termed this “an emergency identity” but cautioned about using it for anything other
than true emergencies. Botterell suggests that the public is remarkably tolerant of false warnings
being issued by government or emergency services, provided that they are promulgated in good
faith and not issued so frequently as to suggest negligence (2008). The fear of many respondents
during interview seemed to be more about the use of the Internet, social media, and personal
contact information for non-emergency purposes like opinion polls, political campaigning or non-
emergency bulletins of a commercial nature. Respondents suggested that any credibility would
be eroded quite quickly if message recipients felt that they were being taken advantage of by
authorities. These concerns are reinforced by the themes identified during the analytical coding
SOCIAL MEDIA AND EMERGENCY WARNINGS
69
phase and that include the need for legitimacy, the importance of trust, an ability to verify
information and the requirement to engage users in advance of an emergency occurring.
Several respondents in both the survey and the interview group made reference to having the
capability to verify warning information with at least one other source and the need for consistent
messaging across multiple sources. This finding is consistent with existing crisis communication
research that suggests the credibility of any message is increased if the same message is received
over multiple channels simultaneously (Botterell, 2008; Mileti, 2007; Scanlon, 2007). What the
initial warning message may serve to do is initiate a desire in the recipient to gain more
knowledge from other sources in order to verify that information. Interview respondents felt that
a warning via social media may serve as an initiator to prompt respondents into seeking
information on another medium or social networking application. Both survey and interview
respondents felt that the initial notification should be brief but also guide respondents to where
they can get more detailed information if they are interested. Respondents suggested that they
should be able to quickly and without cost forward the initial warning message onto friends and
family, thus capitalizing on existing social networks and increasing the spread of an initial
message exponentially. Social media can easily facilitate this rapidity. With the exception of cell
phone or data usage fees, forwarding messages is, for many, also free. Lastly, many respondents
to both the survey and in interview were weary of the many scams and ‘bulk’ emails they receive
and acknowledged that this has affected how they view push notifications and social networking
application credibility. One respondent commented, “Credibility has little to do with the
platform and more to do with the people using it and their motivations.” This goes back to the
point that if the source is trusted the message will likely be successfully conveyed no matter what
the medium. It is this potential for two-way dialogue and personalized connection between an
authority figure and the public via social media that enhances credibility (Seeger, 2006). What
SOCIAL MEDIA AND EMERGENCY WARNINGS
70
social media provides is the facility to engage in discussion, to question warnings, and to seek
additional information if desired. However, as discussed, during emergencies, information flow
is often more unidirectional with warnings being passed through social networks ‘one-way’ (Hui
et al, 2012). Although message recipients could respond, comment, or engage in discussion with
the trusted source of the message, in times of emergency they are less likely to do so.
So in summary, the source of any warning message appears to have a significant impact on the
recipient’s assessment of message credibility. Social media applications that are more
personalized and allow a recipient to be more familiar with a source make that medium more
credible in the eyes of that recipient, even if they have never actually met that source. What
social media does is allow a user to view and become familiar with an online identity, even if
they have never and will never meet that person in reality. A more personalized message from a
distant political figure or emergency services authority enhances credibility not only because of
that figure’s political standing or social ranking, but also because of the seemingly direct and
personal connection that social media application provides. And when such a message is
received second-hand from a trusted source, it has the potential to amplify further that credibility
rating in the eyes of the recipient. The key seems to be that information is attributable.
Recipients prefer to know where the information is coming from in order to apply their own
assessment of credibility in their decision to alter their own behavior. Perhaps this is why
sources like YouTube are deemed less credible as with even the most seemingly compelling
footage, it is often difficult to assess from whom exactly that footage has come and what that
person’s motivations really are.
Receptiveness
The final question of the survey asked respondents if they would wish to receive
emergency warning information via their social media applications. The answer was an
SOCIAL MEDIA AND EMERGENCY WARNINGS
71
overwhelming “yes” with over 92% suggesting they would. However currently, only half of
respondents subscribe to automatic push notifications via the Internet and social media. Based on
the survey and information gathered during interview, the social media applications that are the
most popular are also regarded as the most credible. Furthermore, the most popular and most
credible applications are also the most likely to cause behavioral change in response to a warning
message. What appears to be holding almost half of the respondent population back from using
automatic notifications is their level of trust in the motivations of the sender. Respondents also
stress the negative impact on credibility of an unsolicited message from an unknown source and
the need to establish an emergency identity in advance of any real warning message. So based on
this survey, it appears that there is indeed an appetite amongst Vancouver residents to receive
emergency warning message information via social media and the means available to emergency
managers to make that message credible. Residents have access to the Internet both at home, on
the move and at their place of work. Residents routinely use social media and have a somewhat
shared view of the relative levels of credibility of the social media applications they use. The
recommendations below may serve to contribute to the use of social media in the emergency
warning process and overcome the credibility stumbling block.
Conclusions
Social media can provide emergency managers in the City of Vancouver with a potentially
instantaneous and highly adaptable means of communicating with the majority of the respondent
population. This could be used to complement existing systems including televisions, radio and
print media, posted warning signs, emergency preparedness training, and the use of emergency
services to disseminate information directly. However, it should be noted that the availability of
Internet access reflected in this survey is slightly higher than that found in the other surveys
discussed in the literature, potentially due to the decision to promulgate this survey
SOCIAL MEDIA AND EMERGENCY WARNINGS
72
predominantly via the Internet. While there is an appetite for automatic emergency warning
messages, there is a fear on the part of respondents that it may lead to information overload or
open a gateway to third parties who will abuse the process. While certain social media
applications may seem more popular, greater message saturation will be achieved through posting
information on multiple sites to ensure numerous, possible converging pathways of warnings to
the target audience. Although respondents rate some social media applications as more credible
than others, this is likely more a consideration of the source of the messages being viewed on the
applications as opposed to the applications themselves. Message recipients prefer to know where
information is coming from in order that they can make their own assessment of credibility. Also,
recipients do not need to know a message sender personally for the message to be deemed
credible. A message can gain in credibility if the sender has an online identity that is trusted or if
the message is promulgated indirectly via a trusted contact.
Recommendations
To enhance existing emergency management warning systems in the City of Vancouver, the
following recommendations should be considered by emergency managers:
a. Emergency managers should consider email, linkedin, Facebook, Twitter, YouTube and
Google+ a priority for message promulgation via social media.
b. Emergency Managers must acknowledge the existing trepidation amongst half the
population to subscribe to automatic notifications at present. This figure is consistent with
survey findings from the wider Canadian population. Efforts should be made by
emergency managers in Vancouver to establish and maintain trust with residents via
social media in advance of the need to promulgate any warnings.
c. Emergency managers must be transparent about their motivations for using social media.
Criteria for use should be distributed so recipients know under what circumstances
SOCIAL MEDIA AND EMERGENCY WARNINGS
73
emergency management practitioners are going to use social media to promulgate
warnings.
d. Online contact information for recipients should never be shared with a third party who is
going to use it for non-emergency warning purposes and this should be made clear to
social media users.
e. Warning messages must be attributed, preferably to an individual in authority, such as an
emergency manager, political figure, or spokesperson for the emergency services.
f. Emergency managers who are going to distribute messages must create an emergency
management identity online. Recipients must feel they have the facility to engage with
and ‘get to know’ that warning authority (be it an individual or an organization) in
advance of any crisis.
g. Information should be personalized. That could be in the sense of it coming from a
familiar emergency management authority, or information that is geographically pertinent
to the recipient.
h. Existing social networks should be leveraged. In any promulgated message, recipients
should be encouraged to forward the message onto their own social network by email, as
email from a known sender is most utilized and perceived as the most credible by
respondents.
i. Warning messages should offer a means of verification and should be issued in concert
with multiple mediums to enhance credibility.
j. Any warning process should be free in terms of expense and limited in terms of the time
necessary to understand it. If overly complex or cumbersome, user uptake may be
limited. This is perhaps predominantly relevant in terms of mobile device usage and the
cost of sending and receiving text messages or downloading data.
SOCIAL MEDIA AND EMERGENCY WARNINGS
74
Implications
What this MRP has demonstrated is the presence of a readily available medium for message
promulgation and an appetite amongst Vancouver residents for its use. This research has also
reinforced much of the pre-Internet warning message theory and demonstrated its continued
relevance in the field. Also, the survey data has shown some similarities between recent research
by the Canadian Red Cross into social media usage and the findings of this MRP. The two areas
that have emerged as candidates for subsequent research are as follows:
a. Mapping the flow of warning information via social media from source to recipient could
pay huge dividends to emergency managers. Although this would be quite challenging
due to the ever-changing social media landscape, visualizing how information flows from
primary to secondary to tertiary contacts in a social network via social media could be
beneficial. Noting which routes are the quickest and are likely to produce the highest
level of behavioral change amongst users would provide some specific case study
information on how to pass a message quickly via social media.
b. This MRP has in part focused on the importance of credibility in social media. The next
step would be to focus on decision-making by recipients. Messages can be passed
quickly by social media and found to be credible by the user. However, does that
necessarily translate into action by the user? This research project has in essence looked
at the first half of Sorensen’s six steps in the process of warning response (hearing the
warning, understanding the message content, and believing the warning is credible). The
next stages deal with personalizing the warning to one’s own situation, confirming the
warning is true and if others are taking heed, and finally, responding with protective
action (2000). These next stages are in essence the decision-making that comes out of the
assessment of credibility a recipient makes upon receipt of a message. O’Brien suggests
SOCIAL MEDIA AND EMERGENCY WARNINGS
75
that there is currently no clear understanding of how social media can be used to assist the
public in making decisions that result in recommended actions (2012). To that end,
research into how to take that credible message and make it not just believable, but
convincing enough to support a decision to act by the recipient would be of significant
benefit.
SOCIAL MEDIA AND EMERGENCY WARNINGS
76
Chapter Six: Reflective Learning
To suggest that the MRP process was daunting is an under-statement! From the point
when I first considered taking the Master of Arts (MA) program until the point at which I started
the MRP, was an elapsed period of approximately five and a half years. One of the major
considerations I had when deciding on whether to embark upon the program was whether I would
be able to maintain the level of focus and effort to complete such a major undertaking as the
MRP. I would be lying if I did not say that the MA period has been quite exhausting. Like
many, I have never completed a piece of work as detailed as the MRP and I feared that I would
not be able to write convincingly about one topic for such a length. Being awarded the Social
Sciences and Humanities Research Council of Canada (SSHRC) scholarship just before I began
the MRP was admittedly a sizeable and well-timed boost to my confidence level. During the two
3-week residency periods at RRU, students were required to complete a period of written
reflection at the end of each week. Among other things, we were asked to comment on our
thinking and researching skills and how the previous week had impacted upon and developed
those skills. Like the residency periods, this MRP has been a period of self-discovery and I will
frame my reflection around those two skills: researching and thinking.
Researching Skills
It is ironic that I am considering my researching skills in light of my MRP topic being on
social media. When I first attended University in 1990, the Internet for use by the wider
population was essentially in its infancy. The degree of success of your research in the early
nineties depended mainly on the availability of hardcopy resources in libraries close to where you
lived or studied. Now, I have access to an almost un-ending supply of information 24-hrs a day
almost anywhere. As the subjects of my research pointed out regarding social media and
SOCIAL MEDIA AND EMERGENCY WARNINGS
77
emergency management warnings, information overload is a very real problem with today’s
Internet. This can be a significant problem in even the most basic research project as well I
suspect. I am generally a well-organized person, however the sheer volume of material available
online in addition to in the library means a researcher today must be vigilant to keep track of the
sources he or she uses and discerning in the material they choose to review in detail. A
researcher could spend the entire MRP period sifting through different literature and never really
making any headway. The MA program and the MRP process has forced me to become much
more comfortable with technology and the sheer volume of sources, search engines, survey
platforms and citation management systems available. I think a researcher today has to spend
more time just gathering their sources than ever before. Actually reading and critically analyzing
those sources is arguably the easy part, the hard part is determining the material upon which you
are going to base that analysis and ensuring it is usable and credible. Research is as much about
time management as it is about analysis and writing it seems.
What this last year has also developed in me is the ability to detect and appreciate subtlety
in the information I reviewed. During the literature review phase of the MRP, I would find
myself reading several articles that on a broad level, appeared to be saying the same thing. It was
only through a more careful analysis, and sometimes on a second or third read, that I would
realize that my own misinterpretation of a concept or definition fundamentally yet subtly refined
the argument being presented. On one level, this is because being immersed in a topic I tended
perhaps to approach each new article with my own evolving opinion or sub-conscious definition
for a concept that might or might not be shared by the article’s author. This would configure how
I in turn interpreted that new article. I often wonder if I had read many of the articles I reviewed
in a different order, would it have shaped my research and my opinions in a different way,
thereby impacting on the direction of my MRP? Realistically, I cannot re-invent the wheel and I
SOCIAL MEDIA AND EMERGENCY WARNINGS
78
cannot approach an article completely devoid of any opinion or academic or personal ‘baggage’.
But I did find that recording some of the assumptions I was making and some of the opinions I
was formulating concurrently with my literature review was helpful. It not only assisted me in
being more objective in my research, but it helped me to formulate some of the assumptions I
was making for my research that needed to be pointed out in the body of the MRP. What this also
did was help me challenge some of my existing viewpoints and more thoroughly examine the
opinions that were forming as I conducted my literature review. In my undergraduate studies, I
would have been more inclined to accept any piece of research as valid on the face of it and I
would assume that the opinion of the writer was more worthy than my own. Increasingly over
the last year, I have become more able to read and research critically and make a balanced
judgment on whether I think the author’s view is valid, and if not, why not.
Thinking Skills
During the MRP process, I do not think I used the term ‘systematic enquiry’ once. Yet
from our examination of systematic enquiry during the first period of residency at RRU, I realize
that I have been routinely employing it during this last year. This started with me writing a very
specific problem statement and formulating some equally specific sub-questions for my research.
Like many, I was keen to ‘jump right in’ and start researching and writing the MRP. This was
not only because of my interest in the topic, but also because I was conscious that this year would
pass by quickly. I had to very deliberately restrain myself from hasty and ill-conceived
‘progress’ and ensure that I had a logical foundation upon which to work. In hindsight, I am so
grateful that I was able to stave off my own impatience and adopt a more systematic approach to
my research. The MRP milestone process was also quite helpful in this regard. Having a number
of smaller deliverables and a time schedule to produce them was an excellent handrail. The MRP
schedule called for approximately the same amount of time being spent on formulating a research
SOCIAL MEDIA AND EMERGENCY WARNINGS
79
question as for conducting the literature review. At first glance, this seemed to me to be quite a
slow start and a misuse of time. However, after starting a few literature searches during that
research question formulation phase, I quickly realized that unless I had ‘nailed down’ that
research question and asked myself some second and third order questions about it early on, I
would have real issues identifying the right kind of literature suitable for my MRP. This
reinforces the point I made earlier on the overwhelming wealth of information available to the
researcher. Devoting sufficient time to your research question is time well spent! This does not
mean that you cannot reconsider your research question during your literature review. The MRP
is after all an iterative process. However, I have discovered that a well thought out research
question will hopefully prevent a significant directional change to a project mid-stream if
possible.
I mentioned above the need for a second and third order analysis during the MRP process.
The residency periods emphasized to me the sheer complexity of the multi-disciplinary field of
emergency management and the plethora of stakeholders potentially affected by even the most
‘basic’ emergency situation. The MA program has helped me to wrestle with some of the
underlying concepts to disaster in an effort to not just identify what is happening, but why it is
happening and what it means to a wider audience. A town does not flood ‘just’ because it does
not have flood defenses. Deeper questions about why certain houses and infrastructure were built
in areas we know are prone to flooding and what social factors make certain groups more
vulnerable, help an emergency manager look beyond the obvious argument and mitigation
strategy which might simply entail building a bigger flood wall. During the MRP process I found
myself asking these same ‘so what’ questions. In other words: conducting a subsequent level of
analysis that I might not ordinarily do. I have felt myself becoming a little more critical in my
thinking and looking for a tighter linkage between a question and an answer, both in the sources
SOCIAL MEDIA AND EMERGENCY WARNINGS
80
that I am reading and in my own work.
There is a real temptation after the data for the MRP was collected to just produce a series of
bright colored graphs and charts. Although presentation is a key part of the MRP process,
conducting that next level of enquiry and considering what that information may mean is what I
have gained most from this last year. This reinforces my earlier point regarding the importance
of devoting time to formulating a research question. Part of that initial formulation involved for
me asking questions about what information I was going to gain from posing each sub-question.
I then had to consider what value that information might hold. Would that information actually
have any utility in proving a thesis or answering that sub-question? I pride myself on being a
fairly logical and sequential thinker. Although the MRP process appears somewhat modular and
is established with a helpful system of milestones, I have found that each part does overlap with
other parts of the process. Although it would be premature to start formulating a written survey
before you have decided upon a research question, I would be lying if I said that I was not
considering how I might gather data during the early stages. Likewise, when formulating sub-
questions, I was considering what kind of data I might gather and how that data might be used to
answer that sub-question. So although I am historically more comfortable with a sequential,
ordered, ‘cause and effect’ thought process, the MRP process caused me to think along a number
of pathways simultaneously. Although the process was systematic and I had an ultimate critical
path, represented well by the MRP process and milestones, I had to ask myself some deeper
questions earlier on that did cause me to consider later milestones in the process. This did feel a
little over-whelming at times as I thought my thinking was a little disorganized and I was
‘walking in circles’. But I began to realize early on that as long as I kept to the overall critical
path and MRP structure, thinking ahead and asking some deeper questions about subsequent parts
of the MRP paid dividends in the long-term.
SOCIAL MEDIA AND EMERGENCY WARNINGS
81
Conclusion
I thoroughly enjoyed the MRP process. On one hand, I am elated that it is almost over! On
the other hand, I think there will be a bit of a void in my life being away from academia. Not
being enrolled in some kind of educational courses does not mean that I am going to stop
thinking critically. Quite the opposite in fact: the skills I have honed in the last year have made
me a more considered, logical and empathetic person. I now more readily question the impacts
of a decision or an issue from a wider point of view beyond my own. I would like to think that I
am a lot more inquisitive and critical in the ways in which I think. I also know that my
confidence in my own opinion has increased. I will definitely not always have the ‘right’ answer
(what ever that means) but I am more self-assured that my response will have a greater depth of
understanding of the problem than ever before.
SOCIAL MEDIA AND EMERGENCY WARNINGS
82
References
American Psychological Association. (2010). Publication manual of the American Psychological Association (6th ed.). Washington, DC: American Psychological Association.
Argenti, P. A. (2006). How technology has influenced the field of corporate communication. Journal of Business and Technical Communication, 20(3), 357-370. Retrieved from http://search.proquest.com.ezproxy.royalroads.ca/docview/ 196455658?accountid=8056
Armon, D. (2010). Ready or not: Is your crisis communications plan anti-social? Public Relations Tactics, 17(4), 10.
Auf der Heide, E. (1989). Disaster Response: communication with the public. Principles of preparation and coordination (pp. 122-132) St Louis: CV Mosby.
Botterell, A. (2008). Warnings that work. The American City & County, 123(7), 24. Retrieved from http://search.proquest.com.ezproxy.royalroads.ca/docview /195913904? accountid=8056
Burkhart, F. (1990). Media, emergency warnings and citizen response. (PhD, Arizona State University).
Canadian Red Cross. (2012). Social media during emergencies. Ottawa, Canada: Canadian Red Cross.
Castillo, C., Mendoza, M., & Poblete, B. (2011). Information credibility on twitter. Proceedings from the 20th International Conference on World Wide Web (WWW '11), New York. 675-684. doi:10.1145/1963405.1963500
Clarke, R. (1999). Internet privacy concerns confirm the case for intervention. Communications of the ACM. 42(2). 60-67.
Coombs, W. T., & Holladay, S. J. (2011). An exploration of the effects of victim visuals on perceptions and reactions to crisis events. Public Relations Review, 37(2), 115-120. doi:http://dx.doi.org.ezproxy.royalroads.ca/10.1016/j.pubrev.2011.01.006
Crowe, A. (2012). Disasters 2.0: The application of social media systems for modern emergency management. Boca Raton: Taylor & Francis. Retrieved from http://voyager.royalroads.ca/ vwebv/holdingsInfo?bibId=327769
Cutter, L., & Smith, M. M. (2009). Fleeing from the hurricane's wrath: Evacuation and the two americas. Environment, 51(2), 26-36. Retrieved from http://search. proquest.com.ezproxy .royalroads.ca/docview/224013998?accountid=8056
SOCIAL MEDIA AND EMERGENCY WARNINGS
83
Dwyer, C., Hiltz, S. & Passerini, K. (2007). Trust and privacy concerns within social networking sites: A comparison of facebook and myspace. Proceedings of the 13th Americas Conference on Information Systems, Keystone, Colorado. 339-350.
Fearn-Banks, K. (2011). Crisis communications: A casebook approach (4th ed.). New York: Routledge. Retrieved from http://voyager.royalroads.ca/vwebv/holdings Info?bibId=259172
Federation of Red Cross and Red Crescent Societies Staff. (2004). World disasters report 2004: Focus on community resilience. Bloomfield, USA: Kumarian Press Inc.
Fink, A. (2009). How to conduct surveys: A step-by-step guide (4th ed.). Thousand Oaks, CA: SAGE Publications, Inc.
Ghosh, J. K., Bhattacharya, D., Samadhiya, N. K., & Boccardo, P. (2012). A generalized geo-hazard warning system. Natural Hazards, 62(2), 1273-1289. doi:10.1007/s11069-012-0296-0
Goolsby, R. (2010). Social media as crisis platform: The future of community Maps/Crisis maps ACM Trans. Intell. Syst. Technol., 1(7), 7. doi:10.1145/ 1858948.1858955 http://doi. acm. org/10.1145/1858948.1858955
Handbook of disaster research (2007). Rodríguez H., Quarantelli E. and Dynes R. (Eds.). New York: Springer. doi:10.1007/978-0-387-32353-4
Hui, C., Tyshchuk, Y., Wallace, W., Magdon-Ismail, M., & Goldberg, M. (2012). Information cascades in social media in response to a crisis: A preliminary model and a case study. Proceedings from the 21st International Conference Companion on World Wide Web (WWW '12 Companion), New York. 653-656. doi:10.1145/ 2187980.2188173
Imran, M., Elbassuoni, S. M., Castillo, C., Diaz, F., & Meier, P. (2013). Extracting information nuggets from disaster-related messages in social media. ISCRAM, Baden-Baden, Germany,
Jaeger, P. T., Shneiderman, B., Fleischmann, K. R., Preece, J., Qu, Y., & Fei Wu, P. (2007). Community response grids: E-government, social networks, and effective emergency management. Telecommunications Policy, 31(10–11), 592-604. doi:http://dx.doi. org.ezproxy.royalroads.ca/10.1016/j.telpol.2007.07.008
Jaegera, P., Shneiderman, B., Fleischmann, K., Preece, J., Qu, Y., & Wu, P. (2007). Community response grids: E-government, social networks, and effective emergency management. Telecommunications Policy, 31, 592-604.
Johnson, T. & Kaye, B. (1998). Cruising is believing: Comparing internet and traditional sources on media credibility measures. Journalism and Mass Communication Quarterly 75(2), 325-340. Retrieved from http://www.aejmc.org/home/wp-content/uploads/2012/09/Journalism-Mass-Communication-Quarterly-1998-JohnsonKay-325-340.pdf
SOCIAL MEDIA AND EMERGENCY WARNINGS
84
Kasperson, R., & Kasperson, J. (1996). The social amplification and attenuation of risk. Annals of the American Academy of Political and Social Science, 545, 95-105.
Kavanaugh, A., Fox, E., Sheetz, S., Yang, S., Li, L., Whalen, T., . . . Xie, L. (2012). Social media use by government: From the routine to the critical. Proceedings from the 12th Annual International Digital Government Research Conference: Digital Government Innovation in Challenging Times, New York. 121-130. doi:10.1145/2037556.2037574
Lindsay, B. R. (2011). Social media and disasters: Current uses, future options, and policy considerations. Washington DC: Congressional Research Service.
Martin, N., & Rice, J. (2012). Emergency communications and warning systems. Disaster Prevention and Management, 21(5), 529-540. doi:http://dx.doi.org.ezproxy. royalroads. ca/10.1108/09653561211278671
Mauriello, T. (2013). Entering a new age of disaster response. Pittsburgh Post-Gazette (Pittsburgh, Pa.1978). A.1.
Mersham, G. (2010). Social media and public information management: The september 2009 tsunami threat to new zealand. Media International Australia Incorporating Culture & Policy, (137), 130.
Meyer, J. (2001). Effects of warning validity and proximity on responses to warnings. Human Factors: The Journal of the Human Factors and Ergonomics Society, 43, 563-572. doi:10.1518/001872001775870395
Mileti, D. Public hazards communication and education: The state of the art. Public Education for Earthquake Hazards, 2007(2), 1-14.
Mileti, D., & Peek, L. (2000). The social psychology of public response to warnings of a nuclear power plant accident. Journal of Hazardous Materials, 75(2–3), 181-194. doi:http://dx.doi.org.ezproxy.royalroads.ca/10.1016/S0304-3894(00)00179-5
Ministers Responsible for Emergency Management. (2011). An emergency management framework for canada. Ottawa, Canada: Emergency Management Policy directorate.
O’Brien, M., Bass, E., Bliss, J., Cooke, N., Kalsher, M., & Mayhorn, C. (2012). Improving public response to disaster warnings: A discussion panel. Paper presented at the Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56: 2036-2039.
O'Leary, Z. (2010). The essential guide to doing your research project. Thousand Oaks, CA: SAGE Publications, Inc.
O’Neil, D. (2001). Analysis of internet users’ level of online privacy concerns. Social Science Computer Review. 19(17). 17-31. doi: 10.1177/089443930101900103
SOCIAL MEDIA AND EMERGENCY WARNINGS
85
Page, S., Freberg, K., & Saling, K. (2013). Emerging media crisis value model: A comparison of relevant, timely message strategies for emergency events. Journal of Strategic Security, 6(2), 20-31.
Palen, L. (2008). Online social media in crisis events. Educause Quarterly, 31(3), 76-78.
Pechta, L. E. (2010). Understanding the dynamics of emergency communication: Propositions for a four-channel model. Journal of Homeland Security and Emergency Management, 7(1), 1-18.
Pew Internet. (2009). Generations online in 2009. Washington, DC. Retrieved from http://www.pewinternet.org/Reports/2009/Generations-Online-in-2009.aspx
Quan-Haase, A., Young, A. (2009). Information revelation and internet privacy concerns on social network sites: A case study of facebook. Paper presented at C&T ’09. Retrieved from http://sozio-informatik.net/fileadmin/IISI/upload/2009/p265.pdf
Quarantelli, E. (2000). Disaster planning, emergency management and civil protection: The historical development of organized efforts to plan for and to respond to disasters. (Unpublished Preliminary Paper #301). University of Delaware, Disaster Research Center.
Reynolds, B. (2010). Building trust through social media. Center for disease control experience with H1N1 swine flu response. Marketing Health Services, 30(2), 18.
Richards, L. (2009). Handling qualitative data: A practical guide (2nd ed.). Thousand Oaks, CA: SAGE Publications, Inc.
Scanlon, J. (2007). Chapter 24: Unwelcome irritant or useful ally? The mass media in emergencies. In R. Havidan, E. Quarantelli & R. Dynes (Eds.), Handbook of disaster research (pp. 413-429). New York: Springer.
Schmierbach, M. (2012). A little bird told me, so I didn't believe it: Twitter, credibility, and issue perceptions. Communication Quarterly, 60(3), 317.
Schultz, F., Utz, S., & Göritz, A. (2011). Is the medium the message? Perceptions of and reactions to crisis communication via twitter, blogs and traditional media. Public Relations Review, 37(1), 20-27. doi:http://dx.doi.org.ezproxy.royalroads.ca/ 10.1016/j.pubrev.2010. 12.001
Schonberger, K. (2005). How false information creates ‘true events’. Persistent and recombined forms of activist communication through internet fakes and hoaxes. Hamburg, Germany: University of Hamburg. Retrieved from http://www.eastbound.eu/site_media/pdf/ 060113schoenberg.pdf
Seeger, M. (2006). Best practices in crisis communication: An expert panel process. Journal of Applied Communication Research: JACR, 34(3), 232-244.
SOCIAL MEDIA AND EMERGENCY WARNINGS
86
Sene, K. Flood warning, forecasting and emergency response. (2008). UK: Springer Berlin Heidelberg. doi:10.1007/978-3-540-77853-0
Shankar, K. (2008). Wind, water, and wi-fi: New trends in community informatics and disaster management. The Information Society, 24(2), 116-120.
Simpliflying. (2013). Crises management 2.0. Asiana airlines flight OZ214 in SFO. Message posted to http://www.slideshare.net/shanxz/asiana-flight-214-crash-in-sfo-crises-management-case-study-and-analysis?utm_source=slideshow&utm_ medium=ssemail &utm_campaign=weekly_digest
Sorensen, J. H. (2000). Hazard warning systems: Review of 20 years of progress. Natural Hazards Review, 1(2), 119. Retrieved from https://ezproxy.royalroads.ca/login? url=http:// search.ebscohost.com/login.aspx?direct=true&db=eih&AN=6786276
Sorenson, J. H., & Sorensen, B. V. (2007). Chapter 11: Community processes: Warning and evacuation. In H. Rodriguez , E. Quarantelli & R. Dynes (Eds.), Handbook of disaster research (pp. 183-199). New York: Springer.
Spicer, T. (2013). Being social: Integrating social media into public information support to emergency response. (Unpublished Masters). Naval Postgraduate School, Monterey, CA.
Statistics Canada. (2010). Canadian internet use survey. Retrieved from http://www.statcan. gc.ca/tables-tableaux/sum-som/l01/cst01/comm36a-eng.htm
Statistics Canada. (2011). Focus on geography series, 2011 census. Census metropolitan area of Vancouver, British Columbia. Retrieved from http://www12.statcan.gc.ca/census-recensement/2011/as-sa/fogs-spg/Facts-cma-eng.cfm?LANG=Eng&GC=933
Sutton, J., Palen, L., & Shklovski, I. (2008). Backchannels on the front lines: Emergent uses of social media in the 2007 southern california wildfires. Proceedings of the 5th International ISCRAM Conference , Washington, DC.
Tierney, K. J., Lindell, M. K., & Perry, R. W. (2001). Facing the unexpected disaster preparedness and response in the united states. Washington, DC: Joseph Henry Press. Retrieved from http://voyager.royalroads.ca/vwebv/ holdings Info?bibId=184651
United Nations General Assembly. (2012). The promotion, protection and enjoyment of human rights on the Internet, A /HRC/ 20 /L. 1 3.
Veil, S. R. (2011). A work-in-process literature review: Incorporating social media in risk and crisis communication. Journal of Contingencies and Crisis Management, 19(2), 110-122.
Walaski, P. (2013). Social media. Professional Safety, 58(4), 40-49. Retrieved from http://search.proquest.com.ezproxy.royalroads.ca/docview/1331594111?accountid=8056
SOCIAL MEDIA AND EMERGENCY WARNINGS
87
White, C. M. (2012). Social media, crisis communication, and emergency management: Utilizing web 2.0 technologies. Boca Raton, Fla.: CRC Press. Retrieved from http://voyager. royalroads.ca/vwebv/holdingsInfo?bibId=300703
Widener, J., Horner, M., & Metcalf, S. (2013). Simulating the effects of social networks on a population’s hurricane evacuation participation Journal of Geographical Systems, 15(2), 193-209. doi:10.1007/s10109-012-0170-3
SOCIAL MEDIA AND EMERGENCY WARNINGS
88
Appendix A
SOCIAL MEDIA AND EMERGENCY WARNING INFORMATION
My name is James Embleton-Forrest and this research project is part of the requirement for a Masters Degree in Disaster and Emergency Management at Royal Roads University. If you wish to verify my credentials with Royal Roads University you may contact Associate Faculty Eileen Davenport at [email protected]. This document constitutes an agreement to participate in my research project, the objective of which is to examine how the credibility of social media can be enhanced for use in disseminating emergency management information. The research will consist of a short interview consisting of a mix of open-ended questions and is foreseen to last 10-15mins. The questions will refer to the ways in which you currently use social media in your day-to-day life. I will be submitting my final report to Royal Roads University in partial fulfillment for a Masters Degree in Disaster and Emergency Management. Information will be recorded on computer and where appropriate, summarized in the body of the final report. Participants do not have to include their name or any personal information in order to preserve anonymity. At no time will any specific comments be attributed to any individual unless specific agreement has been obtained beforehand. All documentation will be kept strictly confidential in a locked cabinet and destroyed once my report is completed and accepted by Royal Roads University. If a participant withdraws from the study mid-way, all records of their survey answers will be destroyed upon withdrawal and not included in the study. The final report will be disseminated to Royal Roads faculty electronically. A copy can also be sent to interested survey participants by arrangement. You are not compelled to participate in this research project. If you do choose to participate, you are free to withdraw at any time without prejudice. Similarly, if you choose not to participate in this research project, this information will also be maintained in confidence. Do you wish to continue?
• Yes • No
1. Please state your gender?
• Male • Female • Other
2. What is your age group?
SOCIAL MEDIA AND EMERGENCY WARNINGS
89
• 18-35 years of age • 36-50 years of age • 51-65 years of age • over 65 years of age
3. Do you have Internet access at your home?
• Yes • No
4. Do you have Internet access at your place of work?
• Yes • No
5. If your answer to the previous question was 'Yes', is that access electronically restricted by your employer? (i.e. are there some websites that you are unable to access at work).
• Yes • No • Not applicable
6. Do you use a mobile device to access the Internet? (i.e. smart phone or tablet)?
• Yes • No
7. Do you subscribe to any 'push' notifications via the Internet? (i.e. information like breaking news or weather updates that are automatically sent to your internet device without you having to deliberately access them yourself).
• Yes • No
8. Do you use any of the following social networking accounts? (Please add any social media applications that you use that are not listed).
• Facebook • Twitter • Email (known sender)
SOCIAL MEDIA AND EMERGENCY WARNINGS
90
• Email (unknown sender) • Linked-in • YouTube • Flickr • Other (Please specify below) • Other (Please specify below) • Other (Please specify below) • Other (Please specify below)
9. Of the forms of social media you use, how do you rate each in terms of credibility?
Very credible
Somewhat credible Neutral Lacking in
credibility Very
lacking in credibility
Facebook Twitter
Email (known sender)
Email (unknown sender)
Linked-in YouTube
Flickr Other (Please specify below) Other (Please specify below) Other (Please specify below) Other (Please specify below)
SOCIAL MEDIA AND EMERGENCY WARNINGS
91
10. If you received an emergency warning message through any of the social media forms listed above, how likely are you to immediately alter your personal behavior in response (i.e. evacuate, shelter in place etc)?
Very likely
Somewhat likely
Neither likely nor unlikely
Somewhat unlikely
Very unlikely
Facebook Twitter
Email (known sender)
Email (unknown sender)
Linked-in YouTube
Flickr Other (Please specify below) Other (Please specify below) Other (Please specify below) Other (Please specify below)
SOCIAL MEDIA AND EMERGENCY WARNINGS
92
11. How important to you is knowing the originating source of a message received on social media?
• Very important • Somewhat important • Neither important nor unimportant • Somewhat unimportant • Very unimportant
12. Does knowing the originating source of a message make you more likely to immediately alter your behavior as result?
• Yes • No
13. What contributes more to social media credibility in your opinion: knowing the originating source of a message or the social media application upon which you are viewing the message?
• Knowing the originating source of the message • The social media application upon which the message is viewed
14. Would you want to receive emergency warning information from the City of Vancouver and official agencies via your social media applications?
• Yes • No
15. Do you have any recommendations for how emergency warning information disseminated on social media can be made more credible?
SOCIAL MEDIA AND EMERGENCY WARNINGS
93
Appendix B
Consolidated Response Data for Survey
1. Please state your gender?
Male 41 Female 61 Other 0
2. What is your age group?
18-35 years of age 6 male, 12 female 36-50 years of age 29 males, 39 female 51-65 years of age 6 male, 9 female over 65 years of age 0 male, 1 female
3. Do you have Internet access at your home?
Yes 102 No 0
4. Do you have Internet access at your place of work?
Yes 89 No 7 Not Applicable 6
5. If your answer to the previous question was 'Yes', is that access electronically restricted by your employer? (i.e. are there some websites that you are unable to access at work).
Yes 34 No 55 Not Applicable 13
6. Do you use a mobile device to access the Internet? (i.e. smart phone or tablet)?
Yes 93 No 9
7. Do you subscribe to any 'push' notifications via the Internet? (i.e. information like breaking news or weather updates that are automatically sent to your internet device without you having to deliberately access them yourself).
Yes 52 No 50
SOCIAL MEDIA AND EMERGENCY WARNINGS
94
8. Do you use any of the following social networking applications?
Facebook 81 Twitter 38 Email (from a known contact) 97 Email (from a contact not known to you) 45 linkedin 71 YouTube 77 Flickr 11 Instagram 19 Tumblr 5 Google+ 35 Vimeo 1 Vine 2 Wikispaces 3 Snap Chat 2 Foursquare 1 HeyTell 1 Skype 1 Individual Blogs 1 Flip Board 1 Google Circles 1
9. Of the forms of social media you use, how do you rate each in terms of credibility? Please also rate any applications you added to the previous question under Other 1, 2, 3 or 4.
Very credible
Somewhat credible Neutral Lacking in
credibility
Very lacking in credibility
Facebook 8 40 19 12 2 Twitter 7 19 14 8 1
Email (from a known sender) 58 34 5 Email (from a contact not
known to you) 3 14 19 24 1
linkedin 12 43 10 1 YouTube 4 22 34 15
Flickr 3 6 11 5 Instagram 1 2 14 7 Tumblr 2 13 4
Google+ 8 11 15 Vimeo 1 1 Vine 1 2
Wikispaces 2 1 Snap Chat 1 Foursquare 1
SOCIAL MEDIA AND EMERGENCY WARNINGS
95
Very credible
Somewhat credible Neutral Lacking in
credibility
Very lacking in credibility
HeyTell 1 Skype 1
Individual Blogs 1 Flip Board 1
Google Circles 1
10. If you received an emergency warning message on any of the social media applications listed above (i.e. an evacuation advisory or an extreme weather alert), how likely are you to alter your personal behavior in response? Please also rate any applications you added to the previous
11. How important is knowing the originating source of a message received on social media to you?
Very likely Somewhat likely
Neither likely nor unlikely
Somewhat unlikely
Very unlikely
Facebook 9 31 12 5 2 Twitter 10 20 7 2 3
Email (from a known sender) 46 18 2 Email (from a contact not
known to you) 5 12 15 8 8
linkedin 5 24 13 5 YouTube 1 13 25 11 6
Flickr 1 7 7 1 Instagram 1 9 5 1 Tumblr 7 6 1
Google+ 3 9 7 4 Vimeo 1 Vine 1 1
Wiki Spaces 2 Snap Chat 1 Foursquare 1
HeyTell 1 Skype 1
Individual Blogs 1 Flip Board 1
Google Circles 1
SOCIAL MEDIA AND EMERGENCY WARNINGS
96
Very important 91 Somewhat important 11 Neither important nor unimportant 0 Somewhat unimportant 0 Very unimportant 0
12. Does knowing the originating source of a message make you more likely to immediately alter your behavior as a result?
Yes 99 No 3
13. What contributes more to social media credibility in your opinion: knowing the originating source of a message or the social media application upon which you are viewing the message?
Knowing the originating source of the message 98 The social media application upon which the message is viewed 4
14. Would you want to receive emergency warning information from government (local, provincial or federal) or other official agencies, via your social media applications?
Yes 94 No 8
SOCIAL MEDIA AND EMERGENCY WARNINGS
97
Appendix C
Simulated Emergency Management Messages
Simulated Message Number One: UNSOLICITED EMAIL
Retrieved from http://www.google.org/publicalerts/alert?aid=9a6d2b485c772bd&hl=en&gl=CA&source=web
SOCIAL MEDIA AND EMERGENCY WARNINGS
98
Simulated Message Number Two: TWITTER
Retrieved from https://twitter.com/CityofVancouver
SOCIAL MEDIA AND EMERGENCY WARNINGS
99
Simulated Message Number Three: FACEBOOK
Retrieved from https://www.facebook.com/CityofVancouver?rf=127584693944934
SOCIAL MEDIA AND EMERGENCY WARNINGS
100
Simulated Message Four: TEXT MESSAGE
Unknown
SOCIAL MEDIA AND EMERGENCY WARNINGS
101
Appendix D
Simulated Message Assessment Sheet
This assessment represents a continuation of the previous survey and the same guidelines apply. You may withdraw at any time without prejudice and your anonymity will be maintained unless you specifically authorize otherwise. This assessment refers to four simulated warning messages. The messages are fictional and no real emergency exists. Question One: How credible did you find each of the simulated warnings messages? Very
Question Four: If you stated that you are likely or somewhat likely to pass this message onto family or friends, would you pass it on via social media or via word-‐of-‐mouth? I would forward the message exclusively via social media I would forward the message predominantly via social media I would forward the message via social media and word-‐of-‐mouth equally 2 I would forward the message predominantly via word-‐of-‐mouth 3 I would forward the message exclusively via word-‐of-‐mouth 1 Not applicable
SOCIAL MEDIA AND EMERGENCY WARNINGS
103
Appendix E
The comments provided by respondents have been consolidated and have not been repeated
verbatim for the benefit of brevity.
Descriptive Coding Topical Coding Analytical Coding Age group 18-35: Total of 18 respondents
Male: Total of 6 respondents
Warnings must be automated Facility to authenticate sender Facility to authentication the source Credibility has little to do with the platform. More to do with the people using it and their motivations
Simplicity Ease of use Legitimacy
Female: Total of 12 respondents
Message must come from a credible source such as emergency services
Legitimacy
Age group 36-50: Total of 68 respondents
Male: Total of 29 respondents
Warning messages should be automated Message are most credible from a government source Emergency managers should create an Internet application that people can opt to use creating ‘buy-in’ Systems that people have already taken the time to engage with will be more widely accepted Importance of ‘word of mouth’. Relying on people to promulgate through their own networks Make notifications brief and to the point but also include a link to more detailed information for those who want to research
Simplicity Ease of use Legitimacy Simplicity Engagement Utility Simplicity
SOCIAL MEDIA AND EMERGENCY WARNINGS
104
Age group 36-50 continued:
Male continued:
further Make the notifications free, also including a free function to forward messages on to other friends and family easily Offer financial incentives that are already open for use Remove concerns about scams or being subjected to unrelated mass-emailing An agency must have the ability to override our common social media devices/portals e.g. when an Amber alert appears on your TV I would rely more on conventional media i.e. TV / radio (or perhaps word of mouth from other people who are using social media)
Lack of expense Simplicity Legitimacy Trust Trust
Female: Total of 39 respondents
Show a recorded video of a known political figure and proof of author Public information service campaign via traditional media promoting the service It would have to be from a warranted source that links to a warranted site A post on Facebook from the government that is then linked to their website that says the same thing Verification via news and radio Verification or proof that only that source could provide Establish an 'emergency
identity' before Use only in a legitimate emergency Using trusted newspaper radio or TV news companies with push notifications Use a vast spectrum of media sources Do not rely on any single application or source I would rely on the CBC Establish email link/rapport in advance to avoid skepticism when a warning is received Establish confidence in the system before hand Have a specific social media website designed for this purpose
Trust Saturation Saturation Trust Engagement Engagement Legitimacy Ease of use
Age group 51-65: Total of 15 respondents
Male: Total of 6 respondents
Having a means of verifying the source and content - e.g. a follow-up call or email Simply state who the original message is from and offer an opportunity to verify Information received by subscription tends to be credible; information that gets through anti-spamming controls will be somewhat credible and that credibility will be increased if it carries some sort of verifiable or distinguishing identifier Use multiple medias
Legitimacy Simplicity Legitimacy Saturation
SOCIAL MEDIA AND EMERGENCY WARNINGS
106
Female: Total of 9 respondents
No comments made No comments made
Age group over 65: Total of 1 respondent
Male: Total of 0 respondents
No comments made No comments made
Female: Total of 1 respondents
Pre-authenticated sound and visual alarm to indicate if it is real or a practice or bogus Authentication of source with continuing updates