Introduction during the period of peak storm

“All neighborhoods ... and possibly entire coastal communities ... will be

inundated during the period of peak storm tide,” a National Weather Service (NWS)

advisory said in wake of Hurricane Ike’s predicted landfall on Galveston Island in

September 2008. “Persons not heeding evacuation orders in single-family one- or

twostory homes will face certain death.” The language of “certain death” created an

unprecedented response from citizens all across the Gulf Coast, specifically residents in

Texas, in the path of Hurricane Ike. Over one million people evacuated to places deemed

structurally safe from the hurricane.

The NWS wasn’t the only organization/agency communicating messages of this

magnitude. The United States Department of Homeland Security (DHS) Secretary

Michael Chertoff, “urged people not to succumb to hurricane fatigue,” in referring to

concerns that authorities were overestimating Hurricane Ike's potential impact. He added,

"unless you're fatigued with living, I suggest you want to take seriously a storm of this

size and scale.” In addition to the NWS and DHS having similar messages about the

possible destruction Hurricane Ike could bring, Houston’s Mayor Bill White responded to

reports that people in mandatory evacuation areas planned on staying in their homes and

urged them to reconsider. “If you think you want to ride something out, and people are

talking about a 20-foot wall of water coming at you, then you better think again.”

Message continuity at all levels of government is critical when dealing with hurricane

mitigation and preparedness. A series of diverse evacuation messages during Hurricane

Katrina in 2005 ultimately led to deaths and thousands being stranded without food, water

or humane conditions for days. In addition, the mismanagement of information about

possible levee failures throughout the city during Hurricane Katrina poised agencies

involved in Hurricane Ike to explain all possible outcomes related to the storm’s impact

and to not recreate the scene that unfolded in New Orleans on national television.

Ineffective emergency communication during Hurricane Katrina led to one of the

biggest failures of our government. However, what motivates citizens to respond to

certain messages and not others? What type of sources and messages provoke different

attitudes, beliefs, and behaviors? Will complacency outweigh hurricane preparedness and

mitigation? Will the aftermath of Hurricane Ike support the threatening messages used by

the NWS and others? Should “certain death” language be used again in emergency

communication? It’s critical to understand the attributes of hurricane preparedness and

mitigation messages to diminish future risks. The NWS plays a vital role in emergency

communication. It is often the main source for information regarding future hurricane

projections, track, strength, storm surge, and other hurricane related factors.

The NWS, a division of the National Oceanic and Atmospheric Administration,

released a hurricane preparedness guide that stated that one of the major problems with

hurricanes making landfall in the United States is resident’s perception of risk associated

with these storms. It indicates several reasons for lack of preparedness and mitigation

procedures. Besides infrastructure problems related to urban sprawl, a high percentage of

the population living along hurricane prone areas have only experienced “weaker” storms

and not experienced the “major” storms that cause catastrophic damage. This has led

many individuals to downplay the need to evacuate and remain complacent when experts

urge residents to vacate at risk areas.

This study seeks to further understanding of the effects of emergency message

strategies and message sources on individuals. The importance of understanding message

effects in emergency communication is clear. The findings from this study may provide

information about how communicators can best structure their messages to ensure the

safety of the public.

This study explores message strategy effects in an emergency communication

context using Hazleton and Long’s 1993 public relations process model and Grunig’s

1997 situational theory of publics. The public relations message strategies examined in

this study were derived from Hazleton and Long’s public relations process model.

Hazleton developed a taxonomy of seven public relations strategies that organizations use

when communicating with publics. The seven strategies are: facilitative, informative,

persuasive, promise and reward, threat and punishment, bargaining, and cooperative

problem solving.

Grunig’s situational theory of publics is used to understand publics and measure

their opinions about issues. Grunig and Hunt (1984) stated that communication behaviors

of publics can be best understood by measuring how members of publics perceive

situations in which they are affected. The theory consists of three independent variables

and two dependent variables. The three independent variables are problem recognition,

constraint recognition, and level of involvement. These variables describe “perceptions

that people have of specific situations, especially situations that are problematic or that

produce conflicts or issues” (Grunig, 1997, p. 10). The dependent variables are

information seeking and information processing.

The purpose of this study is to understand what strategy type and message source

is most effective in emergency management communications. Below are the hypotheses

and propositions this study tests.

H1: In emergency communication, message source will influence receiver variables.

P1.1: A national message source (NWS) will produce higher problem recognition

than a local message source (HCG).

P1.2: A national message source (NWS) will produce higher constraint

recognition than a local message source (HCG).

P1.3: A national message source (NWS) will produce higher level of involvement

then a local message source (HCG).

P1.4: A national message source (NWS) will produce higher expertise then a local

message source (HCG).

P1.5: A national message source (NWS) will produce higher trust than a local

message source (HCG).

P1.6: A national message source (NWS) will produce more positive attitudes than

a local message source (HCG).

P1.7: A national message source (NWS) will produce higher information-seeking

than a local message source (HCG).

H2: In emergency communication, message strategy will influence receiver variables.

P2.1: Information seeking will be the highest when the threat and punishment

strategy is used.

P2.2: Information seeking will be the lowest when the informative strategy is

used.

H3: Level of involvement will produce the strongest effect on information seeking

behavior.

H4: Constraint recognition will produce the weakest effect on information seeking

behavior.

The following chapter provides a review of literature important to this study.

Chapter 3 explains the methods and procedures used to gather data for this study. Chapter

4 reviews the results of this study, and Chapter 5 provides discussion of the results and

draws conclusions about the findings of this study.

Chapter Two

Literature Review

The purpose of this study is to understand what strategy type and message source

is most effective in emergency management communications. This chapter reviews the

existing literature relevant to this investigation.

Emergency Management

Emergency management communications is “the dissemination of timely and

accurate information to the general public, elected and community officials and the

media. This plays a major role in the effective management of disaster response and

recovery activities” (Haddow & Bullock, 2003, p. 63). The four phases of emergency

management are mitigation, preparedness, response, and recovery. Figure 1 indicates the

flow of different phases in emergency communication when a disaster occurs.

Figure 1: Four Phases of Emergency Communication

Communication is critical in the mitigation and preparedness phases of

emergency management. The mitigation phase focuses on preemptive measures that can

minimize the damage of a disaster. Mitigation activities are not done overnight. These are

planned activities in advance of a known risk. An example is identifying what schools are

deemed hurricane shelters and how many residents each school can accommodate.

“Federal, state, and local government agencies play a prominent role during this phase

and, in general, are responsible for setting the agenda, engaging the appropriate players in

planning and establishing and enforcing rules and regulations to achieve agreed-on plans”

(Guion, Scammon, & Borders, 2007, p. 21). Mitigation promotes the implementation of

strategies, technologies, and actions that will reduce the loss of lives and property damage

in future disasters (Haddow & Bullock, 2003).

Preparedness focuses on reducing the negative outcomes of disasters. One of the

main characteristics of this phase is “disseminating messages aimed at encouraging

people to make choices about protective behaviors and monitoring compliance with

community plans” (Guion, et al., 2007, p. 21). “During this phase, government agencies

are responsible for ensuring the safety of people in the disaster area and the environment”

(p. 21). An example of the preparedness phase is when the National Weather Service

sends out information regarding tropical storm and hurricane warnings. Below lists the

different characteristics of emergency management at each phase according to Guion,

Scammon and Borders, 2007, p 21.

Table 1: Major Participants in Emergency Management in Disaster Phases

Even though communication messages are disseminated at all phases of

emergency management, this study focuses on the type of messages that get people to act

prior to a potential disaster. The response and recovery phases of emergency management

are exercised when the disaster is happening or has taken place. It includes search and

rescue, support labor and the coordination of aid programs at the response phase, and

shelter coordination and job/training resources at the recovery phase.

Hurricane Classification

The Saffir-Simpson Hurricane Scale is a rating system that measures a hurricane’s

intensity. The scale classifies hurricanes as category 1, 2, 3, 4, or 5 storms. The type of

potential damage depends on the classification. The scale of potential damage ranges

from minimal to catastrophic. Each number estimates the scale of property damage as

related to the strength of the hurricane. Hurricanes classified as categories 3, 4, or 5 are

considered major hurricanes because of the possibility of property damage and loss of

life. Also, the scale gives an accurate representation of the amount and type of property

damage and flooding to expect. The wind speed is the primary factor of the scale. Below,

Figure 2 gives a detailed description of the level of damage a storm can bring by category

classification.

Figure 2: Saffir-Simpson Hurricane Scale

Category

Damage Level

Description

Example

Minimal

Damage primarily to shrubbery, trees, foliage, and

unanchored homes. No real damage to other structures. Some

damage to poorly constructed signs. Low-lying coastal roads

inundated, minor pier damage, some small craft in exposed

anchorage torn from moorings.

Hurricane Earl (1998)

Moderate

Considerable damage to shrubbery and tree foliage; some

trees blown down. Major damage to exposed mobile homes.

Extensive damage to poorly constructed signs. Some damage

of roofing materials of buildings; some window and door

damage. No major damage to buildings. Coast roads and low-

lying escape routes inland cut by rising water 2 to 4 hours

before arrival of hurricane center. Considerable damage to

piers. Marinas flooded. Small craft in unprotected

anchorages torn from moorings. Evacuation of some

shoreline residences and low-lying areas required.

Hurricane Georges

(1998)

Extensive

Foliage torn from trees; large trees blown down. Practically

all poorly constructed signs blown down. Some damage of

roofing materials of buildings; some window and door

damage. Some structural damage to small buildings. Mobile

homes destroyed. Serious flooding at coast and many smaller

structures near coast destroyed; larger structures near coast

damaged by battering waves and floating debris. Low-lying

escape routes inland cut by rising water 3 to 5 hours before

hurricane center arrives. Flat terrain 5 feet or less above sea

level flooded inland 8 miles or more. Evacuation of low-lying

residences within several blocks of shoreline possibly

required.

Hurricane Fran (1996)

Extreme

Shrubs and trees blown down; all signs down. Extensive

damage to roofing materials, windows and doors. Complete

failures of roofs on many small residences. Complete

destruction of mobile homes. Flat terrain 10 feet or less above

sea level flooded inland as far as 6 miles. Major damage to

lower floors of structures near shore due to flooding and

battering waves and floating debris. Low-lying escape routes

inland cut by rising water 3 to 5 hours before hurricane

center arrives. Major erosion of beaches. Massive evacuation

of all residences within 500 yards of shore possibly required

and of single story residences within 2 miles of shore.

Hurricane Andrew

(1992)

Catastrophic

Shrubs and trees blown down; considerable damage to roofs

of buildings; all signs down. Very severe and extensive

damage to windows and doors. Complete failure of roofs on

many residences and industrial buildings. Extensive

shattering of glass in windows and doors. Some complete

building failures. Small buildings overturned or blown away.

Complete destruction of mobile homes. Major damage to

lower floors of all structures less than 15 feet above sea level

within 500 yards of shore. Low-lying escape routes inland cut

by rising water 3 to 5 hours before hurricane center arrives.

Massive evacuation of residential areas on low ground within

5 to 10 miles of shore possibly required.

Hurricane Camille

(1969)

Hurricane Characteristics

When a hurricane makes landfall, the magnitude of destruction is determined by a

variety of factors. These factors include storm surge, storm tide, wind, tornadoes and

inland/freshwater flooding. The level of impact is determined by the strength of the

storm.

Defined by the National Oceanic and Atmospheric Administration (NOAA), storm

surge “is a large dome of water often 50 to 100 miles wide that sweeps across the

coastline near where a hurricane makes landfall” (U.S. Department of Commerce, 2001,

p. 5). The top of the dome consists of battering waves. The impact varies depending on

the strength of the storm and the water level surrounding the coastline in which the

hurricane will make landfall. The more shallow the water is, combined with strength,

determines the height of the surge.

Another factor that determines the impact of a hurricane is storm tide. This is a

combination between storm surge and astronomical tide. The time that a storm makes

landfall determines the effect of storm tide. If a hurricane makes landfall during high tide

the results can be more devastating in terms of property damage and loss of life. Figure 3

documents the difference tides can make on storm surge impact.

Figure 3: Storm Tide Diagram

The main determinant of the Saffir-Simpson Scale as previously mentioned is

wind. A tropical storm becomes a hurricane when winds are measured at a sustained 74

mph or greater. Winds can make ordinary signs, outdoor furniture, lawn decors, etc. into

flying missiles. In addition, winds can be sustained well inland from the initial of landfall

of the storm.

Figure 4: Parts of a Hurricane

The final impact determinant is inland/freshwater flooding. Depending on the

speed of the storm, hurricanes can produce an excessive amount of rainfall in a short

period of time. Flooding is often a major concern for inland residents. Large amounts of

rainfall over a short period of time can also trigger mudslides in more mountainous

regions along the East coast of the United States. According to the U.S. Department of

Commerce (2001), freshwater flooding has accounted for 59% of U.S hurricane deaths

between 1970 and 1999 (p. 7). One of the main reasons is flash flooding. Flash floods

occur when there is a rapid rise in water levels due to substantial rainfall in a short

amount of time.

These five elements determine the potential impact of a hurricane. Mitigation and

preparedness communication informs residents about these attributes and how best to

protect themselves and their property.

Hurricane Watches/Warnings

Once a storm is identified, the National Weather Service releases a series of

advisories regarding the possibility of a tropical storm or hurricane making landfall along

the coast of the United States. According to the NWS, an advisory is official information

issued by the National Hurricane Center describing all watches and warnings in effect and

provides details concerning location, intensity, movement, and precautions that should be

taken. These advisories describe the storms potential landfall location by issuing four

different classifications: tropical storm watch, tropical storm warning, hurricane watch,

and hurricane warning. The NWS Web site defines these terms which are provided in

Table 2 below.

Table 2: Tropical Storm/Hurricane Watches & Warnings4

Type

Description

Tropical Storm Watch

Tropical storm conditions with sustained

winds from 39 to 73 mph are possible in

the watch area with the next 36 hours.

Tropical Storm Warning

Tropical storm conditions are expected in

the warning area within the next 24 hours.

Hurricane Watch

Hurricane conditions (sustained winds

greater than 73 mph) are possible in the

watch area within 36 hours.

Hurricane Warning

Hurricane conditions are expected in the

warning area in 24 hours or less.

After these watches and warnings are in place, the NWS will make predictions

based on various models about the possible landfall location of the storm. These models

are often combined on a single chart to produce a spaghetti model. This model allows you

to see the predicted direction of the storm by a variety of computer models and hone in on

the consensus direction of the storm. See Figure.

This is often described as the “cone of uncertainty.” The “cone of uncertainty,” shown in

Source: http://www.nhc.noaa.gov/aboutgloss.shtml

Source: http://my.sfwmd.gov/sfwmd/common/images/weather/plots.html

Figure 6, takes all the forecast tracks from a variety of different models and concentrates

on a specific area. Once that area has been identified, the local government enacts their

emergency preparedness plans and communicates with the public.

Figure 6: Cone of Uncertainty

Figure 5: Spaghetti Tracking Model

The risk communication literature review is divided into two sections. The first

section focuses on risk communication literature and Hazleton’s (1993) taxonomy of

public relations strategies. In studying past risk communication literature, it is important

to identify the right variables to measure and common language used. The first two

sections will focus on the message. The second section of the literature review will focus

on Grunig’s (1997) situational theory of publics.

Risk Communication

There are numerous definitions for risk communication. Covelo (1992) defined

risk communication as “the exchange of information among invested parties about the

nature, magnitude, significance, or control of risk (p. 359). This involves “the act of

conveying or transmitting information between interested parties about levels of health or

environmental risks; the significance or meanings of such risks; or decisions, actions, or

policies aimed at managing or controlling such risks” (Davies, Covello, & Allen, 1987, p.

112).

Many risk communication studies use the definition of the National Research

Council (1989). They defined risk communication as

“an interactive process of exchange of information and opinion among

individuals, groups, and institutions. It involves multiple messages about the

nature of risk and other messages, not strictly about risk, that express concerns,

opinions, or reactions to risk messages or to legal and institutional arrangements

for risk management” (p. 21).

This definition stresses the importance of communication to all possible

stakeholders. “Stakeholder involvement is pivotal in the development of a dialogue

intended to result in a risk management or mitigation consensus” (Cole & Fellows, 2008,

p. 214).

In addition, Palenchar (2005) stated that “risk communication provides the

opportunity to understand and appreciate stakeholders’ concerns related to risks generated

by organizations to engage in dialogue to address differences and concerns, carry out

appropriate actions that can reduce perceived risks, and create a climate of participatory

and effective discourse to increase harmony and mutuality” (p. 752-753).

It is critical to understand how to communicate this information. Heath and Abel

(1996) noted that risk communication studies often center upon how technical experts

frame and present technical information to concerned publics in language they can

understand. The National Research Council (1989) stated that risk communication is

“successful only to the extent that it raises the level of understanding of relevant issues or

actions and satisfies those involved that they are adequately informed within the limits of

available knowledge” (p. 21). Satisfaction of risk communication messages relies on two

components. Cole and Fellows (2008, as cited in Rowan 1991) stated that “first, it must

communicate the probabilities and consequences of known risks to affected audiences”

(p. 213). This is a critical part of risk communication. The risk communicators must

present information to the public to instill an act of urgency in mitigation and

preparedness phases. Second, “it should seek consensus among these audiences regarding

a specific course of response and mitigation” (p. 213). It is important to have one

message strategy when communicating risks. Once various messages enter the public

sphere, the public is unsure of the issue and what source to believe. Two common themes

emerged in risk communication literature: trust and credibility.

Trust and credibility are important components in risk communication. “The

source of an organization’s perceived trust and credibility comes from its ability to care,

competent commitment to solve the risk, honesty, and expertise” (Cole & Fellows, 2008,

p.214). Spokespersons, either local or national, must be trusted in communicating this

information. “Residents who demonstrated trust in industry and emergency response

personnel were more likely to gather information, be knowledgeable, and exhibit positive

behavioral intentions regarding emergency response procedures” (Palenchar & Heath,

2002). The more trust individuals have in these officials, the more likely they will be

proactive in adhering to their message. Having universal trust is important. The risk

communicator must know how to communicate to the various publics that will potentially

be affected by the risk. “Risk communication becomes a tool for communication values

and identities as much as being about the awareness, attitudes, and behaviors related to

the risk itself” (Palenchar & Heath, 2007, p.127). The public has to relate to the

individual disseminating the message.

Another aspect of risk communication is care communication. In care

communication, “risks are already known to the audience or appropriate experts, and

risks for which management processes are scientifically determined and accepted by the

audience” (Cole & Fellows, 2008, p.213). This message strategy is informative rather

than persuasive. Using Hurricane Katrina and the New Orleans levees as an example,

Cole and Fellows (2008, cite Lundgren and McMakin, 2004) that the objective of care

communication is to alert an audience to the presence of a risk and to advise appropriate

risk avoidance behavior.

The core focus of my study was derived from two hurricanes: Hurricane Katrina

in 2005 and Hurricane Ike in 2008. These two storms, similar in size, but not strength

garnered two different response plans. Hurricane Katrina was an awful display of

emergency management and Hurricane Ike was a strong representation of message affects

and coordination and trust in the source of the message as well as the message itself. Cole

and Fellows (2008) highlighted the poor display of emergency management mitigation

and preparedness during Hurricane Katrina.

Cole and Fellows (2008) conducted a case study that documented the risk

communication failures during Hurricane Katrina (2005) in the city of New Orleans.

They concluded that inadequate clarity, insufficient credibility, and failure to properly

adapt to critical audiences resulted in a failure of consensus communication and crisis

communication (p. 211). Their findings highlighted some important issues in risk

communication. They found that crisis messages were inadequate, message preparation

prior to the crisis is essential, effective messages must be delivered by credible sources,

and messages must be adapted to encompass a wide variety of different demographic

characteristics.

“Risk communicators are faced with the dual challenge of translating existing and

emergent technical and/or scientific material regarding the anticipated event into lay

person’s terms and arousing an understanding of the severity of the potential

consequences an event may have on the populace” (Cole & Fellows, 2008, p. 211-212).

Due to the difficulty of forecasting landfall coordinates of hurricanes days out,

meteorologists predict the different characteristics of these storms such as paths, landfall,

and strength. “Individuals who do not perceive the risk as personally relevant may

minimize such messages” (p. 212).

In order to have a better understanding of the core principles of risk

communications, studies must be conducted to document the challenges/mishaps. During

Hurricane Katrina, people were hesitant to leave their valuables behind and received

unclear messages from officials or the lacked knowledge on how to evacuate. According

to Cole and Fellows (2008), “hurricane roulette” was present during Hurricane Katrina.

This means that citizens felt lucky they would be able to ride out the storm. Conflicting

evacuation messages left vague and uncertain understanding of what to do and what was

required. “When individuals perceive themselves at risk, their ability to comprehend and

to process information declines significantly” (Cole & Fellows, 2008, p. 224). Language

used in these messages conveyed several meanings. Another variable that led to many not

evacuating was the lack of spokesperson credibility. This led to people not trusting the

messages they received. Messages must be credible from credible people or

organizations.

Another important aspect of understanding risk communication messages is use of

common vocabulary. Hurricane Katrina communication messages were filled with

confusing advisory language and inconsistent messages. This includes adapting these

messages to a variety of target audiences. Factors involved in target audiences are

income, education, race, ethnicity, and residential location. In risk communication, it is

important that the communicator understand the various target audiences that will be

receiving the message and how to respond to their needs. Cole and Fellows (2008) study

featured different factors that led individuals to not respond to risk management

messages. These are past experiences, trust in public officials, lack of knowledge,

vocabulary, target audience, and the role of media.

Past Experience

Dombroski, Fischhoff, and Fischbeck (2006) offer a general approach to

predicting public compliance with emergency recommendations (p. 1675). The approach

starts with a general risk assessment that includes factors that could affect behavior. The

implications of these factors should be used to improve emergency risk assessment

models and improve preparedness for disasters. Different variables play different factors

in risk preparedness. Baker (1991) concluded the most important determinants are actual

risk levels, citizen’s beliefs that their homes are at high risk, and official

recommendations and warnings. The impact of many variables, including risk area,

evacuation notices, housing, storm threat information, hurricane probability forecasts,

hurricane experience, length of residence, hurricane awareness, crying wolf, and

demographics have an effect on preparedness.

Past experience factors into whether you respond to risk communication

mitigation and preparedness messages. Siegrist and Gutscher (2008) investigated the

affects of past experience on mitigation behavior. Results suggested that “people without

flood experience envisioned the consequences of a flood differently from people who had

actually experienced severe losses due to a flood” (p. 771). Weinstein (1989) said that

“past experience seems to be an important factor influencing people’s perception of

hazards” (p. 772). On the contrary, people who had not been affected by a flood strongly

underestimated its effect. Mitigation campaigns are needed to increase knowledge about

these risks. “Risk communication must not focus solely on technical aspects, in order to

trigger motivation for mitigation behavior, successful communication must also help

people to envision the negative emotional consequences of natural disasters” (Weinstein,

1989, p. 771). This study highlighted the need for people to understand that

nonexperience should not equate to low knowledge of a potential hazard and what you

can do to protect yourself and your property. People can be knowledgeable about disaster

risks and still not have the motivation to act accordingly.

Kapuca (2008) examined the role of household preparedness in response to

disasters. Findings suggested that household and individual preparedness is an important

factor in preparedness for natural disasters. Kapuca’s study reconfirms a common theme

in risk communications: complacency. Kapuca found that “households, even with

significant experience with disasters, can be complacent in response to disasters” (p.

526). Why does personal experience lead to complacency?

Martin, Bender, and Raish (2007) investigated the cognitive perceptual process

people go through when faced with risks. The variables they looked at were the role of

motivation, decision stages of risk readiness, and subjective knowledge. Subjective

knowledge, based on someone’s direct or indirect experience, was essential in

preparedness. They investigated a number of risk-mitigating actions taken by those in risk

situations. They concluded that “personal experience can have a powerful impact on

recognition of risk and the willingness to protect oneself from risk” (p. 897). These past

experiences become the basis for individual beliefs in their own knowledge about risk.

Halpern, Millstein, Ellen, Adler, Tschann, and Biehl (2001) found that “participants who

had experienced a natural disaster or engaged in a particular risk behavior estimated their

chance of experiencing a negative outcome resulting from that event or behavior as less

likely then individuals without such experience” (p. 120). The findings suggest that

behavioral experiences drive risk judgments.

Trust and Credibility

An important aspect in risk communication is the source of the message and the

perceived trustworthiness of the source. During Hurricane Katrina, a segment of the

population of New Orleans didn’t trust officials disseminating the evacuation messages.

This was because different federal, state, and local officials were disseminating different

messages to the same audience. In addition, it’s important to understand how trust can be

built around these issues.

Heath and Abel (1996) discovered “that communities that engage in more

extensive efforts to create emergency response systems and inform residents of those

measures increase the risk of tolerance of community members” (p. 151). They

concluded, “when community officials provide emergency response systems and the

information citizens need to protect themselves in the event of an emergency those efforts

can be demonstrated to foster support for the industry” (p. 151). The argument here is that

“responsible parties -- industry and government -- in communities where potentially

dramatic risks exist are wise to acknowledge those risks and to work proactively to

inform members of the public about the protective measures they can take in the event of

emergency” (p. 153).

Risk communicators must be proactive in the way they disseminate information to

the public. This involves knowing that a risk exists from the declaration of experts in the

field. Heath and Abel (1996) found that “although community members are concerned

that unfavorable events will occur, they believe emergency response personnel are

prepared to respond properly” (p. 158).

In addition, Heath and Abel’s (1996) found that television messages were the

preferred way to be contacted by emergency response personnel. A key aspect of their

findings was how respondents trusted government officials. Respondents “seem to trust

their own judgment more than that of officials or do not know the advantages of taking

the emergency response measures recommended by emergency response experts” (p.

165). They concluded that “people may trust government and industry more when those

entities acknowledge potential dangers and give proactive solutions to problems rather

than attempt to downplay them by stressing the improbability that emergencies will

occur” (p. 170).

Lack of Knowledge

Lack of knowledge also has been found to impact people’s response to emergency

situations. Previous research suggests that people did not know what to do during an

emergency. According to Heath and Abel (1996) “Residents believe government officials

are prepared to respond properly and to serve as credible sources of opinion” (p.

166). Trust is a fluid attribute the public looks for in government officials.

Baker (1995) studied the effect of hurricane probabilities on public response.

Numerous hypothetical threat scenarios were used to assess hurricane probability

forecasts and risk variables associated with public response. “The most important

practical finding in hurricane preparedness is the local officials’ advice or orders

regarding evacuation (p. 146). “This was the most important element affecting

evacuation, regardless of whether probabilities are included in people’s information of

not” (Baker, 1995, p. 146). The importance of trust in public officials was critical to the

outcome of Hurricane Katrina in 2005. Residents said that they had the least trust for

public officials and the messages they were disseminating. Baker (1995) found that

people often feel that they are more knowledgeable than public officials in mitigation and

preparedness activities.

Heath and Palenchar (2000) found that because “concern remains high that risk

events are likely to occur and harm community safety, citizens are willing to become

knowledgeable of emergency response measures” (p. 131). This knowledge “gives

citizens a greater sense of control, which may translate into trust for industry and city

emergency response efforts” (p. 131). Mitigation campaigns can build trust in favor of

community and government officials.

McEntire and Myers (2007) discussed what local governments must do to prepare

for various disasters. They identified a step-by-step approach to establish a process of

local ordinances, assessing risk, creating emergency operations plans, and educating the

public.

Through the mitigation phase, “effective public relations efforts can build

community support through collaborative, community-based decisions regarding the

kinds of risks that exist, and the emergency response measures that can be initiated as

needed for public safety” (Heath & Palenchar, 2000, p. 132). According to the authors,

risk management, perception, and communication research address five themes. These

themes are the likelihood that specific risks will occur, who will be affected if they occur,

magnitude of effect, mitigation of the occurrence, and mitigation of impact.

One of the primary goals of the practitioner is to disseminate information to

“potentially affected communities so that they know that emergency warning and

response systems are in place and that measures can be taken to reduce personal exposure

to the risk if it occurs” (p. 135). According to Heath (1995), “risk communication

campaigns are best when they are coupled with community relations efforts that include

messages that respond to citizens’ desire to know what to do to increase their safety in the

event of a health or life threatening emergency” (p. 135).

Some of the key variables in these campaigns are trust and cognitive involvement.

“Trust is a central factor in predicting whether members of a community accept and rely

on the conclusions and recommendations of people who are trained in science, business

operations, engineering, and emergency management.” (Heath, 1995, p. 135) Cognitive

involvement states that the “more people believe that some dire consequence can result,

the greater their level of cognitive involvement” (p. 136).

Some characteristics of individuals who are cognitively involved are that they

“acquire, pause to consider, and evaluate information more thoroughly” (Heath, 1995, p.

136). An interesting result of the study showed that “people who are cognitively involved

have a higher sense of risk, are less trusting of government and industry officials” (p.

149). Public officials must understand the different comprehension levels between

professional risk communicators and the general population.

Knocke and Kolivras (2007) studied flash flood awareness in southwest Virginia.

They concluded that there is a “knowledge gap between flood experts and the general

public about the level of perceived risk that the latter has toward the powerful flood

waters” (p. 155). The knowledge gap affects communication capabilities and efficiency of

the warning process. Their research found that even though people had knowledge of

flash floods, it wasn’t enough to garnered a proper level of awareness. To effectively

communicate this information to the public, new warning methods must be developed.

Vocabulary

Vocabulary plays an important part in how individuals understand mitigation and

preparedness messages. It must consist of general terms that the majority of the

population can comprehend. Communicating a message with unclear language can cause

individuals not to take the recommended action.

Hellier, Aldrich, Wright, Daunt, and Edworthy (2007) studied warning signal

words and the meaning of their usage. These signals are often used on “warning signs and

labels to denote the level of hazard implied by the situation they indicate” (p. 323). They

conducted a multidimensional analysis of rating 17 signal words. In doing so, three

dimensions emerged: the level of hazard implied by the signal words, the extent to which

they explicitly implied risk and the explicitness of the instruction given.

The results support a general code of using certain signal words for certain

hazards. This “suggests that that there might be utility in mapping signal words to the

conditions that they indicate in terms of the extent to which the situation or product

constitutes explicit risk” (Hellier, et al., p. 323). Citing Wogalter and Silver (1990), they

argue that signal words recommended for use are too limited in number and are over

used. This results in desensitization to them and habituation. However, previous research

studies have revealed a consistent relationship between signal words and perceived

hazards. This research “supports the use of signal words in warning implementation to

quantify hazard and also suggest that the dimensionality of signal words can be further

refined to include not only hazard but also the explicitness with which risk is implied” (p.

337). Common vocabulary needs to be agreed and used among various stakeholder

organizations.

Manoj and Baker (2007) discussed the lack of common vocabulary between

response organizations, organizations and citizens. This can be attributed to the problems

related with mitigation and preparedness. They indicate that the primary challenge in

responding to both natural and man-made disasters is communication.

The use of signal words and warnings were echoed during Hurricane Ike. The

same message was being disseminated through all communication channels. It will be

interesting to see if ‘certain death’ language will face desensitization due to prior use and

the outcome ‘better than expected’ outcome of the Hurricane Ike.

Target Audience

When communicating risk mitigation and preparedness messages, it is apparent

that the communicator knows their target audiences and how best to communicate with

them. Highlighted in the mess that followed Hurricane Katrina, it is important that all

demographics are receiving the same message regardless of ethnicity, religion, gender,

socio-economic status, etc. It was evident in New Orleans that the African-American

population wasn’t communicated with effectively.

Eisenman, Cordasco, Asch, Golden, and Glik (2007) studied the factors that

influenced evacuation decisions in impoverished communities. Using Hurricane Katrina

as a case study, they indicated that family, friends, and community organizations played a

positive and negative role evacuation decisions. Through a series of qualitative

interviews, they concluded that disaster plans must account for situations in less affluent

communities. Questioning the orders of local officials, one respondent said, “the last

storm we had there, it was more people got hurt on the highway traveling away from the

storm, running out of gas, accidents, than it would have been if they stayed home” (p.

S111). The obstacles they encountered should lead to new strategies that have an

emphasis on community-based communication and preparation strategies. Subjective

norms also played a role in whether to evacuate.

McIvor and Paton (2007) sought to further develop a model for natural hazard

preparedness. They examined the role of attitudes, mitigation and social norms plays in

natural hazards. Their study examined “whether social-cultural factors influence the

decisions people make regarding their relationship with natural hazards” (p. 79). People’s

attitudes and social norms influence their perception of these hazards and ultimately how

they prepare for them.

McIvor and Paton (2007) concluded that mitigation activities must be on-going

within the community in future risk communication: “People living in communities at

risk from natural hazards continue to demonstrate poor knowledge of risk mitigation

procedures and a reticence to adopt protective measures” (pp. 79-80). It is important to

know your target audience when developing risk communication messages. “Perceptions

of risks and hazards are culturally and socially constructed; people interpret it in the

context of their experience, beliefs, and expectations” (p. 80). Subjective norms and

attitudes influence the efficacy of engaging in risk mitigating behavior. Knowing how to

communicate to your target audience is critical.

Connelly and Knuth (1998) studied how information format can influence the

extent to which target audiences understand and respond to risk-related information. The

“purpose of the study was to measure anglers’ perceptions and anticipated responses to

various health advisory presentation formats so that risk communicators producing

advisories could consider likely audience response when preparing information for

anglers” (p. 652). Their study examined four components of risk information

presentation: reading level, diagram vs. text, tone, use of information. They stated that

“the manner in which risk information is presented to target audiences is a critical

influence on their ultimate response in terms of attitudes, behaviors, and perceptions

related to the risk” (p.650). Doing research on your audience “regarding information

needs and communication formats may help clarify which approaches to take” (p. 649).

They found that the use of graphics could improve the understanding of risk information

by certain audiences. In addition, the study found that there is no consensus

communication strategy that has similar effects on all target audiences. The media also

plays an important role in risk communications.

Media

Perez-Lugo (2004) analyzed the role of the media in the sociology of natural

disasters. Media has been “mainly viewed as management tools used to influence

people’s preparedness and response to natural disasters” (p. 210). Focusing in on the

media-audience relationship during natural disasters, it was revealed that “media also

have latent functions in disasters, which consist of emotional support and

companionship” (p.222).

Perez-Lugo (2004) concluded that disaster research points out the role of the mass

media during disasters as crucial in disseminating information in a quick efficient

manner. “There importance lies in their power to increase preparedness and facilitate

recovery by changing people’s attitudes about natural hazards” (p. 211). Mitigation and

preparedness is key when dealing with risk communications. During the mitigation phase,

“the media are considered a disaster information provider through coverage of non-local

disasters, which helps the community raise disaster awareness and prepare for future

events” (p. 212). In the preparedness phase, “the mass media provide factual information

about the approaching hazard and tips to prepare for its impact” (p. 212). Through

quantitative interviews, the study found a lack of interest in the preparedness phase.

“Instead of looking for ways on how to secure life and property, they wanted the physical

location of the hurricane” (Perez-Lugo, 2004, p. 218). People said they take action based

on previous personal and collective experiences with hurricanes. “The media-audience

relationship remains a very important aspect of the people’s coping strategies during

disasters” (p. 223). Many news organizations have the capabilities to broadcast live in

disaster situations.

“Media presence during Hurricane Katrina allowed the world not only to see the

atrocities experienced by the evacuees but also to see clearly and repeatedly the

contradictions and failings by all levels of government” (Guion, et al., 2007, p. 23). The

mass media serve various functions for society, one of which is a channel for emergency

managers to disseminate information in times of imminent danger (p. 25). Media

disseminate this information voluntarily. Mass media can play a critical role during the

mitigation phase because media coverage contributes to the formation of public attitudes,

which in turn influence legislative actions (p. 21). Citing Fishman and Casarett (2006),

Guion, et al. (2007) noted that this information can shape beliefs, attitudes, and perceived

norms and can subsequently influence behavior.

Public Relations Strategies

Hazleton (1993) noted that symbols are the primary means of accomplishing

public relations. In order to better understand how symbols are developed and used for

the purpose of communicating with others (p. 89), Hazleton developed a matrix to

analyze public relations symbols. The function element of the matrix associates the

audience and the assumptions about message affects (Hazleton, 1993). Characteristics of

the audience must be taken into account in the classification of these messages according

to their functional characteristics (Hazleton, 1993). This allows communicators to

postulate about motivational, cognitive, and behavioral characteristics of audiences

(Hazleton, 1993).

The matrix also analyzes message effects and message processing at the

psychological level. These “messages may be understood as objects to be understood by

individuals” (Hazleton, 1993 p. 91). The psychological level is most apparent in the

public relations strategic planning process. This level seeks to understand how people

respond to and understand communication (Hazleton, 1993). Organizations use symbols

to accomplish goals related to the public relations function. The functions proposed at

this level represent “the goals of public relations in terms of the impact and meaning of

messages to individual recipients” (p. 94).

Influenced by the social change literature of Zaltman and Duncan (1976),

Hazleton derived four message functions that may accurately capture mass media based

strategies (Hazleton, 1993). He also developed two message functions based on Grunig’s

two-way symmetrical and two-way asymmetrical public relations models. These are the

most commonly used in public relations strategy creation at the psychological level.

Hazleton used these six functions to develop a taxonomy of public relations strategies

that organizations typically use when communicating with publics (Werder, 2006). The

seven strategies are: facilitative, informative, persuasive, promise and reward, threat and

punishment, bargaining, and cooperative problem solving. The definitions Hazleton

(1993) used to describe these strategies are provided below.

Facilitative

A facilitative strategy is accomplished by making resources available to an

audience that allow them to act in ways that they are already predisposed to act.

Resources may be tangible artifacts, such as tools or money, or they may be directions

which tell someone how to accomplish a particular action.

Informative

An informative strategy is based upon the presentation of unbiased facts. This

strategy does not draw conclusions, but presume that the audience will infer appropriate

conclusions from accurate data. Informative messages may suggest a variety of

alternative solutions to problems. Informative messages are characterized by the use of

neutral language, and organic or natural patterns of organization.

Persuasive

Persuasive strategies are characterized by appeals to audience values, or affect

and a biased presentation of information. They may use language which is not neutral and

reflects the importance of the issue and the involvement of the source in the situation.

These types of messages are directive in the sense that they provide a call for action

either tacitly or explicitly.

Promise and Reward

Promise and reward strategies involve the exercise of power to obtain compliance.

They include a directive and a contingent outcome which may be explicitly or tacitly

linked to performance of the directive request. They imply or point out that the source of

the message controls an outcome desired by the receiver of the message.

Threat and Punishment

Threat and punishment strategies involve the exercise of power, threats and

promises to obtain compliance. They include a directive and a contingent outcome which

may be explicitly or tacitly linked to performance of the directive request. They imply or

point out that the source of the message controls an outcome feared or disliked by the

receiver of the message.

Bargaining

Bargaining is characterized by an organized exchange of messages between

communicators. Strategic withholding of information and deceptions designed to mislead

others concerning your acceptable range of alternatives and to discover the other party’s

acceptable range of alternatives are used. Bargaining communication is characterized by

the use of contrasting symbols which differentiate groups, such as “we” and “they”.

Cooperative Problem Solving

Cooperative problem solving messages reflect a willingness to jointly define

problems and solutions to problems. Cooperative problem solving messages are

characterized by the use of inclusive symbols, “we” and not “they.” In contrast to

bargaining, problem solving is characterized by an open exchange of information.

Situational Theory of Publics

Using the concept of ‘publics’ derived from classic public opinion theorists

Dewey and Blumer, Grunig “formalized those theories and provided means for

identifying and measuring publics and their opinions” (Grunig, 1997, p. 9). Dewey and

Blumer concluded that “publics arise around issues or problems that affect them.” After

they recognize that a problem affects them, “publics organize into issue groups to

pressure organizations that cause the problems or to pressure government to constrain or

regulate those organizations” (p. 9). Grunig and Hunt (1984) describe a public as a

loosely structured system whose members, existing with a population or linkage, detect a

problem and behave as though they were one body to solve the problem.

According to Grunig and Hunt (1984), the situational theory of publics “states that

communication behaviors of publics can be best understood by measuring how members

of publics perceive situations in which they are affected by such organizational

consequences” (Hamilton, 1992, p. 124). In essence it describes different aspects of

communication effects on publics. Grunig (1997) noted that publics begin as

disconnected systems of individuals experiencing common problems; but they can evolve

into organized and powerful activist groups” (p. 9).

“The situational theory provides a means of segmenting a general population into

groups relevant to public relations practitioners” (Grunig, 1997, p. 8). This is an

important concept when it comes to creating public relations campaigns. The foundation

of the theory rests on the balance of trying to “predict the differential responses most

important to public relations professionals: responsiveness to issues; amount of and

nature of communication behavior: effects of communication on cognitions, attitudes, and

behaviors; and the likelihood of participation in collective behavior to pressure

organizations” (Grunig, 1997, p. 9). Vasquez (1993) adds that “publics are recognizable

based on their shared behaviors, and the communication behavior publics can be

understood by measuring how members of a public perceive situations in which they are

affected by organizational consequences” (p. 208).

Aldoory (2001) stated that “the situational theory of publics is one of the most

useful theories for understanding why publics communicate and when they are most

likely to communicate.” She points out that there is a significant gap in our understanding

of the situational theory regarding any antecedent factors that may help explain

involvement, constraint recognition, and problem recognition, the three independent

variables in the theory. Most research has studied the dependent variables and the

predictability of the independent variables, and most research has found strong support

for the theory.

Hallahan (1999) defined a public as a group of people who relate to an

organization, who demonstrate varying degrees of activity or passivity and who might or

might not interact with others concerning their relationship.

An active public is described as one who seeks information. Hamilton (1992)

citing Grunig (1989) characterized active publics. These “people communicating actively

develop more organized cognitions, are more likely to have attitudes about a situation,

and more often engage in a behavior to something about the situation” (p. 124). A passive

public is the opposite of an active public. They make little to no effort to seek

information.

The situational theory of publics consists of three independent variables and two

dependent variables. The two dependent variables consist of active and passive publics.

They have also been described by their characteristics as information-seeking and

information processing. Clarke and Kline (1974) described the two dependent variables

as premeditated information seeking, “the planned scanning of the environment for

messages about specified topic.” Information processing “describes message discovery

the unplanned discovery of a message followed by continued processing of it.”

The theory is comprised of three independent variables; problem recognition,

constraint recognition and level of involvement. “The three concepts together predict not

only when people will communicate; they also predict that active communication

behavior more often results in effects of communication -- cognitions, attitudes,

individual and collective behaviors --than does passive communication behavior” (Grunig

& Repper, 1992, p. 136). Grunig (1997) describes the independent variables as

situational. “They describe perceptions that people have of specific situations, especially

situations that are problematic or that produce conflicts or issues” (p.10).

Problem Recognition

Problem recognition stated in Aldoory (2001) is the extent to which individuals

recognize that issues or events are problems to be concerned about” (p. 165). Major

(1993) found that the likelihood of communication is increased by problem recognition,

such that, among people who face problems, information seeking and processing are

likely to occur even under low involvement situations. Hamilton (1992) found that

problem recognition did not account for active media use and that other variables played

a role.” Citing Major (1993), Aldoory and Sha (2007) found that the likelihood of

communication is increased by problem recognition, such that among people facing

problems, information seeking and processing are likely to occur even under low

involvement situations.

Constraint Recognition

The second independent variable is constraint recognition. “People do not

communicate about problems or issues about which they believe they can do little or

about behaviors they do not believe they have the personal efficacy to execute” (Grunig

and Repper, 1992, p. 135). Constraint recognition represents the extent to which

individuals perceive obstacles, or barriers, in a situation that limit their freedom to plan

their own behavior (Werder, 2005, p. 226). Citing Grunig and Ipes (1983), Aldoory and

Sha (2007) concluded, “for a campaign to move people to develop organized cognitions

and perhaps to change their behavior, it must show people how they can remove

constraints to their personally doing anything about the problem.

Level of Involvement

Level of involvement is the extent to which an issue, problem, or situation has

personal relevance to an individual (Werder, 2005, p. 226). Pavik (1988) defines it as “a

perceived emotional connection or relevance, involvement increases the likelihood of

individuals attending to and comprehending messages.” This creates an active audience.

These high-involved publics often seek additional information to supplement their

beliefs. Lovelock and Weinberg (1984) stated that level of involvement is the degree of

importance or concern that a product or behavior generates in different individuals.

Research on involvement and other independent variables has led to a greater

understanding of the probability of information seeking and information processing.

These studies and others have shown that, in general, members of a public are more likely

to seek information and communicate actively when they perceive an issue to be a

problem. People communicating actively develop more organized cognitions, are more

likely to have attitudes about a situation and more often engage in a behavior to do

something about the situation. An active public often seeks information through a variety

of media, interpersonal contacts, and specialized channels. Passive publics are more

likely to process information from mass media (Aldoory, 2007).

The situational theory of publics has been extended over the years by researchers

testing different situations on the independent variables. Personal and impersonal

dimensions of the independent variables have been created and tested. They argued that

independent variables, such as level of involvement, were driven by either egoistic

concerns or altruistic concerns. Results of these studies indicated that distinguishing

situations by personal and impersonal dimensions usefully extended the situational theory

because of the dimensions’ improvement in segmenting publics. Goal compatibility is the

extent to which goals or objectives of one party are similar to and coincide with the goals

and objectives of another party (Page & Hazleton, 1999).

Hazleton’s (1993) public relations strategies will be the foundation for my

message treatments. These strategies are typically used by organizations to communicate

a message. Current research does not measure these strategies in a risk communication

context. This study will help better understand the type of message strategies that are

most effective when the public is at risk. Messages for my treatments will be derived

directly from the definitions of the strategy types indentified by Hazleton.

Grunig’s situational theory of publics will help identify the effects of these message

strategies on the public. By understanding how the public responds to problem

recognition, constraint recognition and level of involvement will allow experts in the risk

communication field to better craft messages to garner the necessary/desired response

from the public. In risk communications, reaching your target audience is critical. By

understanding message strategies and the variables identified in the situational theory of

publics, organizations will be able to reach their target publics and yield the desired

outcomes.

The purpose of this study is to understand what message source and strategy type

is most effective in emergency management communications. Below are the hypotheses

and propositions this study seeks to test.

H1: In emergency communication, message source will influence receiver variables.

P1.1: A national message source (NWS) will produce higher problem recognition

than a local message source (HCG).

P1.2: A national message source (NWS) will produce higher constraint

recognition than a local message source (HCG).

P1.3: A national message source (NWS) will produce higher level of involvement

than a local message source (HCG).

P1.4: A national message source (NWS) will produce higher expertise than a local

message source (HCG).

P1.5: A national message source (NWS) will produce higher trust than a local

message source (HCG).

P1.6: A national message source (NWS) will produce more positive attitudes than

a local message source (HCG).

P1.7: A national message source (NWS) will produce higher information-seeking

than a local message source (HCG).

H2: In emergency communication, message strategy will influence receiver variables.

P2.1: Information seeking will be the highest when the threat and punishment

strategy is used.

P2.2: Information seeking will be the lowest when the informative strategy is

used.

H3: Level of involvement will produce the strongest effect on information seeking

behavior.

H4: Constraint recognition will produce the weakest effect on information seeking

behavior.

The next chapter explains the methodology used in this study. The experimental

procedures, the treatment conditions, the instrumentation, and the data analysis

procedures are included.

Chapter Four

Methodology

The purpose of this study is to understand what strategy type and message source

is most effective in emergency management communications. Below are the hypotheses

and propositions this study tests:

H1: In emergency communication, message source will influence receiver variables.

P1.1: A national message source (NWS) will produce higher problem recognition

than a local message source (HCG).

P1.2: A national message source (NWS) will produce higher constraint

recognition than a local message source (HCG).

P1.3: A national message source (NWS) will produce higher level of involvement

than a local message source (HCG).

P1.4: A national message source (NWS) will produce higher expertise than a local

message source (HCG).

P1.5: A national message source (NWS) will produce higher trust than a local

message source (HCG).

P1.6: A national message source (NWS) will produce more positive attitudes than

a local message source (HCG).

P1.7: A national message source (NWS) will produce higher information-seeking

than a local message source (HCG).

H2: In emergency communication, message strategy will influence receiver variables.

P2.1: Information seeking will be the highest when the threat and punishment

strategy is used.

P2.2: Information seeking will be the lowest when the informative strategy is

used.

H3: Level of involvement will produce the strongest effect on information seeking

behavior.

H4: Constraint recognition will produce the weakest effect on information seeking

behavior.

Methodology

A controlled experiment was conducted to test the four hypotheses and nine

propositions posited by this study. Werder (2005) designed an experiment to test the

effect of Hazleton’s public relations strategies on the receiver variables of problem

recognition, constraint recognition, and level of involvement. This study seeks to

replicate and extend that study in the context of emergency communication.

Research Participants

Research participants were recruited from five undergraduate classes at a large

southeastern public university: Principles of Public Relations; Introduction to

Advertising; Mass Communications and Society; Public Relations Research; and

Marketing Management. The sample consisted of a total of 147 participants. The

participants included 96 females and 51 males. The age range of students who

participated in this study was 18 to 44. The average age was 22. Those who volunteered

to participate in the study did not receive any incentive.

Stimulus Materials

Using a 2 x 6 factorial design, 12 treatment conditions and two control treatments

were created. Each participant in the survey was randomly assigned one of twelve

different treatment conditions or one of two control conditions. A minimum of 10

questionnaires were completed for each treatment and control condition. A total of 147

questionnaires were completed. These treatments varied by message source and strategy

type. Six of the treatment conditions created were from the National Weather Service and

six were from the Hillsborough County Government.

The six message types were derived from Hazleton’s (1993) public relations

message strategies: informative, facilitative, persuasive, promise and reward, threat and

punishment and cooperative problem solving. The message treatments were in the form

of hurricane press advisories with information coming directly from past hurricane

releases. Each of the press advisories for all 12 treatments was almost identical. The

content was identical in all 12 treatments. The format was replicated from an official

hurricane advisory sent out to the media/public during a hurricane. The name of the

hurricane, Jacob, was chosen. The name of the storm is not related to any previous

hurricane that had hit Florida.

The information in the press advisory was changed to simulate a category three

hurricane approaching Tampa Bay. Cities/locations in the advisory were changed to make

the simulation as real as possible. The only changes that were made to distinguish the

differences between the two message sources were the headline, logo and sidebar. The

headline read, “Hillsborough County Government Hurricane Advisory” and “National

Weather Service Hurricane Advisory.”

The logos of both these organizations were placed above the sidebar in the same

location and of the same size. The manipulations used to test message strategy were

placed in the sidebar. The sidebar was offset from the rest of the material in a grayshaded

box. The boxes on all the treatments were the same size. The manipulations ranged

between 13 and 17 lines of text. The word count ranged from 36 to 50. The control

messages were unrelated to the issue featured in the treatment materials. They consisted

of a press release about a new hiring at the National Weather Service Office in Wisconsin

and a hurricane preparedness convention in Tampa. An identical survey followed all 12

treatments and the control condition.

Instrumentation

The questionnaire consisted of the variables identified in Grunig’s situational

theory of publics. Twenty-six items were created to measure problem recognition,

constraint recognition, level of involvement and information-seeking behavior. Responses

to were rated on a 7-point Semantic-differential scale. Participants were asked to rate the

indicated statements, 1 (strongly disagree), 2 (disagree), 3 (somewhat disagree), 4

(neutral/no opinion), 5 (somewhat agree), 6 (agree), and 7 (strongly agree) on a

continuum.

Each variable was tested by the following statements. Problem recognition was

measured by the subsequent statements: 1) Based on the weather advisory I read, I

believe this situation qualifies as an emergency; 2) I recognize the existence of a

weatherrelated emergency situation; 3) I don’t believe this emergency situation is serious;

4) I want to understand this emergency situation better; 5) I need to seek out additional

information to better understand this emergency situation.

Constraint recognition was measured by these statements: 1) I believe that I am

not able to evacuate; 2) I cannot do anything about this emergency situation; 3) I believe

that there are constraints or obstacles that limit my ability to evacuate; 4) I do not

understand this evacuation enough to do anything about it; 5) I do not understand this

emergency situation enough to do anything about it; 6) I do not have the ability to make a

difference in the outcome of this emergency situation.

Level of involvement was measured by the following statements: 1) I am

personally affected by this emergency situation; 2) I have strong opinions about this

emergency situation; 3) I have strong opinions about evacuation; 4) I am personally

affected by this evacuation; 5) This evacuation does not involve me; 6) This emergency

situation does not involve me.

Three statements were used to measure information-seeking behavior: 1) I will

actively seek more information about this emergency situation; 2) I plan to seek out

additional information on ways I can better prepare for this emergency situation; 3) I

don’t want any more information about this emergency situation.

In addition, trust and expertise were measured by using three statements each.

These questions changed if someone received a National Weather Service treatment or a

local government treatment. Trust was measured on the National Weather Service

questionnaire by the following items: 1) I trust information provided by the National

Weather Service; 2) The National Weather Service is a trustworthy organization; 3) The

National Weather Service is a credible organization. Expertise was measured by: 1) The

National Weather Service has adequate expertise to handle this emergency situation; 2)

The National Weather Service provides the most accurate information about emergency

situations; 3) The National Weather Service is knowledgeable about emergency

preparedness.

Trust was measured on the local government questionnaire by the following

statements: 1) I trust information provided by the local government; 2) The local

government is a trustworthy organization; 3) The local government is a credible

organization. Expertise was measured by: 1) The local government has adequate expertise

to handle this emergency situation; 2) The local government provides the most accurate

information about emergency situations; 3) My local government is knowledgeable about

emergency preparedness.

Attitudes toward the National Weather Service and the Hillsborough County

Government were measured by a semantic-differential scale. The terms measured

consisted of bad : good, negative : positive, incompetent : competent, and not important :

important.

The last section of the questionnaire asked participants to provide demographic

information. This included gender, age, ethnicity, type of dwelling, highest level of

education completed, registered voter, and number of children they have.

A pretest was conducted to determine the validity of the message treatments used

in this study. This manipulation check was performed to assess the degree to which the

weather emergency situation message treatments agreed with Hazleton’s (1993) public

relations strategy definitions. An expert panel, consisting of 7 graduate students agreed

that the treatments reflected the definition.

Data analysis was conducted using SPSS 15.0 for Windows. An alpha level of .05

was required for significance in all statistical procedures, which included reliability

analysis, correlation analysis, linear regression analysis and analysis of variance

(ANOVA). Due to the sampling procedure used, not all treatments had the same number

of responses. The next chapter presents the results from the study.

Chapter Five

Results

The purpose of this study is to understand what message source and strategy type

is most effective in emergency management communications. This study tested the

following hypotheses and propositions:

H1: In emergency communication, message source will influence receiver variables.

P1.1: A national message source (NWS) will produce higher problem recognition

than a local message source (HCG).

P1.2: A national message source (NWS) will produce higher constraint

recognition than a local message source (HCG).

P1.3: A national message source (NWS) will produce higher level of involvement

than a local message source (HCG).

P1.4: A national message source (NWS) will produce higher expertise than a local

message source (HCG).

P1.5: A national message source (NWS) will produce higher trust than a local

message source (HCG).

P1.6: A national message source (NWS) will produce more positive attitudes than

a local message source (HCG).

P1.7: A national message source (NWS) will produce higher information-seeking

than a local message source (HCG).

H2: In emergency communication, message strategy will influence receiver variables.

P2.1: Information seeking will be the highest when the threat and punishment

strategy is used.

P2.2: Information seeking will be the lowest when the informative strategy is

used.

H3: Level of involvement will produce the strongest effect on information seeking

behavior.

H4: Constraint recognition will produce the weakest effect on information seeking

behavior.

Descriptives

Descriptive statistics were conducted to find out information about the

participants. Of the 147 participants in the study, 34.7% (n = 51) were male and 65.3% (n

= 96) were females. All percentages reflect the valid sample. Table 3 shows the

participants race identification.

Table 3: Participants Race Identification

Frequency

Percent

Valid Percent

Cumulative

Percent

Valid

Caucasian

Hispanic

100

68.0

68.5

78.8

10.2

10.3

African-American

9.5

9.6

88.4

Asian

American Indian

3.4

91.8

92.5

Pacific Islander

1.4

93.8

Other

6.1

6.2

100.0

Total

146

99.3

100.0

Missing

100.0

Total

147

Participants identified themselves as Caucasian 68.5% (n = 100). The highest

level of education completed was ‘some college’ at 92.5% (n = 135). The type of

dwelling was either ‘house’ 42.6% (n = 58) and ‘apartment’ 52.9% (n = 72). Those who

surveyed did not have any children 97.1% (n = 132). Most people surveyed were

registered voters at 92.6% (n = 126). When asked if they would need to assist an elderly

family member during an evacuation, 39.7% (n = 54) indicated yes and 60.3% (n = 82)

indicated no. The mean age of the participants was 21.86. The ages ranged from 18 to 44.

Data analysis began with an examination of descriptive statistics for all items used

to measure the receiver variables of problem recognition, constraint recognition, level of

involvement, expertise, trust, attitude, and information seeking behavior. The item means

and standard deviations are shown in Table 4. Item means ranged from 2.65 to 6.12. The

item “I believe that I am not able to evacuate,” which measured constraint recognition,

produced the lowest mean of 2.65. The item “I recognize the existence of a weather

related emergency situation,” which measured problem recognition, produced the highest

mean of 6.12.

For problem recognition, the item “I recognize the existence of a weather-related

emergency situation” produced the highest mean (M = 6.12, SD = 1.101). “I need to seek

out additional information to better understand this emergency situation” produced the

lowest mean (M = 4.71, SD = 1.640).

Constraint recognition is reversed when looking at the mean. The item “I believe

that I am not able to evacuate” produced the highest mean (M = 2.65, SD = 1.488) and the

item “I believe that there are constraints or obstacles that limit my ability to evacuate

produced the lowest mean (M = 3.85, SD = 1.650).

For level of involvement, the item “This emergency situation does not involve me

had the highest mean (M = 5.55, SD = 1.540). “I have strong opinions about this

emergency situation” produced the lowest mean (M = 4.02, SD = 1.474).

The source in the items used to measure expertise and trust were changed to

match the different treatment conditions. The source was the National Weather Service

and the Hillsborough County Government in the different treatment conditions.

“The source is knowledgeable about emergency preparedness” produced the

highest mean under expertise of the source, (M = 5.26, SD = 1.339). “Expertise of the

source provides the most accurate information about emergency situations” produced the

lowest mean (M = 4.45, SD = 1.350).

For trust of the source, “The source is a credible organization” produced the

highest mean (M = 4.92, SD = 1.405). “The source is a trustworthy organization’

generated the lowest mean (M = 4.81, SD = 1.577).

An “important” attitude toward the source produced the highest mean (M = 5.72,

SD = 1.213). A “positive” attitude toward the source produced the lowest mean (M =

5.08, SD = 1.243)

For information seeking behavior, the statement “I don’t want any more

information about this emergency situation” had the highest mean (M = 5.70, SD =

1.655). “I will actively seek more information about this emergency situation” produced

the lowest mean (M = 4.71, SD = 1.703).

Table 4: Item Mean and Standard Deviation

Minimum

Maximum

Mean

Std. Deviation

1 PR Based on the advisory I

read, I believe this situation

qualifies as an emergency.

146

5.47

1.405

2 PR I want to understand

this emergency situation

better.

145

4.83

6.12

1.692

1.101

13 PR I recognize the

existence of a weatherrelated

emergency situation.

146

Table 4. Continued

Minimum

Maximum

Mean

Std. Deviation

14 PR I need to seek out

additional information to

better understand this

emergency situation.

146

4.71

1.640

21 PR REV I don't believe

this emergency situation is

serious.

146

5.51

3.57

3.66

3.33

2.98

3.85

2.65

5.14

1.496

3 CR I cannot do anything

about this emergency

situation.

146

1.922

4 CR I do not have the ability

to make a difference in the

outcome of this emergency

situation.

146

1.884

7 CR I do not understand this

evacuation enough to do

anything about it.

146

1.801

1.595

1.650

11 CR I do not understand

this emergency situation

enough to do anything about

it.

146

15 CR I believe that there are

constraints or obstacles that

limit my ability to evacuate.

146

23 CR I believe that I am not

able to evacuate.

146

1.488

Table 4. Continued

10 LI I am personally

affected by this emergency

situation.

145

1.597

Minimum

Maximum

Mean

Std. Deviation

12 LI REV This evacuation

does not involve me.

146

5.14

1.772

16 LI REV This

emergency situation does not

involve me.

17 LI I have strong

opinions about this

emergency situation.

146

5.55

4.02

4.74

1.540

1.474

146

20 LI I am personally

affected by this evacuation.

145

1.625

26 LI I have strong opinions

about evacuation.

146

4.05

4.71

5.26

4.45

1.379

1.424

1.339

6 EXP The SOURCE has

adequate expertise to handle

this emergency situation.

147

9 EXP The SOURCE is

knowledgeable about

emergency preparedness.

146

18 EXP The SOURCE

provides the most accurate

information about emergency

situations.

146

1.350

Table 4. Continued

8 TRUST The SOURCE is a

trustworthy organization.

146

4.81

4.88

1.577

1.402

1.405

19 TRUST I trust

information provided by the

SOURCE.

146

25 TRUST The SOURCE is

a credible organization.

146

4.92

Minimum

Maximum

Mean

Std. Deviation

ATT GU11536902ood

146

5.25

1.166

ATT Positive

146

5.08

1.243

ATT Competent

146

5.10

5.72

1.363

ATT Important

146

1.213

1.655

5 IS REV I don't want any

more information about this

emergency situation.

145

5.70

22 INFOSEEK I plan to seek

out additional information on

ways I can better prepare for

this emergency situation.

146

4.88

1.659

1.703

24 INFOSEEK I will actively

seek more information about

this emergency situation.

146

4.71

Valid N (listwise)

139

Table 4. Continued

Prior to hypothesis testing, Cronbach’s alpha was used to assess the internal

consistency of the multiple-item indexes for problem recognition, constraint recognition,

level of involvement, information-seeking behavior, expertise, trust and attitude.

Reversed items were transformed before performing the reliability analysis. The results

of the analysis are shown in Table 5

. Table 5: Final Cronbach’s Alpha for Multiple-Item Indexes

Variable

Number of items

Problem Recognition

.65

Constraint Recognition

.66

Level of Involvement

.79

Information Seeking-

Behavior

.81

Expertise of the Source

.73

Trust of the Source

.90

Attitude Toward the Source

.87

Five items were included to test problem recognition; however the alpha indicated

scale reliability was higher by dropping the item “I need to seek out additional

information to better understand this emergency situation.” The four remaining items

produced a reliability coefficient of .65. Six items were used to test constraint

recognition. The alpha indicated scale reliability was higher by dropping the item “I

believe that there are constraints or obstacles that limit my ability to evacuate.” The five

remaining items produced a reliability coefficient of .66. Six items were also used to test

level of involvement. The alpha indicated scale reliability was higher by dropping the

item “I have strong opinions about this emergency situation.” The five remaining items

produced a reliability coefficient of .79. The three items used to test information-seeking

behavior produced a reliability coefficient of .81. The three items used to test the

expertise of the source produced a reliability coefficient of .73. The three items used to

measure trust of the source produced a reliability coefficient of .90. Finally, the four items

were used to test attitude of the source and produced a reliability coefficient of .87.

While alpha values between .80 and 1.00 indicated high reliability (Berman,

2002), it is generally agreed that the lower limit of .70 is still a useful measure of

constructs (Broom & Dozier, 1990; Stacks, 2002). While the situational theory of publics

is a strong theory, a strong criticism is the weak internal reliability of the items that

measure its constructs.

Hypothesis 1

Hypothesis 1 stated that in emergency communication, message source will

influence receiver variables. Receiver variables are problem recognition, constraint

recognition and level of involvement. To test this hypothesis, regression analysis was

performed. Message source, the independent variable, was regressed on the measures of

problem recognition, constraint recognition and level of involvement, the dependent

variables. The results of the analysis are shown below.

Proposition 1.1 stated that a national message source (NWS) will produce higher

problem recognition than a local source (HCG). The local government produced a higher

mean score (M = 5.63, SD = .94943) than the National Weather Service (M = 5.36, SD =

1.04410). However, the results of ANOVA were not significant, F(1,143)=2.629, p=1.07.

Therefore, P1.1 is not supported.

Proposition 1.2 stated that a national message source (NWS) will produce higher

constraint recognition than a local message source (HCG). While the ANOVA was not

significant, F(1,144)=3.142, p=.078, the National Weather Service produced a higher

mean score (M = 3.49, SD = 1.07795) than the local government (M = 3.19, SD =

.99157). Thus, there is mixed support for P1.2.

Proposition 1.3 stated that a national message source (NWS) will produce a higher

level of involvement than a local message source (HCG). The local government produced

a higher mean score (M = 4.94, SD = 1.04625) than the National Weather

Service (M = 4.65, SD = 1.10609); however, the ANOVA test was not significant,

F=(1,142)=2.620, p=.108. Thus, P1.3 is not supported.

Proposition 1.4 stated that a national message source (NWS) will produce higher

expertise than a local message source (HCG). The results of the ANOVA test were

significant, F(1,143) = 26.39, p =.000. The NWS produced a significantly higher mean

score (M = 5.23, SD = .95076) over the local government source (M = 4.36, SD =

1.07876). P1.4 was supported.

Proposition 1.5 stated that a national message source (NWS) will produce higher

trust than a local message source (HCG). The results were significant, F(1,142) = 44.79,

p = .000. The NWS produced a significantly higher mean score (M = 5.51, SD = 1.05697)

than the local government source (M = 4.21, SD = 1.27383). Thus, P1.5 was supported.

Proposition 1.6 stated that a national message source (NWS) will produce more

positive attitudes than a local message source (HCG). The ANOVA results were

significant, F(1,143) = 27.14, p = .000. The NWS produced a significantly higher mean

score (M = 5.69, SD = .88395) than the local government source (M = 4.86, SD =

1.05531). Thus, P1.6 was supported.

Proposition 1.7 stated that a national message source (NWS) will produce higher

information-seeking than a local message source (HCG). The local government source

produced a higher mean score (M = 5.28, SD = 1.43204) than the National Weather

Service source (M = 4.92, SD = 1.41293), and the ANOVA results were not significant,

F(1,143)=2.297, p=.132. Thus, P1.7 is not supported.

Hypothesis 2

A series of one-way ANOVAs were conducted to test this hypothesis. Hypothesis

2 stated that, in emergency communication, message strategy will influence receiver

variables. Results indicated that problem recognition was significantly affected by

strategy type, F(6,138)=3.00, p=.001. An evaluation of mean scores indicated that the

threat and punishment strategy produced the greatest influence on problem recognition

(M = 5.85, SD = .76691), followed by the persuasive strategy (M = 5.69, SD = .83256)

and the promise and reward strategy (M = 5.68, SD = 1.06646). The means for problem

recognition across all treatments are shown in Table 6.

In addition, results indicated that level of involvement was significantly affected

by strategy type, F(6,137)=2.821, p=.013. An evaluation of the mean score indicated that

the threat and punishment strategy produced the greatest influence on level of

involvement (M = 5.23, SD = 1.8875), followed by the facilitative strategy (M = 5.03, SD

= .78658), and the cooperative problem-solving strategy (M = 4.90, SD = 1.11252). The

means for level of involvement across all treatments are shown in Table 6.

The other receiver variables were not significant; however, the mean scores are

shown in Table 6. For constraint recognition, the facilitative strategy had the highest

mean (M = 2.96, SD = .83875), followed by the threat and punishment strategy (M =

3.17, SD = .94281). For expertise, the persuasive strategy produced the highest mean (M

= 5.32, SD = 1.25230), followed by the cooperative-problem solving strategy (M = 4.98,

SD = .89738). The persuasive strategy also had the highest mean for the trust receiver

variable (M = 5.24, SD = 1.25230). Second was the cooperative-problem solving strategy

type (M = 5.21, SD = 1.09279). For attitude, the persuasive strategy produced the highest

mean (M = 5.62, SD = .86826), followed by the cooperative-problem solving strategy (M

= 5.49, SD = .79668).

Hypothesis 2 stated that in emergency communication, message strategy will

influence receiver variables. Threat and punishment had the highest mean for

information-seeking behavior (M = 5.48, SD = 1.36572), followed by promise and reward

(M = 5.37, SD = 1.35465). Proposition 2.1 stated that information-seeking behavior will

be the highest when the threat and punishment strategy is used. Thus, proposition 2.1 is

supported.

Proposition 2.2 stated that information-seeking behavior will be the lowest when

the informative strategy is used. However, the persuasive strategy produced the lowest

mean for information-seeking among strategies, (M = 4.80, SD = 1.698). Thus,

proposition 2.2 is not supported. While the informative strategy did produce a low mean

for information-seeking behavior, the persuasive strategy produced the lowest mean

score.

Table 6: ANOVA Strategy Type/Receiver Variables

Strategy Type/Receiver Variables

Mean

Std. Deviation

PR Threat and Punishment

5.8500

.76691

Persuasive

5.6905

.83256

Promise and Reward

5.6875

1.06646

Cooperative

5.5595

.88000

Facilitative

5.5238

.74542

Informative

5.4896

1.11920

Control

4.5139

1.12613

Total

145

5.4897

1.00384

CR Facilitative

2.9683

.83934

Threat and Punishment

3.1667

.94281

Persuasive

3.2222

.96944

Promise and Reward

3.3000

.83875

Informative

3.4028

1.00111

Cooperative

3.5317

1.34125

Control

3.8070

1.21241

Total

146

3.3390

1.04210

LI Threat and Punishment

5.2333

1.08875

Facilitative

5.0333

.78658

Cooperative

4.9048

1.11252

Table 6. Continued

Promise and Reward

4.8333

1.21876

Persuasive

4.8016

1.06278

Informative

4.7319

.98311

Control

3.9649

1.00551

Total

144

4.7917

1.08309

EXPERT Persuasive

5.3175

.85943

Cooperative

4.9841

.89738

Promise and Reward

4.7667

.96791

Informative

4.7536

1.27215

Threat and Punishment

4.7167

1.14593

Facilitative

4.5238

1.04654

Control

4.4912

1.37153

Total

145

4.7977

1.10225

TRUST Persuasive

5.2381

1.25230

Cooperative

5.2063

1.09279

Threat and Punishment

4.9000

1.25237

Facilitative

4.8413

1.19545

Informative

4.7727

1.33090

Promise and Reward

4.6500

1.63469

Control

4.3684

1.54718

Total

144

4.8611

1.33770

ATT Persuasive

5.6190

.86826

Table 6. Continued

Cooperative

5.4875

.79668

Informative

5.4063

.96631

Facilitative

5.2857

1.17070

Promise and Reward

5.2375

1.02108

Threat and Punishment

5.1500

1.09545

Control

4.7750

1.33007

Total

146

5.2860

1.05495

INFOSEEK Threat and Punishment

5.4833

1.36572

Promise and Reward

5.3667

1.35465

Cooperative

5.2857

1.37148

Facilitative

5.1429

1.20909

Informative

5.1111

1.67004

Persuasive

4.8000

1.69761

Control

4.4211

1.11023

Total

145

5.0943

1.42871

Hypothesis 3

Hypothesis 3 stated that level of involvement will produce the strongest effect on

information-seeking behavior. Linear regression analysis was conducted to test this

hypothesis. For the analysis, problem recognition, constraint recognition, and level of

involvement were the independent variables and information-seeking behavior was the

dependent variable. Results indicated that the three independent variables, together

contributed to 32% of the variance in information-seeking, R = .572, R² = .327,

F(3,138)=22.34, p=.000. However, an examination of the coefficient analysis, shown in

Table 7, indicated that problem recognition produced the strongest effect on

informationseeking behavior. Therefore, H3 was not supported.

Hypothesis 4

In addition, Hypothesis 4 stated that constraint recognition will produce the

weakest effect on information-seeking behavior. The results shown in Table 7 indicate

that H4 is supported.

Table 7: Coefficientsa

Model

Unstandardized Coefficients

Standardized

Coefficients

Std. Error

Beta

Sig.

(Constant)

.193

.750

.258

.797

.393

.122

.308

3.224

.002

.485

.125

.351

3.885

.000

.121

.102

.089

1.179

.241

a. Dependent Variable: INFOSEEK

The next chapter provides a discussion of the results and draws conclusions about

the findings of this study. It also includes limitations of the study, future research, and

conclusions.

Chapter Six Discussion

The purpose of this study is to understand what message source and strategy type

is most effective in emergency management communications. Results indicated that a

national message source produced higher constraint recognition. The national message

source produced higher expertise, trust, and attitude then a local message source. The

threat and punishment strategy produced the highest level of information-seeking

behavior. Information-seeking behavior was the lowest when a persuasive strategy was

used. Constraint recognition produced the weakest effect on information-seeking

behavior.

This study examined Hazleton’s (1993) public relations strategies. Specifically,

message strategy effects were examined using J.E. Grunig’s (1997) situational theory of

publics to determine strategy influence on individuals’ problem recognition, constraint

recognition, level of involvement, information seeking behavior, expertise, trust and

attitude toward the source of emergency communication. Four hypotheses were tested.

Hypothesis 1: In emergency communication, message source will influence receiver

variables.

P1.1: A national message source (NWS) will produce higher problem recognition

than a local message source (HCG). This proposition was not supported. Respondents

indicated that they are most likely to recognize the existence of a problem if it is

communicated at the local level. This makes sense because local officials have a better

understanding of the local environment and the potential effects of an emergency

situation. The local source plays a significant factor in whether publics will be active or

passive. Local officials are a prominent factor in disseminating emergency information to

the public. They are better educated on the perceived risks and how to minimize the

effects of the emergency.

P1.2: A national message source (NWS) will produce higher constraint

recognition than a local message source (HCG). This proposition produced mixed

support. “Constraint recognition represents the extent to which individuals perceive

obstacles or barriers, in a situation that limit their freedom to plan their own behavior”

(Werder, 2005, p.226). This is important because it states that national messages that are

not localized for a public can lead to the public feeling constraints on what actions they

should take during an emergency situation. Local officials must communicate this

information for the public to feel that they can do something in the emergency situation

and not feel constrained.

P1.3: A national message source (NWS) will produce higher level of involvement

then a local message source (HCG). This proposition was not supported. This means that

publics are more likely to be more involved in the emergency situation if they message

source is from local officials. “An emotional connection of relevance increases the

likelihood of individuals attending to and comprehending messages” (Pavik 1988). The

public perceives themselves to be more involved in an emergency situation when the

information is communicated at a local level. Local communicators understand the

geospatial map of the area and can best make decisions to help the public from a risk.

Once the public understands the existence of an emergency situation, having a local

official conveying the message will make the public more active.

The local officials have a significant role in emergency communication. The local

source creates higher problem recognition, produces lower constraint recognition and a

higher level of involvement. These three items help shape the role local officials have

when communicating emergency messages. Not only must they be knowledgeable on the

local effects of the emergency, they must shape public opinion to get the public to act in a

certain manner that is most beneficial to the community. The next three propositions

yielded significant results on the attributes of the message source.

P1.4: A national message source (NWS) will produce higher expertise then a local

message source (HCG). This proposition was supported with significance. The national

message source yielded a higher level of expertise then the local source. In emergency

communication, especially with the hurricane situation used in our treatments, people

believe that there is a higher level of expertise at the national level. This makes sense

because the National Hurricane Center makes predictions about these storms using

sophisticated resources. Local officials often communicate that they are waiting for

updated hurricane information from the National Hurricane Center via their scheduled

advisories. This may have engrained the public to view the national message source as

having more expertise. The local officials must take that information and shape messages

to reach their publics. For local officials, it may be best to communicate the source of the

message as from a national source to establish the expertise of the message.

P1.5: A national message source (NWS) will produce higher trust than a local

message source (HCG). This proposition was supported with significance. This was an

interesting discovery about the trust of local officials in emergency situations. Results

indicated that the public has higher trust in a national message source then a local

message source. This may reflect back to the argument that the public is used to hearing

that they are waiting on information from the National Hurricane Center. Local officials

seem to communicate the trust in the national message source in terms of their

communication strategy. Communicators need to factor the trustworthiness of the

national message source into their messages to become more credible. When

communicating an emergency situation, having interviews and live updates from a

national source may be a resourceful way to build trust in local officials. This shows the

communication with the national source and having them share their expertise to make

the best local emergency decisions.

P1.6: A national message source (NWS) will produce more positive attitudes than

a local message source (HCG). This proposition was supported with significance. The

attitudes tested were bad : good, negative : positive, incompetent : competent, and not

important : important. The attitude ‘important’ produced the highest mean score.

‘Positive’ and ‘Competent’ yielded the lowest means. This finding confirms the expertise

and trust of the message source. The public views the national message source as being

more important, positive, competent, and good. The local message source must

communicate the message as being from a national message source to have the message

be perceived by the attitudes tested in this study.

P1.7: A national message source (NWS) will produce increases

informationseeking than a local message source (HCG). This proposition was not

supported. This means the public looks toward the local message source to provide

specific information on what actions they should take. The public needs to understand the

local risks involved with the emergency and how best to protect themselves. While the

national message source may produce higher expertise, trust, and attitude, the local

component is still an important factor in the public’s actions to seek out additional

information.

Hypothesis 2: In emergency communication, message strategy will influence receiver

variables.

P2.1: Information-seeking will be the highest when the threat and punishment

strategy is used. This proposition was supported. This correlates directly with my

introduction situation with Hurricane Ike. Residents reacted toward evacuation orders at

unprecedented levels when certain death language was used regarding the intense nature

of the hurricane. This indicates that when communicating emergency information, it may

be best to communicate the ‘worst possible outcome’ as the message. Fear tactics, since

2001, seem to have a greater effect on the public’s actions toward an emergency situation.

In addition, it makes the public want to seek out additional information on the emergency

situation because the severity of it is the message.

When communicating to the public about emergency situations, the threat and

punishment strategy generated the highest information-seeking behavior. Communicators

can’t use ‘certain death’ language every time a hurricane approaches the United States

and expect to get the same reaction each time. The punishment aspect of the message is

critical. By saying that emergency response personnel are not going to risk their own

lives to rescue you because you failed to listen to emergency evacuation messages sets a

precedent that if you disobey mandatory evacuation messages that you will be dealt to

deal with the consequences.

P2.2: Information-seeking will be the lowest when the informative strategy is

used. This proposition was not supported. Information-seeking was the lowest when the

persuasive strategy was used. This means that the public doesn’t want to be told how to

act during emergency situations. They want to make their own decisions on how best to

deal with the situation by using the information they have. This is an important finding

because it indicates the type of message that should not be used when communicating

emergency communication.

Hypothesis 3: Level of involvement will produce the strongest effect on information

seeking behavior.

This hypothesis was not supported; however the results were significant. This

means that level of involvement does play a significant role in information-seeking

behavior. The more a person is involved in the emergency situation means that they will

seek out additional information. It is important to target the public that has the greatest

potential of being affected by the emergency situation because it will allow them to seek

out additional information on how to respond. Level of involvement and problem

recognition are vital to information-seeking behavior. If communicators target these

publics with the correct message type, the desired response indicated by national and

local officials will be executed.

Hypothesis 4: Constraint recognition will produce the weakest effect on information

seeking behavior.

This hypothesis was supported. Constraint recognition produced the weakest

effect on information-seeking behavior. A public that feels constrained during an

emergency situation is most likely to become passive about the situation because they

feel that they cannot do anything to prevent it. These individuals see obstacles and

barriers in the emergency situation that prohibits them from seeking additional

information or taking action.

Limitations

The results are limited to the demographic surveyed. They are not generalizable to

a greater population. The source of the message may not make a difference in how risk

communication messages are received. The use of ‘emergency’ is also a limitation

because that definition varies among people. A hypothetical situation was used which

may not be reality. It is tough to argue how people in Tampa Bay would act to a direct hit

by a hurricane because the area hasn’t been hit by a hurricane directly since the early

1900s.

In addition, the population surveyed was college undergraduate students who

don’t own their own dwelling. There ‘life’ as a college student may not depict the typical

lives of those living in the Tampa Bay area that may deal with this type of emergency

situation. Traditionally, college-students are not known to be an information-seeking

public. Students were surveyed during the beginning of hurricane season. The season

doesn’t usually increase in activity until August and September. If this survey was

conducted during a period of higher activity, hurricane communication may have been

more recent in their minds.

In conclusion, this study discovered the type of message source and message type

that is best to use in emergency communication. This study extends the situational theory

of publics to emergency communication. Future research would need to experiment with

other types of emergency situations. Hurricanes are only a major concern for part of the

country. Extending this study to other emergency situations could further validate the

results of this study. The variables tested also provide communicators with structure on

how to communicate these messages to the public. Emergency communication is critical.

By understanding how a public responds to these messages is vital to getting the message

out to its intended publics.

Students also viewed