PUBLIC ADVISORY COMMUNICATIONS SYSTEMS

These class notes were composed by Dr. Tom O'Connor for his class on Homeland Security at NORTH CAROLINA WESLEYAN COLLEGE, original documents can be found here

PUBLIC ADVISORY COMMUNICATIONS SYSTEMS
"There is no limit to the good you can do if you don't care who gets the credit" (General George Marshall)

No warning system is 100% perfect, especially with natural disasters, which are quite unpredictable. However, in recent years, with technological, social, and terrorist hazards, efforts have been made to improve (and possibly perfect) emergency warning systems. This lecture attempts to outline and analyze the many issues associated with alarm and alert procedures and warning systems. We shall see that a number of principles in this area, which have survived more by custom than by science, may be in need of revision and modification. Disasters, of course, may never be eliminated solely by better prediction through science, and the political assessment of threat may never be completely congruent with scientific assessments. All that might be hoped for are better "information delivery systems" which provide sufficient early warning for the maximum number of people impacted and result in the maximum amount of mitigation possible, although predictability, controllability, and mobilization are worthy goals. Let's begin with some common definitions and/or distinctions that should be discussed.

The word "alert" is the more collective term for both alarms and warnings, and indicates when a predefined set of parameters have reached a warning or alarm state. The word "alarm" generally refers to a condition where human scientists or a mechanical signal device indicates at least one parameter has reached a critical or abnormal state. The word "warning" technically refers to action carried out by public authorities who translate alerts and alarms into recommendations for action. As Alexander (2002) points out, alerts and alarms are generally the responsibilities of scientific and technical staff, while warnings are generally the responsibility of political authorities. The relationship between alert, alarm, and warning is an interesting case of symbiosis between scientists, who pursue prediction in terms of relative ambiguous statistical probabilities, and politicians, who tend to pursue absolute certainties and crystal-clear warning messages. Prediction and warning are very different things.

Furthermore, alert and alarm procedures are different from warning systems, at least in terms of communications channels. Two communications systems are needed: (1) an alert and alarm system that connects all the monitoring, sensor, and signal devices (for which there are many for multiple hazards), and also allows the technicians and scientists to securely communicate with one another; and (2) a warning system for the general public and specialized groups, which needs to be connected, in some way, to the scientific system so that any misunderstandings about the meaning or interpretation of scientific findings can be cleared up. The phrase "specialized groups" also refers to that part of the public who is most likely to be impacted, and in this sense, may constitute a private, or third, system of communications. There are pros and cons about specialized communications systems in the field of disaster management, as sometimes it is more efficient to warn a sector of society, and sometimes there is a legal obligation to avoid "failure to warn" the rest of society. The channel which connects technicians and scientists together is sometimes called the "priority" communications channel; and the one which connects politicians is sometimes called the "civil authorities" channel.

The term "mobilization" refers to the need for a rapid response to the prediction of danger, and the concept of mobilization highlights the important necessity of time requirements. Survival rates fall exponentially if a warning is not given at least an hour before impact, and the preferable amount of time, established by custom with the scientific prediction of natural disasters, is eight hours. The concept of mobilization also highlights the way in which alert and alarm procedures (aimed mainly for emergency workers) are connected with warning systems (aimed mainly for the general public). Adapted from Alexander (2002), the following is the general model for a full-scale mobilization process:

According to Foster (1980), once a scientific prediction of impact has been made, an emergency manager should have the authority to decide, at any phase, whether or not to warn the public. However, there are many issues and dilemmas involved. One dilemma is that failure to warn at an early enough phase may be grounds for negligence lawsuits, and premature warnings may be perceived as false alarms and reduce public confidence in the warning process. Ideally, Section 9.1.1 of the emergency plan (see Emergency Management lecture) should outline the obligations of all parties on the interpretation of scientific evidence, but as a practical matter, the decision to warn the public must take other issues into account; e.g., how educated the public is; the sociological and psychological factors; the economic and political effects; and whether or not the warning system functions as expected thru prior testing.

Of the above issues, perhaps the public education issue is most important because the public needs to know exactly what to do and needs to believe that the warning is legitimate and authoritative. It is helpful to remind ourselves here that we are talking about sharing warning, not prediction, with the general public. Many in the public, especially the media, will want to know exactly what is expected to occur, and that is prediction, not warning. Warning tells the public what to do or what action to take, but the public is not generally confident in warning unless they are convinced the underlying prediction mechanisms are legitimate (have adequate foundations in law) and effective (valid and reliable). This informational dilemma is usually resolved by issuing a public advisory which contains a little of the prediction, but not enough to overwhelm the public with scientific gobblygook. For analysis purposes, the following is a sample public advisory warning, the form of which should be familiar to most people:

A Sample Public Advisory Type of Warning

The National Weather Service has issued a winter storm warning for the counties of ABC, MNO, and XYZ in which snowfall in depths up to six feet may unexpectedly hit eastern portions of those areas within the next eight to twelve hours accompanied by freezing rain and ice which may disrupt utilities and along with snow make travel impossible. According to both Doppler and Infrared radar, there is an eighty-five percent chance of this event happening with a ten percent confidence interval for error. Residents are urged to make immediate preparations for shelter and plan for at least two days until public services and utilities are restored. Local police will deploy roadblocks to curtail traffic, and emergency shelters will be available for those who feel they do not have adequate personal shelter for the possible two-day blackout. Do not delay. Make preparations now for staying inside, and remain tuned to emergency frequencies on a battery-operated radio.

In the above example, one can easily see the public education problem. The public may question the underlying legitimacy of the prediction mechanism; e.g., Doppler & Infrared radar with 85% plus or minus 10% accuracy. The public may also question the credibility of the warning source; e.g., the National Weather Service has been wrong before. There are also sociological and psychological factors that come into play, as many residents will rush out to grocery and supply stores to stock up on things like non-perishable food, water, and portable heating systems. How traffic is to be "curtailed" is questionable unless drills or prior tests have been done with police before. Panic is likely to result from some of the wording in the advisory. The public also needs to know where the emergency shelters are located. This is not a perfect sample, but there may be no perfect examples of perfect warnings. There may only be issues which must be addressed as carefully as possible.

There are three kinds of messages an emergency official ought to be willing and able to communicate to the general public: risk, warning, and crisis (Bullock et. al. 2005). A risk message can be an alert or have something important to do with public education. A warning message usually involves advising the public to evacuate or take shelter. A crisis message is about keeping the public informed of facts and developments during the response and recovery phase of emergency management. The choice of medium to communicate with the public, as we shall see, is an important matter, but there may be times when no means of communication is open, and that's where amateur (HAM) radio volunteers come into play because FEMA support for groups like RACES is strong. It is impossible to overstate the importance of having good, strong working relationships with whomever can get messages out to the public, whether it be traditional delivery systems -- television, radio, or print -- or whether it be one of the newer forms of media, like the Internet. Also, communicating with the public about terrorism hazards presents some unique challenges. Let's examine the three kinds of emergency messages.

Risk Communication:
This type of message essentially serves a public education, or persuasive function (Mullis 1998). The goals are to teach people what they can do to protect themselves and obtain their cooperation. There have been critics of this kind of communications, on grounds that it causes fear (e.g., Furedi 1997; Glassner 1999), or is just government propaganda of the Walter Lippman variety (see Lippman 1922) because it doesn't tell people everything. However, there is much research on risk communication indicating that it is the best way to reduce both the likelihood and consequence components of risk (Ansell & Wharton 1992; Willis 1997; Morgan et. al. 2002). Teaching people how to protect themselves is also an ongoing process called citizen preparedness, and the Ready.gov website is the primary DHS effort in this regard.

The research on risk communication tends to support some commonsense principles. First of all, the quality of information must be high. People will die if they don't get good, quality information. By "quality" is meant that the depth of information must be good enough that people can make intelligent, informed judgments. Secondly, it is known that people are more likely to follow safety suggestions if the risk involved is an old one they are familiar with, rather than a new risk they are unfamiliar with (Morgan et. al. 2002). Thirdly, it is important to avoid providing too much, exhaustive, unrelated information such as the kind you get from in-depth, investigative journalism. In fact, the journalism and broadcast industries tend to go into too much depth at times, and produce fear of risk rather than the appropriate response, awareness of danger. Risk, by itself, is not to be feared because it is simply a matter of odds, likelihoods, or probabilities; it's the danger or consequence of risk that should be feared, if fear must be involved at all. Trying to handle the new media's bloodthirsty taste for what is "newsworthy" without succumbing to needless, emotional fear appeal is a major challenge in emergency management, and there may also be times when it will be apparent that the news media seems "biased" against the purposes of risk communication. Singer & Endreny (1993) as well as Mileti (1999) have outlined some of the key ingredients for good risk communication, as follows:

Warning Communication:
This type of message is best delivered to the public via some easily recognized and popular system which represents a consensus on alert states. A color-coded scheme, a thermometer, sound tones, or other symbol of iconic stature are usually used. This kind of delivery system has traditionally worked well with natural disasters, but questions about utility and effectiveness have been raised for its applicability to terrorism (Ethiel 2002). The advantages and disadvantages of color coding schemes for terrorism are discussed later. Suffice it to say that groups like the Partnership for Public Warning have been at the forefront of examining the issue. One of the enduring controversies regarding warning communication is whether or not a consensual terror alert ought to be announced nationwide, regionally, or location-specific. There is some wisdom on all sides of this controversy. A nationwide warning certainly wastes precious resources by forcing governors and mayors all over the place to increase spending on security, but a location-specific warning might elevate fear and anxiety in one particular region (if done repeatedly) and play into the terrorist's hands.

Crisis Communication:
This type of message is best thought of as a consumer-friendly attempt to reach target audiences and/or stakeholders. In the public relations business, they are known as press briefings, and sometimes they have to be given out every hour or two, depending upon consumer demand. Their basic goal is to provide timely information during an ongoing crisis, and also to exercise care in not revealing any information that might "taint" an ongoing criminal investigation. The D.C. Sniper Case of 2002 is a good case study for this because Sheriff Moose, at times, seemed to be using the media to "talk" to the snipers, and there were numerous "leaks" and other panic reactions which might have been avoided. Standard hostage negotiation doctrine does not favor the idea of using the media to communicate with criminals, but there has been some precedent in serial killer profiling and counterterrorism, so maybe the idea is worthy of further exploration.

There are controversies over the best way to deliver crisis communication. Best practice has it that even if an oral press briefing is held, a printed copy of remarks should be handed to the press afterwards (or before, if the script can be followed). However, in some cases, the demand for information is going to be so urgent that frequent updates to an Internet web page might be the best delivery system. It is also important to remember one's audiences, and crisis communication typically is the kind of thing that first responders and other "internal customers" tune in to.

The study of panic reactions is an area where more interdisciplinary research is needed. In criminal justice, there is a diverse literature on panics, scares, and crime waves, but little of it is organized for applicability to homeland security. Sociological perspectives also exist of dubious value. Warr's (1987) line of research on the fear of crime tends to be somewhat on-target, and explores some of the psychological syndromes that are of interest to emergency managers. Slovic, Fischhoff & Lichtenstein (1981) have identified four "risk perception conclusions" that can be drawn from the research, which are paraphrased as follows:

Reaching out to the public must not only involve careful calculation of the time requirements, but the space requirements also. There are important spatial or geographic considerations in public safety, and these are generally addressed by a system of emergency zonation and micro-zonation. "Zonation" is a term from ecology which refers to an area of protection or reserve that has established boundaries and a buffer zone where transboundary activities can be coordinated. "Microzonation" is a term from the field of hazard management which attempts (thru extensive GPS mapping) to isolate the safer areas within a disaster impact zone. Emergency zonation usually involves the application of a color coding scheme for different places, and color coding is used as well for different levels of hazard. The use of color coding has long been regarded as the best way of imparting complex scientific information to an audience that is mainly interested in the overall pattern but may also be interested in the details (Bertoline et. al. 1997).

A basic, generic scheme for color coded zonation would make lighter colors, like white, the safest areas, followed by green zones, which are potentially dangerous but no special precautions need be taken. The color yellow is usually reserved for places where monitoring, surveillance, or intense awareness is needed. It then gets complicated after that because some theories and systems use red as the next level, indicating minor destruction possible; and other theories advocate blue or purple as the next level, whereas purple usually indicates major destruction possible. Regardless of whatever color coding scheme is used, it is important to use it consistently, clearly, make sure it is widely disseminated, and that the boundaries, once set, are firmly imposed. Alexander (2002) recommends the following color-coded alert model for general hazard warning:

*A Color Coded Alert Model for Hazard Warning*
White (or Green) level	Area subject to possible hazard, but none expected. Monitor for possible secondary hazards.
Yellow level	Precursor signs of hazard; at least one anomaly or enhanced activity detected; open priority channels.
Orange level	Repeated and accelerating signs of hazard; open civil authorities channel; issue bulletins or advisories; shut down hazardous and sensitive operations.
Red level	Intense signs of hazard where minor destruction may have already started; real-time monitoring and specific warnings issued; general warnings advice-oriented; evacuation plan and worker deployments in effect.
Purple level	Major impact of hazard in at least one area; evacuation should cease and quarantine begin; all emergency services working; forecasting of secondary impacts.

Warning, whether specific or general, should be a iterative, or repetitive process. The first message (most likely a bulletin) should be followed by further bulletins which update any changes in the status of the hazard as well as any additional forecasts or predictions of expected impact. Bulletins are usually disseminated via secure, or priority, channels, so there's little need to monitor the appropriateness of the response, unless one is conducting a drill, exercise, or fictitious scenario. Once an advisory has been issued, on the other hand, it is important to monitor the appropriateness of response to ensure that public panic does not result and that civil authorities are safe and secure. Advisories tend to become more and more advice-oriented up to and including the stages of evacuation and search and rescue. Bulletins are normally issued at fixed and regular times, like 6 minutes after every hour, and advisories are normally issued every 10 minutes. Every warning system should contain a way to measure feedback, which means that the recipients of warning should be able to confirm what they hear.

Warning systems must be delivered by some logical means of dissemination, and there are important logistical and sociological matters at stake. Logistically, one may need to consider any special arrangements necessary for night-time warnings, and sociologically, one may need to consider minority language groups, people with special needs, those who are deaf and blind, and those who live in remote areas. The general rule of thumb is to carry the warning on as many robust mediums as possible and that the overall system of dissemination have redundancy. A robust medium is one that is likely to stay in operation thru many levels of the emergency, and redundancy means that more than one medium is relied upon. No one channel of warning dissemination is likely to reach all the people who need to receive the warning message. Some media channels are notoriously unreliable under hazard conditions, and other channels are more vulnerable to hacking than others. Some basic choices, as outlined by Alexander (2002) with pros and cons, modified and extended upon, are as follows:

Usually, a public information officer is the specialist relied upon to choose the best medium for dissemination, but the procedures for public information are also usually spelled out in the emergency plan. It may be helpful to understand that whenever a warning is issued thru a news media that any particular media becomes "intermediaries" who may take it upon themselves to decipher, translate, and even challenge or criticize the warning. The tradition of a free press tends to foster something of a mistrust of civil authorities that often impedes the close public-private partnership needed for a successful warning system. The Partnership for Public Warning's (pdf) document on this matter is an educational read.

Since 1994, all radio and television stations have been mandated by the FCC (Federal Communications Commission) to have EAS (Emergency Alert System) equipment. EAS used to be called the EBS (Emergency Broadcast System) which had been around since 1963. FCC rules require a station to interrupt programming whenever there is a national alert, and stations are free to choose whether they want to run local alerts. The national alert warnings are sent as a digital packet to the stations and cannot be longer than two minutes in runtime. The station either must display the warning full-screen or add it as a "crawl" along the margins. Because many weather advisories and other warnings are local, and because producers want to minimize interruptions, the public only sees a small percentage of all the warnings that are issued via EAS. Check out the FCC's EAS page for recent rule changes, if any. EAS equipment is tested on all stations weekly by some sound tones, and monthly by a two-tone test message which goes like this: "This is a test of the Emergency Alert System -- this is only a test...." EAS automatically translates into Spanish for areas where there is a significant Spanish-speaking population.

The Common Alerting Protocol (CAP) is a movement that got started in the year 2000 and is based historically on a document called "Effective Disaster Warnings (pdf)" released by the Working Group on Natural Disaster Information Systems under guidance from the President's National Science & Technology Council. That document suggested a standard method be developed to relay instantaneously and automatically all types of hazard warnings locally, regionally, and nationally. The National Weather Service is the only agency with anything close to that capability at present, and since about 70% of EAS interruptions are weather-related, CAP initiatives have attempted to build on the NWS model, with FEMA guidance. CAP is also essentially a proposal for warning-system "interoperability" which is based on XML (Extensible Markup Language) standards. XML, like HTML (Hypertext Markup Language - what you're reading right now) can be understood by both humans and a variety of machines, especially those machines used to help the deaf and blind. A group called OASIS (Organization for the Advancement of Structured Information Standards) is involved in helping refine and showcase the latest CAP projects that software developers come up with. The most recent (November 2004) project showcased was a new way to handle alerts during incident management of a chemical plume cloud disaster [View Slides]. On a more significant level, the word "interoperability" frequently means "compatibility" and refers to the heavily grant-funded effort to get county-level emergency service departments on the same radio frequencies as city-level police departments. Most EMS departments use VHF frequencies which run in the 30 to 300 megahertz range, and most police departments use 800-megahertz frequencies.

In 2001, the National Center for Missing & Exploited Children launched the AMBER Alert System. The AMBER Plan was initially suggested in 1997 in honor of 9-year-old Amber Hagerman, a little girl who was kidnapped and murdered while riding her bicycle in Arlington, Texas. The way it works is that in serious child abduction cases where law enforcement agencies think time is of the essence and the law enforcement agency has met Justice Department requirements for entering data into NCIC (National Crime Information Center), descriptive information about the child is then sent to radio and television stations via the EAS (Emergency Alert System). Some states also incorporate electronic highway billboards as part of AMBER alerts, and these are those highway billboards ordinarily used to disseminate traffic information to drivers. Some AMBER alerts also contain pertinent information about abductor or what the abductor might look like. The system is operational in all the United States and most parts of Canada. There are penalties for false reporting, and to date, AMBER Alerts have been responsible for the successful recovery of 176 children. The Center for Missing Kids also runs a successful Cyber Tipline for crimes against children. CodeAmber is where one can get real-time tickers to add to one's website.

Using Amber Alert for Homeland Security Purposes

Amber Alerts are based on what is known as the capillary model of information distribution, where the alert starts out in a central conduit, like a main website or web portal, and then it spreads or flows out to a variety of devices, such as electronic roadway signs (pictured at right), E-mail, text messages, pop-up ads for people who opt in to alerts, and the capability exists to send the alert (and other digital signals) through public TV stations, cable companies, mobile phones, PDAs, satellite broadcasts, and other devices. NASCIO (National Association of State CIOs) and DHS are conducting pilots in early 2005 to see if Amber has the potential to be the nation's needed All Alert System.

Sigma Communications, Inc. developed the Reverse 911 System, a Microsoft Windows-based program, in 1993 to help public safety agencies contact citizens quickly in specific areas about emergency situations. Reverse 911 is a system that sends a recorded message to telephone owners from the police. The message can include a warning of the emergency and/or important information regarding the emergency. The calls can be targeted to specific neighborhoods or can be city-wide. The way it works is that residents receive a call that goes like this: "This is a Police Bulletin", DO NOT HANG-UP. The following is important information! This is a reverse 911 call from the police department, calling to let you know of important information. If you are an unlisted number, you must callback at 555-555-5555 to ensure that you will receive future reverse 911 calls." Reverse 911, especially when combined with Enhanced 911, which allows GPS tracking of calls and callbacks, allows for powerful law enforcement and public safety applications. While the system is finding widespread application, the system is most likely to be found in jurisdictions that are near prisons, and to notify households in case of an escape from the facility.

Slightly less than half the states and a good number of the nation's major cities have a 211 System, which was originally created in Atlanta during 1997 as a kind of non-emergency backup for overburdened 911 systems. The 211 systems generally only operate during normal business hours, and connect callers with agencies in an United Way database, other social service providers, charitable organizations, and emergency response information. Most 211 systems are localized or regionalized, but some state homeland security agencies (e.g., Connecticut, Hawaii, Idaho, Minnesota, and Arizona) in the last couple of years have been looking at utilizing the system as part of a statewide alert program. Connecticut, for example, used 211 in the aftermath of the Sept. 11 attacks to connect families with victims, refer people to mental health services, and get volunteers to donate blood. Some statewide 211 systems have VoIP (Voice-over IP) technology, and a group known as the Alliance of Information & Referral Systems has been in the forefront of enhancing 211 capabilities. Research to date indicates the system has excellent promise in directing people's generosity and may be useful in the recovery stages of emergency management.

There are also some interesting and innovative pure web-based alert-like systems on the Internet. For example, Rollins College in Florida (see Rollins emergency plan), created a system called the Student Storm Tracker, which allows students to check in and keep in touch from anywhere. Instead of tracking the storm, the tracker system was a web-based database to record where students went and a contact phone number where they could be reached. Every parent also received e-mails about the location and safety of their children in school. In addition, the system could send post-hurricane-status e-mails. During the 2004 hurricane season, Rollins got about an 85% usage rate with the system. It turned out to be exactly the kind of service people needed and wanted.

If and when signed into law, the Department of Homeland Security will be mandated by House Bill H.R. 2250 (pdf) to develop the READICall (Responsive Emergency Alert and Dissemination of Information Call System) emergency alert system which will both alert and inform citizens about imminent or current acts of terrorism. The READICall system is designed to be activated only by order of the Secretary of Homeland Security, and will reach roughly all the 104 million households out of 109 million that have landline telephone service, nearly every business in the United States, and 141 million cellular phones. In the meantime, as of late 2004, DHS is considering revising the Homeland Security Advisory System (HSAS), which most people are familiar with (below):

Homeland Security Advisory System

Severe Condition (Red) reflects a severe risk of terrorist attacks. Emergency managers increase personnel for emergency needs; mobilize specially trained teams; monitor or constrain transportation systems; and close public and government facilities.
High Condition (Orange) reflects a high risk of terrorist attacks. Agency-specific protective measures are implemented, and inter-government security efforts are coordinated; additional precautions at public events, including alternative venues or cancellations; critical workforce dispersed; and facility access to essential personnel only.
Elevated Condition (Yellow) reflects a significant risk of terrorist attacks. Increased surveillance of critical locations; emergency plans are coordinated with nearby jurisdictions; precise characteristics of threat are refined; and emergency response plans are implemented.
Guarded Condition (Blue) reflects a general risk of terrorist attacks. Communication systems and plans are checked; and the public is provided with any information that would strengthen its ability to act appropriately.
Low Condition (Green) reflects a low risk of terrorist attacks. Volunteers are trained; and all facilities and sectors are assessed for vulnerabilities to terrorist attacks, with reasonable measures taken to mitigate these vulnerabilities.

A threat condition is assigned on the basis of risk, where risk is a combination of the probability of attack and its potential gravity. A qualitative, not quantitative, assessment of integrated intelligence will be used to determine risk, with qualitative factors of such intelligence including being credible, corroborated, specific and/or imminent, and having grave consequences. There can be no guarantee that, at any given threat condition, a terrorist attack will not occur. Threat conditions are assigned by the Attorney General in consultation with the Assistant to the President for Homeland Security the appropriate Homeland Security Principals (heads of directorates). Threat conditions may be assigned for the entire Nation, or they may be set for a particular geographic area or industrial sector. They are binding on the executive branch and suggested, although voluntary, to other levels of government and the private sector. Federal agencies must submit an annual report describing the steps they have taken in terms of Protective Measures for each Threat Condition.

Most of the practical problems that the Advisory System (HSAS) has experienced center on the transition from yellow to orange. Reportedly, DHS officials receive so many complaints every time they change the threat level from yellow to orange, they’re simply not going to change it again, unless they can do so for very specific cities or industry sectors. Nobody really seems to know what to do at Condition Orange anyway, and Condition Yellow seems like what everyone is comfortable with. The CATO Institute has been a long-time critic of the HSAS, and the Partnership for Public Warning's pdf document outlines many problems with the HSAS, as does a 2004 GAO Report (pdf), and a Univ. of Delaware Disaster Center paper (doc) argues that HSAS is NOT even a warning system. Wikipedia's article on HSAS outlines the six times the Threat Condition has been raised to orange. The threat level has never been raised to red (high), nor has it ever been lowered to blue (guarded) or green (low).

PRINTED RESOURCES
Alexander, D. (2002). Principles of Emergency Planning and Management. NY: Oxford Univ. Press.
Altheide, D. (2002). Creating Fear: News and the Construction of Crisis. NY: Aldine de Gruyter.
Ansell, J. & Wharton, F. (1992). Risk Analysis, Assessment, and Management. NY: Wiley.
Bedient, P., Holder, A., Benavides, J. & Vieux, B. (2003). "Radar Based Flood Warning Systems." Journal of Hydrologic Engineering 8(6), 308-318.
Bertoline, G., Wiebe, E., Miller, C. & Mohler, J. (1997) Technical Graphics Communications. NY: McGraw Hill.
Bullock, J., Haddow, G., Coppola, D., Ergin, E., Westerman, L. & Yeletaysi, S. (2005). Introduction to Homeland Security. Boston: Elsevier.
Burkhart, F. (1991). Media, Emergency Warnings, and Citizen Response. Boulder, CO: Westview.
Donley, M. & Pollard, N. (2002). “Homeland Security: The Difference between a Vision and a Wish.” Public Administration Review 62:138-144.
Dory, A. (2004). "American Civil Security: The U.S. Public and Homeland Security." The Washington Quarterly 27(1): 37-52.
Ethiel, N. (Ed.) (2002). Terrorism: Informing the Public. Chicago: McCormick Tribune Foundation.
Fischoff, B., Gonzalez, R., Small, D. & Lerner, J. (2003). "Evaluating the Success of Terror Risk Communications." Biosecurity and Bioterrorism 1(4): 255-258.
Flynn, S. (2004). America the Vulnerable. NY: HarperCollins.
Foster, H. (1980). Disaster Planning: The Preservation of Life and Property. NY: Springer.
Glassner, B. (1999). The Culture of Fear. NY: Basic Books.
Haddow, G. & Bullock, J. (2003). Introduction to Emergency Management. Boston: Elsevier.
Lippman, W. (1922). Public Opinion. Cambridge, MA: Harvard Univ. Press [online edition].
Mileti, D. (1999). Disasters by Design. Washington DC: Joseph Henry Press.
Morgan, M., Fischoff, B., Bostrom, A. & Atman, C. (2002). Risk Communication. NY: Cambridge Univ. Press.
Mullis, J. (1998). "Persuasive Communication Issues in Disaster Management." Australian Journal of Emergency Management (Autumn): 51-58.
National Research Council. (1989). Improving Risk Communication. Washington DC: National Academy Press.
Partnership for Public Warning. (2003). A National Strategy for Integrated Public Warning Policy and Capability. McLean, Virginia: Partnership for Public Warning.
Perry, R. & Lindell, M. (2003). “Understanding Citizen Response to Disasters with Implications for Terrorism.” Journal of Contingencies and Crisis Management 11(2): 49-60.
Schneier, B. (2003). Beyond Fear. NY: Copernicus.
Schoch-Spana, M. (2002). "Educating, Informing, and Mobilizing the Public." Pp. 118-135 in B. Levy & V. Sidel (eds.) Terrorism and Public Health: A Balanced Approach to Strengthening Systems and Protecting People. NY: Oxford Univ. Press.
Singer, E. & Endreny, P. (1993). Reporting on Risk: How the Mass Media Portray Accidents, Diseases, Disasters, and Other Hazards. NY: Russell Sage Foundation.
Slovic, P., Fischhoff, B., & Lichtenstein, S. (1981). The Assessment and Perception of Risk. London: The Royal Society.
Subcommittee on Natural Disaster Reduction. (2000). Effective Disaster Warnings: Report by the Working Group on Natural Disaster Information Systems. Washington, D.C.: National Science and Technology Council.
Warr, M. (1987). "Fear of Victimization and Sensitivity to Risk." Journal of Quantitative Criminology (3):29-46.
Waugh, W. (1990). Terrorism and Emergency Management. NY: Marcel Dekker.
Waugh, W. (2000). Living with Hazards. NY: M.E. Sharpe.
Willis, J. (1997). Reporting on Risks: The Practice and Ethics of Health & Safety Communication. Westport, CT: Praeger.
Zschau, J. & Kuppers, A. (2002). Early Warning Systems for Natural Disaster Reduction. NY: Springer-Verlag.