Theory based approaches to understanding

PLEASE SCROLL DOWN FOR ARTICLEOn: 28 July 2010 Access details: Access Details: [subscription number 918013331] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Regi

PLEASE SCROLL DOWN FOR ARTICLE

On: 28 July 2010

Access details: Access Details: [subscription number 918013331]

Publisher Taylor & Francis

Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,

37-41 Mortimer Street, London W1T 3JH, UK

Journal of Health Communication

Publication details, including instructions for authors and subscription information:

http://www.informaworld.com/smpp/title~content=t713666566

Theory-Based Approaches to Understanding Public Emergency Preparedness: Implications for Effective Health and Risk Communication

Hye-Jin Paeka; Karen Hilyardb; Vicki Freimuthc; J Kevin Barged; Michele Mindline

a Department of Advertising, Public Relations, & Retailing, College of Communications Arts & Sciences, Michigan State University, East Lansing, Michigan, USA b School of Advertising & Public Relations, College of Communication and Information, University of Tennessee, Knoxville, Tennessee, USA c Department of Speech Communication and the Grady School of Journalism and Mass

Communication, University of Georgia, Athens, Georgia, USA d Department of Communication, Texas A&M University, College Station, Texas, USA e Georgia Division of Public Health, Atlanta, Georgia, USA

Online publication date: 21 June 2010

To cite this Article Paek, Hye-Jin , Hilyard, Karen , Freimuth, Vicki , Barge, J Kevin and Mindlin, Michele(2010) 'Theory-Based Approaches to Understanding Public Emergency Preparedness: Implications for Effective Health and Risk

Communication', Journal of Health Communication, 15: 4, 428 — 444

To link to this Article: DOI: 10.1080/10810731003753083

URL: http://dx.doi.org/10.1080/10810731003753083

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes Any substantial or

systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or

distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly

or indirectly in connection with or arising out of the use of this material.

Theory-Based Approaches to Understanding Public Emergency Preparedness: Implications for Effective

Health and Risk Communication

HYE-JIN PAEK Department of Advertising, Public Relations, & Retailing, College of Communications Arts & Sciences, Michigan State University, East Lansing, Michigan, USA

KAREN HILYARD School of Advertising & Public Relations, College of Communication and Information, University of Tennessee, Knoxville,

Tennessee, USA

VICKI FREIMUTH Department of Speech Communication and the Grady School

of Journalism and Mass Communication, University of Georgia, Athens, Georgia, USA

J KEVIN BARGE Department of Communication, Texas A&M University, College Station, Texas, USA

MICHELE MINDLIN Georgia Division of Public Health, Atlanta, Georgia, USA

Recent natural and human-caused disasters have awakened public health officials to the importance of emergency preparedness Guided by health behavior and media effects theories, the analysis of a statewide survey in Georgia reveals that self-efficacy, subjective norm, and emergency news exposure are positively associa-ted with the respondents’ possession of emergency items and their stages of emerg-ency preparedness Practical implications suggest less focus on demographics as the sole predictor of emergency preparedness and more comprehensive measures of

This research is a part of emergency and disaster preparedness project, supported by the Georgia Department of Human Resources–Public Health Division, with funds from the U.S Department of Health and Human Services, Centers for Disease Control and Prevention An earlier version of this article was presented at the 2008 National Communication Annual Convention in San Diego, CA, November 21–24

Address correspondence to Hye-Jin Paek, Department of Advertising, Public Relations,

& Retailing, College of Communication Arts & Science, Michigan State University, 320 CAS, East Lansing, MI 48824, USA E-mail: paekh@msu.edu

DOI: 10.1080/10810731003753083

428

preparedness, including both a person’s cognitive stage of preparedness and checklists of emergency items on hand We highlight the utility of theory-based approaches for understanding and predicting public emergency preparedness as a way to enable more effective health and risk communication

Following a series of natural and human-caused disasters ranging from Hurricane Katrina to 9=11 to epidemic disease outbreaks such as West Nile virus and avian flu, the public’s level of emergency preparedness is capturing the attention of pub-lic health officials, risk communicators, and academicians The Federal Emergency Management Agency (2007) has allocated more than $16 billion for state and local government initiatives to prepare for terrorism and more than $7 billion to prepare the nation for pandemic flu (Morse, Garwin, & Olsiewski, 2006) and established new protocols such as the National Incident Management System (NIMS) to enhance government response to disaster Despite the intense interest in emergency planning and preparedness, relatively little is known about the determinants of the public’s level of emergency preparedness beyond demographic characteristics This study explores the extent to which three key predictors—efficacy, perceived norm, and attention to emergency-related news media—are related to levels of emergency preparedness Drawing on literature from social psychology, persuasion, public health, and mass communication, we highlight the theoretical rationale for select-ing these three predictors Our analysis of a statewide emergency preparedness sur-vey provides several practical implications for public health officials and risk communication practitioners when designing messages to promote emergency pre-paredness

Emergency Preparedness as a Critical Public Health and Safety Issue

Emergency preparedness refers to the capability of individuals, as well as public health and risk organizations and communities, ‘‘to prevent, protect against, quickly respond

to, and recover from emergencies’’ (Nelson, Lurie, Wasserman, & Zakowski, 2007,

p S9) The term ‘‘emergency’’ connotes the sense of a sudden and unexpected onset

of a disaster, but it also can be applied to other situations, which may not reach the level

of ‘‘disaster.’’ From a public health perspective, emergencies tend to be characterized by disruptions in essential services, such as utilities, transportation, and food supply; the potential damage or destruction of dwellings and businesses; and the need for evacu-ation or rescue Although many different checklists exist to help citizens prepare, the most common emergencies such as fires, floods, hurricanes, and earthquakes require

a similar set of supplies; human-caused emergencies such as terrorism and industrial accidents often build on the same lists Accordingly, we take this broader conception

of emergency and consider preparedness as a coordinated and continuous process of planning to take corrective action (Nelson et al., 2007)

The majority of research related to emergency management deals with response rather than preparation (McEntire & Myers, 2004) Studies of emergency prepared-ness behavior among the public primarily have been confined to descriptive statistics

of degrees of preparedness or to scrutinizing government actions, with little attention being paid to predictors other than demographic characteristics Also absent in the research is any examination of preparedness campaigns within the context of a con-tinuum or staged process—a critical oversight, given that being prepared for an

emergency is not an all-or-nothing proposition, but a process in which people may adopt recommended precautions incrementally

Existing literature reveals that Americans generally are not well prepared for emergencies, although there are some differences in preparedness levels among particular demographic groups (Eisenman et al., 2006; Enarson, 1998; Fothergill

& Peek, 2004) For example, in a February 2007 poll, the American Public Health Association (2007) found that 23% of the public had taken no steps to prepare for

an emergency; 27% claimed to be prepared, but only half that group had a 3-day supply of food, water, and medicine First aid kits and cash on hand also were lack-ing, but even low- or no-cost preparations were overlooked, such as planning how to communicate with family members in an emergency Other nationwide polls found similar results, with about 23–28% of respondents not prepared at all (e.g., American Red Cross and the Council for Excellence in Government, 2007; Harris Interactive, 2007)

Generally, it is the poor, the elderly, and households headed by women that are believed to be the least likely to be prepared for emergencies (American Red Cross, 2007; Morrow, 1999) Post-Hurricane Katrina studies have examined demographic differences in perceptions of community preparedness and the ability of the govern-ment to respond to disasters, but the same data do not exist with regard to individual

or household preparedness (Perry & Lindell, 2003; Sattler, Kaiser, & Hittner, 2000; Tierney, Lindell, & Perry, 2001)

It seems that personal vulnerability and resilience also contribute to individual variance in preparedness Other influences on preparedness, however, appear coun-terintuitive For example, awareness of risk is not associated with increased levels of preparedness A 2002 poll of more than 8,000 residents of Los Angeles County showed that despite high perceived risk in which 60% believed a terrorist attack was likely within the next year, only 37% had emergency supplies or plans As might

be expected, young people and single people were found to be the least prepared One unexpected finding was that the disabled, African Americans, and residents born outside the United States tended to be prepared better than other groups (Eisenman et al., 2006) This contrasts with conventional wisdom that these popula-tions are inherently more vulnerable to emergencies due to language barriers or lack

of resources, indicating a need for further research

Lessons Learned from the Past Studies: Toward Theory-Based Understanding and Prediction

Inherent in many of the typical preparedness surveys such as those detailed above are two measures of emergency preparedness: One is defined by a checklist of how many items have been gathered or plans made; and another describes the cognitive process behind preparedness behavior, including reflexive perceptions of prepared-ness There are both practical and theoretical reasons to discern two types of emergency preparedness Practically, the first measure (counting the number of emergency items) has been employed as an indicator of preparedness by many public health organizations (e.g., American Red Cross, King County Office of Emergency Management) This measure does not capture whether people own certain emerg-ency items coincidentally or whether they deliberately have collected these items to prepare for an emergency As a result, a measure of how mindful people are of plan-ning for emergencies may better describe, explain, and predict the public’s level of

emergency preparedness The utility of measuring the respondent’s self-appraisal of deliberate planning and preparation is also supported by the Transtheoretical Model (TTM) The TTM posits that people change their behavior through stages of action (Glanz, 1997; Prochaska & DiClemente, 1983; Prochaska, DiClemente, & Norcross, 1992) The number of stages of action has changed over time and may vary across health contexts, but most research incorporates some version of the following five stages: (1) precontemplation—the stage where people have no intention to change behavior in the near future; (2) contemplation—where people are aware that a prob-lem exists and are seriously thinking about overcoming it, but they have not yet made a commitment to take an action; (3) preparation—where people intend to take action and have started to make some changes; (4) action—where individuals modify their behavior or environment to overcome problems and reach certain goals; and (5) maintenance—in which people maintain behavior changes for at least 6 months or more (Prochaska & DiClemente, 1983)

The TTM suggests that communication messages should be different for people in different stages of change For example, action-oriented messages may be quite effective with individuals who are in the decision or action stages, but these same campaigns may be ineffective with individuals in precontemplation or contemplation stages These individuals may not yet be aware of a need to prepare, or they may not yet be persuaded of the urgency

or efficacy of preparation (Prochaska et al., 1992) Research has shown the importance of targeting each stage in the process to produce incremental change Even moving just one stage initially, for example, from precontempla-tion to contemplaprecontempla-tion or contemplaprecontempla-tion to preparaprecontempla-tion can double the chance of

an individual’s likelihood of getting to the action stage within 6 months or in the ‘‘near future’’ (Prochaska & Norcross, 2001) The model successfully has been applied to many health contexts including smoking and drug cessation, weight control, condom use, and mammography screening (Prochaska et al., 1994)

In this study, we apply the TTM in two ways First, we use the stages-of-change model to construct the dependent variable measuring the stage of emerg-ency preparedness actions This allows us to examine the degrees by which people consider, learn, prepare, maintain, and update emergency supplies and plans It makes sense to examine behavior as it is associated with these stages, because response to emergency preparedness campaigns often is incremental or graduated

in nature, a phenomenon similar to that documented in literature from many dis-ciplines (Marcus et al., 1992; Nigg et al., 1999; Prochaska & DiClemente, 1983) Second, we use the TTM to examine some of the key theoretical concepts or determinants that are related to both the number of emergency items that indivi-duals have and the individual’s stage of emergency preparedness Three important theoretical determinants are identified in the TTM that cut across several theories

of behavioral change (see Slater, 1999): (1) self-efficacy; (2) normative influence; and (3) media effects

Efficacy Hypothesis

In order to take action, a person must first recognize the existence of the problem and a need to improve the situation (also see Situational Theory in public relations literature; Grunig & Hunt, 1984) Second, the person must take into account any

constraints that might limit their control over the situation This idea is closely related to perceived self-efficacy, which has been defined in Social Cognitive Theory (SCT; Bandura, 1997) and others (Witte, 1992) as an individual’s belief that he or she

is capable of executing a particular behavior.1The Extended Parallel Process Model (EPPM; Witte, 1992) and Protection Motivation Theory (Rogers, 1983) also propose that people must have the confidence to undertake the recommended behaviors Finally, the Theory of Planned Behavior (TPB; Ajzen, 1985) also notes the signifi-cant role of perceived behavioral control in predicting behavioral intention Similar

to self-efficacy, perceived behavioral control is defined as an individual’s perceptions

of his or her ability or feelings of self-efficacy to perform a behavior Numerous studies have tested and confirmed that self-efficacy is a significant predictor of behavior or behavior change in various health and risk contexts (Glanz et al., 2002; Maddux, 1991)

Based on conceptual and empirical support from SCT, EPPM, and TPB, we offer the following hypothesis:

H1 Self-efficacy is positively related to emergency preparedness

Normative Influence Hypothesis Although self-efficacy serves as an internal driving force to make people adopt a recommended risk reduction behavior (like emergency preparedness), people also often follow what is acceptable in society and among their significant others The idea of normative influence is reflected most prominently in the Theory of Reasoned Action (TRA)=Theory of Planned Behavior (TPB; Ajzen, 1985; Fishbein & Ajzen, 1975) Both theories explain how behavior is influenced by the interplay among beliefs, attitudes, and behavioral intentions That is, behavior is dependent on intention, which is determined by an individual’s attitude (i.e., beliefs and values about the outcome of the behavior) and subjective norm (i.e., a belief about how one’s significant others think she or he should engage in the behavior) Although attitudes may best explain behavioral intention, subjective norm also has been found to explain behavioral intention to a significant extent (Gottlieb, Gingiss, & Weinstein, 1992)

While the subjective norm identified in TRA and TPB is limited in scope to a person’s ‘‘significant others’’ (e.g., parents, close friends, spouses, children), another line of normative influence literature embraces perceived norm in broader terms: an individual’s perception about what most people do or what

is typical or normal (also called descriptive norm by Cialdini, Reno, & Kallgren, 1990) Whether subjective norm or perceived norm, the idea behind this norma-tive influence is that people are influenced by what others think and how others behave (Paek & Gunther, 2007)

Guided by the normative influence literature, varying levels of emergency preparedness among individuals are likely tied to their perception of what their

1SCT proposes causal, triadic interactions among the environment (both social and physi-cal), people (and their intrinsic cognitive, affective, and biological traits), and the behavior (Bandura, 1997) In this theory, self-efficacy is critical to performing a particular behavior successfully (Glanz, Rimer, & Lewis, 2002)

significant others approve of or what a majority of Americans think and do Thus, this study proposes the following normative influence hypothesis:

H2 Subjective norm (H2–1) and perceived norm (H2–2) are positively related to emergency preparedness

Media Effects Hypothesis

People can learn from the news media about current disasters and emergencies and how to prepare for them First, one of the main functions of media, ‘‘surveillance,’’ is the watchdog function to keep people informed of what is happening in the environ-ment and aware of threats or dangers With media monitoring the world and gath-ering and summarizing important information, people can go about their daily lives, knowing they will be alerted when something relevant or dangerous needs their attention (Perse, 2001)

Second, through agenda setting and priming, the media play a key role not only

in making certain issues more salient (McCombs & Reynolds, 2002) but also in alter-ing criteria for evaluatalter-ing political leaders and government policies (Iyengar, 1991) Third, media expand the boundary of an individual’s experiences, allowing them

to vicariously experience various human-caused and natural emergencies such as Hurricane Katrina, 9=11, and anthrax (Mutz, 1998) The vicarious experience idea also is supported by the SCT, which posits that people observe and learn from role models and respond to an environment, in this case provided by media (Bandura, 1994)

Studies of disaster news have examined its functions in communicating warnings and directives to individuals directly affected, as well as the policy implications, edu-cational potential, and emotional impact of coverage on a wider audience (Garner & Huff, 1997; Keselman, Slaughter, & Patel, 2005) While emergency-related news may make disaster more salient to the public, coverage that emphasizes the random nat-ure of disaster may result in higher levels of fatalistic thinking (McClnat-ure, Sutton, & Sibley, 2007; Tierney et al., 2001) Therefore, the third hypothesis explores the relationship between attention to the emergency-related news and preparedness behavior:

H3 Exposure to emergency-related media news is positively related to emergency preparedness

Methods

Data

The study analyzed a state-representative telephone survey of 1,302 adult respon-dents age 18 and older conducted in the state of Georgia in July 2006 Georgia may be an ideal microcosm for the study of emergencies and disasters in general, because it experiences the same disasters that many other states do, yet there is no single overwhelming threat like earthquakes or hurricanes that might dominate resi-dents’ thoughts While many areas of Georgia are rural, Atlanta is the ninth largest metropolitan statistical area in the United States The city’s size and the presence of institutions such as the Centers for Disease Control and Prevention (CDC) and

Cable News Network (CNN) make it a possible terrorist target, and its importance

as a travel hub makes it a potential vector for disease

A professional research firm was employed to conduct the 15-minute telephone sur-vey, collected through a stratified, list-assisted, random digit-dialing (RDD) sampling method among the state’s general population.2 The CASRO response rate was 40.5%.3Compared with the 2006 census data, the sample appeared to slightly over-represent females (62%), Whites (69.6%), older citizens (M¼ 48.59, SD ¼ 15.53), and the better-educated (M¼ 5.06—some college education, SD ¼ 1.02)

Measures The variables used in the present analysis can be categorized into five groups, with one group serving as criterion variables and the four other groups serving as inde-pendent variables: (1) criterion variables; (a) the number of emergency items and (b) stage of emergency preparedness; (2) control variables: (a) demographics (gender, race, age, and the number of children at home), (b) social structural variables (social status, home ownership), (c) residential area, and (d) past experience; and (3) main predictors: (a) respondents’ own beliefs and perceptions—self-efficacy, subjective norm, and perceived norm regarding emergency preparedness; and (b) exposure to emergency-related media news Question items were drawn from existing literature and modified to fit the study context Appendix A reports the individual question items used to operationalize the variables Table 1 reports descriptive statistics for all the variables used in the study

The number of emergency items was measured by counting the number of actual emergency supplies people possessed The survey asked a series of ‘‘yes’’

or ‘‘no’’ questions about specific preparedness behaviors and supplies The question items were drawn from other emergency preparedness programs (e.g., King County Office of Emergency Management; American Red Cross) and con-firmed by public health officials Due to the limited time of a telephone interview, the Computer-Aided-Telephone-Interview (CATI) system randomly divided the sample into two groups (N¼ 654 and 648), each of which was asked either 10

2Using a multistage sampling method, all possible telephone numbers were divided into blocks that were defined by the last two digits of a 10-digit phone number Strict procedures were implemented to improve response rate and to ensure that the hard-to-reach respondents had a chance to be interviewed, including the following: (1) a requirement of distributing call attempts by each interviewer at different times across different days; and (2) strict call-back rules stipulating that a number could not be abandoned unless it had been verified as being invalid or attempted at least 25 times Within each household, the interviewer randomly selec-ted the adult in the household who most recently celebraselec-ted a birthday in order to randomize within-household selection and to prevent bias associated with only speaking to the person who answers the telephone The survey was composed of two sections: (a) one about emerg-ency preparedness in general; and (b) the other soliciting citizens’ knowledge and perceptions

of avian flu and pandemic flu The present analysis is based on the first section of the survey, where respondents were asked about their media use and attention related to emergency news, their perceptions and self-reported knowledge about emergency preparedness, various emerg-ency supplies that they currently owned, and their stages of emergemerg-ency preparedness

3The CASRO rate is a formula suggested by the Council of American Survey Research Organizations (CASRO) and is largely used in nationwide and statewide surveys (e.g., Beha-vioral Risk Factor Surveillance System) Similar to American Association for Public Opinion Research (AAPOR) response rate formula 3, the CASRO rate was calculated as the number of interviews divided by the number of known and unknown eligible cases

or 11 of the 21 questions Among these items, one item, ‘‘Extra supplies for pet,’’ was eliminated from the final analysis since it could not be applied to all participants, especially those without pets (21.5% of respondents) The ‘‘yes’’ responses then were summed to create a measure of the number of emergency items owned by individuals (range¼ 0–10)

To measure stage of emergency preparedness, the authors adopted the five stages of action developed and tested by Prochaska and his associates (1983, 1992,

1994, 1997, 2001) and modified them to fit the emergency preparedness context Those who answered ‘‘no’’ for the first question (‘‘Have you thought about planning for emergencies at all?’’) were considered to be in the pre contemplation stage; those who answered ‘‘yes’’ for the first question but answered ‘‘no’’ for the second question (‘‘Have you tried to learn or find more information about how to prepare for emerg-encies?’’) were placed in the contemplation stage; and so on, with the following questions corresponding similarly to the other three stages (i.e., two ‘‘yes’’ answers followed by ‘‘no’’ to ‘‘Have you taken any action to gather supplies or make a plan for emergencies?’’ was classified as preparation; ‘‘no’’ to ‘‘Have you updated your emergency plans or restocked your supplies for emergencies in the past 6 months?’’ was considered action and ‘‘yes’’ considered indicative of the maintenance stage.) The stage of emergency preparedness measure was a 5-point ordinal scale The correlation between the two dependent variables, number of emergency items, and stage of emergency preparedness was 57, indicating a moderate association

Table 1 Descriptive statistics (N¼ 1,302)

Dependent variables

Stage of emergency preparedness 2.37 1.45 Control variables

Beliefs & perceptions

Media use & attention

aGender and race were measured with dummy variables with female (male¼ 0) and White being 1 (other races¼ 0) Homeownership, urban residence, and past experience are measured with binary scale of yes or no The numeric values reported are percentages

Self-efficacy was measured with a single item, ‘‘How confident are you about your own ability to manage an emergency?’’ with a 4-point scale ranging from

‘‘not at all confident (1)’’ to ‘‘very much confident (4).’’

Subjective norm was measured with a single item, ‘‘To what extent do most of your family or friends think you personally should prepare for an emergency?’’ with

a 4-point scale ranging from ‘‘not at all (1)’’ to ‘‘very much (4).’’

Perceived norm was measured with a one-item question, ‘‘How well prepared do you think most people in the U.S are for an emergency?’’ with a 4-point scale ranging from ‘‘not at all prepared (1)’’ to ‘‘very much prepared (4).’’

To measure media use, participants were asked about the amount of attention they paid to emergency preparedness news using a 4-point scale (1¼‘‘no attention

at all,’’ to 4¼ ‘‘very close attention’’) The Exploratory Factor Analysis (EFA) result shows that the four items clearly constituted one factor explaining 59.8% of the total variance (Cronbach’s alpha¼ 76) The four items were averaged to create an index

of attention to emergency news, with a higher score indicating a higher level of attention

For control variables, guided by existing risk communication literature (e.g., Griffin & Dunwoody, 2000), the authors included gender, age, race (White as a dummy variable), number of children at home, social status, home ownership, resi-dential area (i.e., urban=rural), and past experiences with various emergency situa-tions (see Paek, Yoon, & Shah, 2005; Tierney et al., 2001, for the relevant rationales) For social status, an averaging index of education and income was created (6-point scale), because the two variables often represent individuals’ socioeconomic status and are highly correlated (inter-item correlation¼ 50; see Griffin & Dunwoody, 2000, for the same measure)

Analytic Strategy First, the variables were evaluated to see if they violated normality assumptions that might threaten the results Diagnostic statistics using skewness and kurtosis showed that the normality assumption was not violated

Second, two hierarchical multiple regression models were computed to assess the predictive value of all the variables on the two dependent variables, the number of emergency items and stage of emergency preparedness In the regression models, predictors were entered in the following order: control variables (demographics, social structural variables, and past experience), respondents’ own beliefs and perceptions, and then media attention variables (see Table 2) Although the stage of emergency preparedness is measured with an ordinal scale (that is not truly an interval scale), the diagnostic statistics indicate normal distributions without any outliers or truncation of the data Ordinary least squares (OLS) regression was used because it yields results similar to those from ordinal techniques such as probit or logit models but allows clearer interpretations (Fox, 1991; Knoke & Bohrnstedt, 1994)

Results H1 Efficacy Hypothesis Hypothesis 1 predicted that the more confident people are about their ability to pre-pare for an emergency, the more likely they will have emergency items and the higher