Making Data Talk - National Cancer Institute

Transcription

National Cancer InstituteU.S. DEPARTMENTOF HEALTH ANDHUMAN SERVICESNational Institutesof HealthMaking Data Talk:A Workbook

Making Data Talk:A Workbook

How to use this WorkbookThis workbook provides an overview of the main points contained in the book Making Data Talk: CommunicatingPublic Health Data to the Public, Policy Makers, and the Press, as well as practical exercises for applying the book’sconcepts and communication principles to your unique situation.The first three chapters review basic communication concepts, from analyzing your audience to building a storyline.Chapters 4 and 5 shift the focus from conceptual to practical by introducing guidelines for presenting data, aswell as the Organize, Plan, Test, and Integrate (OPT-In) framework developed by the textbook’s authors to aid inplanning and executing data-related communications. Chapters 6 and 7 focus on the application of concepts andthe OPT-In framework to the real world in scenarios, such as crisis situations or advocacy.The ultimate goal of this workbook—and the book Making Data Talk: Communicating Public Health Data tothe Public, Policy Makers, and the Press—is to help you select and communicate quantitative data in ways layaudiences can understand. You will gain the most from this workbook by reviewing its contents in concert withthe book Making Data Talk: Communicating Public Health Data to the Public, Policy Makers, and the Press,making note of the tips and guidelines it presents, and completing the practical exercises beginning in Chapter 3to ensure your understanding of the concepts and ability to successfully apply them.

Table of ContentsIntroduction . 1Chapter 1: You CAN Make Data Talk and Be Understood . 2Table 1.1 Contrasts Between Scientists and Lay Audiences . 3Table 1.2 Tips for Presenting Audience-Friendly Data . 4Chapter 2: Use Communication Fundamentals to Your Advantage .Figure 2.1 Basic Communication Model .Table 2.1 Types of Sources .Table 2.2 Types of Channels .Table 2.3 Comparison of Selected Lay Audiences .55789Chapter 3: Help Lay Audiences Understand Your Data . 10Table 3.1 Audience Biases that Influence Quantitative Data Processing . 12Practice Exercise . 14Chapter 4: Present Data Effectively . 16Table 4.1 Basics of Visual Symbols . 19Practice Exercise . 21Chapter 5: Use the OPT-In Framework to Make Your Data Talk . 23Table 5.1 Roles of Data in Communication . 24Practice Exercise . 26Chapter 6: Show What You Know: Communicating Data in Acute Public Health Situations . 28Table 6.1 Acute Public Health Situations: Communication Phases and Objectives . 29Table 6.2 Questions Lay Audiences May Have in Acute Public Health Situations . 30Table 6.3 Higher-Controversy Situations: Characteristics and Communication Implications . 31Practice Exercise . 33Chapter 7: Show What You Know:Communicating Data in Health Policy or Program Advocacy Situations . 34Figure 7.1 Public Policy Cycle . 35Practice Exercise . 38Conclusion . 39References . 40

IntroductionCommunicating scientific data to lay audiences is difficult. Public health practitioners, researchers, clinicians, andothers in the public health field often have the responsibility of communicating “the numbers” to individuals fromall walks of life. How do you summarize and convey data so they make sense to someone who may not be familiarwith the topic, let alone the basics of epidemiology or statistics? How do you package and present data to answerthe question often asked by busy people with competing demands and time constraints: why should I care?The National Cancer Institute (NCI) is pleased to introduce Making Data Talk: A Workbook, which is based onthe groundbreaking book Making Data Talk: Communicating Public Health Data to the Public, Policy Makers,and the Press.1 This workbook is designed to be a companion piece that enhances the information presentedin the text by Drs. David E. Nelson, Bradford W. Hesse, and Robert T. Croyle, NCI researchers with significantexpertise in their own fields. The information presented in Making Data Talk: Communicating Public Health Datato the Public, Policy Makers, and the Press reflects a careful synthesis of research from many disciplines, so theprinciples described in the book can be applied to a variety of public health issues, not just cancer. This workbookcomplements the various communication and education tools and materials available through the NCI.The content presented in the following chapters will take you through communication concepts, an easy-to-understandframework for communicating data, and the application of that framework to actual public health situations. Manychapters also include practice exercises that use real-world examples to reinforce key concepts and help you applywhat you have learned. We hope this workbook will serve as a guide for those looking to successfully communicatescientific evidence to improve public health.Office of Communications and EducationNational Cancer Institute1

CHAPTER ONE:You CAN Make Data Talk and Be UnderstoodSharing information with the public is now one of the standard responsibilities of scientists and public healthpractitioners, such as epidemiologists, researchers, statisticians, health care providers, public relations officers,and others. Communication is a complex process that involves a series of choices about how to convey whatyou know or discover in a way that others can understand and, if applicable, use to make decisions abouttheir beliefs, attitudes, or behaviors.The Organize, Plan, Test, and Integrate (OPT-In) framework (presented and explained in Chapter 5)helps health communicators organize the communication process. OPT-In relies on a variety of basiccommunication concepts, including audience analysis. In Chapter 1, audiences are discussed in terms ofwhat they expect when receiving data and how those expectations can be used to craft more effectivecommunication. After reading this chapter, you will be able to: Identify some of the differences between health communicators and their audiences. Explain some basic strategies for making data more audience-friendly.You are likely to be successful if you use whatis known about your audiencesEffective communication starts with having a strong understanding of your audiences. It is important to notethat the people with whom you wish to communicate have their own areas of expertise, but those areas ofexpertise may fall outside of science or public health. The scientific community shares a common culture,so people outside of that culture may not share the same terminology, beliefs, or interests. See Table 1.1 formore detail on some common differences between scientists and lay audiences.2

Table 1.1 Contrasts Between Scientists and Lay AudiencesScientistsLay audiencesSources and definitionof acceptable evidenceNarrowBroadBelief in rationaldecision makingStrongVariableAcceptance of uncertaintyHighLowLevel of interestin scientific topicHighMedium to lowaQuantitative andscience literacyHighLowAbility and interest to reviewextensive amounts of dataHighLowaN ote: Except for audience members with high levels of involvement for a specific issue.Source: Making Data Talk: Communicating Public Health Data to the Public, Policy Makers, and the Press by David E. Nelson,Bradford W. Hesse, and Robert T. Croyle (2009), Table 1.2, p. 14. By permission of Oxford University Press, Inc. (www.oup.com).Each of the three lay audiences presented in the textbook – the general public, policy makers, and the press – isimportant to the practice of public health, and each has unique characteristics. See Chapter 2 for more informationabout these audiences and their characteristics.No matter the audience, people generally have certain expectations for receiving scientific data: They expect to be told why they should believe or do what scientists and other healthpractitioners recommend. They expect to be given the rationale for how these individuals reach their conclusions. Sincepeople are influenced by pre-existing beliefs and other factors, they may not be convinced to changetheir thinking without a sound and logical rationale for doing so. Finally, audiences expect to know what to do with the information they receive. In other words,they want to know what action they or others should take.In communicating with various audiences, you must acknowledge the role of your own values and ethics. Becausemany lay audiences inherently trust scientific experts, scientists and other communicators have an important ethicalresponsibility to maintain that trust. The selection and presentation of information can have a strong influenceon people and the way they interpret data. The goal is to lead people to conclusions based on sound data that arewell-reasoned and well-presented. To accomplish this, you should avoid emphasizing, minimizing, or ignoringcertain themes that would persuade someone to draw inaccurate conclusions from data.To succeed in effective communication, scientists and other health practitioners must consider these differencesand present data in a way that audiences will understand. Table 1.2 provides some basic tips for presenting datain an audience-friendly way. Chapter 4 of this workbook builds on these concepts by providing more practicaltips for presenting data to audiences.3

Table 1.2 Tips for Presenting Audience-Friendly DataTipExample/Explanation Avoid terms not frequently used outsideof the scientific community.Cohort, longitudinal Avoid terms with multiple meanings.Surveillance Avoid science and math concepts that canbe misunderstood. If these term(s) or conceptsmust be used, be sure to explain them in aneasy-to-understand way.Proportions, relative risk Focus on the main message instead of detailedscientific arguments or outcomes.When making decisions, many people use heuristics(shortcuts) rather than the rational decision-makingmodel used by most scientists.2 Explain how the data may impact audiences.Demonstrating impact can help audiences understandwhy the data are relevant to them. Present data in a distinctive way that helpsyou gain the attention of your audiences.For a majority of people in the United States, health issuesare of moderate-to-low interest.3 Presenting relevant andinteresting information can reduce the likelihood thatpeople will filter it out due to lack of interest.After reading this chapter, you should be able to recognize that effective communication with audiences outsideof the scientific community requires consideration of how those audiences differ from the scientific communityand how communication can be modified to account for those differences. For further detail on conceptspresented in this chapter, refer to Chapter 1, Introduction, of Making Data Talk: Communicating Public HealthData to the Public, Policy Makers, and the Press.4

CHAPTER TWO:Use Communication Fundamentalsto Your AdvantageAll efforts to share information—whether discussing a simple issue or a complex topic—consist of a few basiccommunication elements. By understanding these elements and how they work together, you can makeinformed choices about your communication approach. After reading this chapter, you will be able to: Identify and differentiate the four main elements of the basic communication model. Name three lay audiences key to public health communication. Recognize how messages can be developed to support a storyline.Consider the basicsA variety of elements are involved in the basic framework of communication. Although many more complexmodels of communication exist, this workbook uses the basic communication model presented in Figure 2.1as the foundation for discussion.Figure 2.1 Basic Communication ModelContextSource andchannelMessageAudience(receiver)ContextSource: Making Data Talk: Communicating Public Health Data to the Public, Policy Makers, and the Press by David E. Nelson, Bradford W. Hesse,and Robert T. Croyle (2009), Figure 2.1, p. 31. By permission of Oxford University Press, Inc. (www.oup.com). See References for additional sources.This basic communication model presents four main elements:1) Messages, or WHAT is used to convey information (e.g., words, symbols, or pictures).2) Sources (or senders), or WHO SENDS the message (e.g., individuals or organizations).3) Channels, or HOW messages are sent (e.g., newspapers, conversations, or e-mail).4) Audiences (or receivers), or WHO RECEIVES the message and interprets it.5

This workbook primarily focuses on helping people who work in public health (the senders) effectively communicatequantitative data as part of the health messages they send to the general public, policy makers, and the press(audiences) using various channels.In order to make the best decisions about the individual elements of the communication process (e.g., messages,channels, etc.), you should first consider the following: Purpose (i.e., why the message is being communicated). There are four purposes for communicatingpublic health information: to increase knowledge, to instruct, to facilitate informed decision-making,and to persuade. It is important to know which of these applies to the messages you are sending. Strategy (i.e., the approach for gaining attention). Some communicators use an active strategy,such as employing a mass media campaign or encouraging word-of-mouth communication. Othersuse a passive strategy, such as adding information to a Web site and relying on information-seekingaudiences to find it. The “push-pull” model combines both strategies by sending messages toaudiences (the push: active), while also making information and materials available to interested parties(the pull: passive). Context (i.e., factors that may influence receipt and/or interpretation of the message). Contextualfactors—including other sources of information, personal experience, and competing priorities—areoften outside the control of those sending messages. These factors can have influence at variouspoints during the communication process and can even prevent effective communication.Determining your purpose, planning a strategy, and considering the context are all crucial steps in the communicationprocess. In fact, these elements are three of the five fundamental pieces of the “Plan” step in the OPT-In frameworkthat will be presented in Chapter 5.MessagesMessages – and the storylines they support – play a critical role in both the basic communication modelpresented in this chapter and the OPT-In framework presented in Chapter 5.The term “storyline” must be defined and explained before messages can be developed and communicated toaudiences. In this case, the term “storyline” refers to the major conclusion(s) that scientists and other health practitionerswant audiences to understand. In other words, the storyline is the science-based bottom line. This differs according tothe type of information the story is based on.Once storylines are determined, messages must be developed. Messages – chunks of information that supportthe storyline – should be based on scientific knowledge and understanding. Each message should be able tostand alone by communicating a single idea, but, collectively, the messages should provide rationale for the largertheme (i.e., the storyline). “ Settled science,” or science that has received a clear consensus based on many studies over time, makesfor the strongest storylines since it provides a clear rationale. As a result, messages supporting settledscience storylines can be persuasive or instructive in nature. S cience that has little supporting knowledge and/or no consensus among scientific experts is moredifficult to address. Messages supporting these types of storylines should focus on increasing knowledgeor informing the decision-making process.6

These concepts are an important part of the OPT-In framework presented in Chapter 5, with storylines beingcrucial to the “Organize” step and message development being one of the five elements of the “Plan” step.SourcesAs noted in Table 2.1, sources are differentiated based on the intimacy of contact, with interpersonal sourcesinvolving one-on-one interaction and mediated sources involving one-to-many interactions. Communication ofteninvolves a mix of both interpersonal and mediated sources, such as when health information received from massmedia (e.g., a radio talk show host) becomes part of interpersonal communication (e.g., conversations with friends).Table 2.1 Types of riptionExamplePeople who share informationthrough one-on-one interactionFamily members, friends, colleagues,health care providersPeople who share informationthrough one-on-many interactionJournalists, politicians7

ChannelsLike sources, channels can also be divided into two main types: interpersonal and mediated (see Table 2.2).Table 2.2 Types of escriptionExampleWays of sharing information thatinvolve personal contactPhone conversations, oral presentations,personal e-mails, doctor visits, textmessages, social media/networkingWays of sharing information that aremore impersonal and typically reachlarger numbers of people at a timeNewspapers, newsletters, Web sites, TVChannel selection is a key component of message development and distribution. Research shows thatmany health campaigns have failed because only a small percentage of the intended audience was actuallyexposed to the message(s).4 To have a better chance of reaching the intended audience, scientists andhealth practitioners should consider the following factors: Availability, or whether audiences can access certain sources or channels (e.g., television, Internet,personal health care provider). Preference, or where and how audiences obtain information, which is closely related to availability. Credibility, or how believable a source is, based on perceived trustworthiness and expertise.Audience trends related to these factors change frequently, so you may want to consult the latest researchto understand the current habits and behaviors of your intended audiences.AudiencesThe following lay audience segments are important to public health communication: General public: individuals within the population at large. Policy makers: administrators and elected officials with the authority to make decisionsthat affect public health. Press: print, broadcast, or online journalists who obtain or report news.Table 2.3 provides descriptions and characteristics of each of the three lay audiences.8

Table 2.3 Comparison of Selected Lay AudiencesIndividual Occupational andinstitutional factorsVariable by audience subgroup,but common factors include: Level of interest in and involvementwith health issues Geographic location Varying levels of education Socioeconomic status Health insurance status E xisting health beliefs, social beliefs,and worldviews Gender Age Various social networks and cultures Ambitious, hard-working, savvy Attuned to financial implications Intuitive decision-making is common Want certainty from experts Usually have progressive “mainstream”values and beliefs Concerned about individualfreedom issues May be intimidated by scientists orhealth professionals General reporters, specialty reporters,and editorialistsRegular sourcesof informationVariable by audiencesubgroup, but trustedsources may include: Healthcare providers Television news Internet Web sites Other people (e.g.,friends, relatives,neighbors, co-workers) Radio/ethnic media Public vs. private systems Elected vs. appointed individuals Formal and informal processes Public policy typically made Interpersonal sourcesby legislators, executives, or A ttend to relevant newsadministratorsmedia coverage Interpersonal relationships crucial Rely on gatekeepers Busy and subject to multiplecommunication efforts and requests Business considerations: attunedto topics of interest to the public Short deadlines common Differences between specific newsmedia (e.g., newspapers, TV) Certain characteristics makestories more newsworthy”“(e.g.,local tie-in) Prefer personal stories (narratives) Much competition for news space Follow news outlet leaders”“(e.g.,elite papers such as The NewYork Times) Preselected listof trusted expertsSource: Making Data Talk: Communicating Public Health Data to the Public, Policy Makers, and the Press by David E. Nelson, Bradford W. Hesse,and Robert T. Croyle (2009), Table 2.2, p. 49 and Table 2.3, p. 54. By permission of Oxford University Press, Inc. (www.oup.com). See Referencesfor additional sources.Audience segmentation refers to the process of dividing an audience into smaller subgroups based on sharedcharacteristics (e.g., demographic information, geographic location, habits, and behaviors). Segmentation isa part of audience analysis—research that helps you better understand the people with whom you wish tocommunicate. Audience analysis can aid in planning your communication approach, thus, it is one of the fivefundamental pieces of the “Plan” step in the OPT-In framework.After reading this chapter, you should have a better understanding of the basic model of communication and itsfour elements: messages, sources, channels, and audiences. For further detail on concepts presented in this chapter,refer to Chapter 2, Communication Fundamentals, of Making Data Talk: Communicating Public Health Data to thePublic, Policy Makers, and the Press.9

CHAPTER THREE:Help Lay Audiences Understand Your DataWhen people receive messages, they process and interpret them based on their own literacy level, tendencies,and biases. As a result, these factors must be considered and addressed when communicating quantitative datato audiences. After reading this chapter, you will be able to: Identify audience tendencies that can influence how people receive data. Describe biases that audiences can have when interpreting data. Recognize techniques to overcome these tendencies and biases.Be aware of audience tendenciesPeople are not always well-prepared to receive and process messages containing quantitative data. Quantitativeliteracy (i.e., the skills required to apply mathematical operations) varies from person to person, and even themost educated audiences may have only a basic or intermediate level of familiarity with mathematical concepts.Common mistakes people make when interpreting numbers include:Misunderstanding probability estimates5 (people may believe that a risk of 1 in 200 is greater thana risk of 1 in 25).Misunderstanding percentages.Improperly converting proportions to percentages.6 To account for differences in quantitative literacy, health communicators should simplify messages, provideadditional explanation, or modify their approach to increase audience understanding.In addition to literacy considerations, health communicators should also be aware of general informationprocessing factors that, although not specific to data or public health topics, can be strongly influential aspeople process quantitative data. Here is a list of these tendencies along with explanations and examples.Cognitive processing limits. Individuals have a limited capacity to process large amounts of information at onetime and simplify or “chunk” the information to which they are exposed. T he 7-digit telephone numbering system was based on research suggesting that people can optimally retainonly 7 ( 2) discrete pieces of information at a time.7Satisficing. People tend to limit the amount of mental energy they spend obtaining information until they believethey have “enough” for their purposes.8 S tudies show that visitors will usually leave a Web site within 15 minutes or less if they do not find theinformation they need.910

Expectations of experts and the challenge of uncertainty. Most lay audiences want experts with experienceand credentials to provide definitive, prescriptive information.10 T o use a non-health example, people look to mechanics to definitively diagnose automobile problems —instead of estimating that there is a 30 percent chance that the alternator is the problem — as well as torecommend specific solutions.Processing risk information. Many people misunderstand concepts related to risk, such as absolute risk, lifetime risk,and cumulative risk.11 M ost people do not recognize that repetition of low-risk behavior — such as failing to wear a seat belt withevery car ride — increases a person’s cumulative risk of adverse outcomes during their lifetime.Framing. “Framing” is presenting data in a way that is consistent with common public frames or models. E mphasizing the possibility of colon cancer over the minor discomforts of a colonoscopy is an exampleof a loss frame. A ssociating rewards, such as losing weight and looking fit with exercise, is an example of a gain frame.Scanning. People often do a quick scan of written or visual material to decide if it interests them, draw conclusionsabout what the major points might be, and try to identify the bottom line.12 W hen an Internet search for specific information returns hundreds or thousands of potential Web sites,people scan the first few results before deciding which link to follow.Use of contextual cues. People tend to look for cues to help them better process and understand information,especially in cases where the data presented is complex, detailed, or in an unfamiliar format.13 R egular reports on breast cancer data can be of more use to audiences by highlighting what haschanged since the last report.Resistance to persuasion. People have a natural resistance to persuasion and often engage in a practice ofdefensive processing, an approach that blunts messages that are inconsistent with current behavior. S mokers may blunt messages emphasizing that smoking is bad since those messages are inconsistentwith the smoker’s own attitude toward tobacco use.Role of emotion. Emotions have the potential to be a motivating influence on behavior by heightening arousal,orienting attention, and prompting self-reflection.14 C ommunicating that 440,000 Americans will die from smoking in a given year may cause a variety ofemotional reactions based on the reader’s own relationship or attitude towards smoking.11

Be aware of audience biasesThere are also biases people have when interpreting data, particularly if they are not well-trained in statisticalmethods. For instance, people can process incoming information by using heuristic shortcuts, or highly ingrained,subconscious patterns that run automatically. These shortcuts can lead to systematic error15 and illogicalreasoning16 and are summarized in Table 3.1.Table 3.1 Audience Biases that Influence Quantitative Data ivenessheuristic P eople can sometimes use their implicit knowledge and stereotypes aboutan object’s category to make judgments about the object itself.—P eople perceive cancer to be a highly aggressive, lethal disease. As a result, it isdifficult to communicate that cancer is a broad set of diseases, that many typesare slow-growing and easily detectable, and that early diagnosis may not be a“death sentence.”Anchoring andadjustment bias P eople tend to be “anchored” by the first number they see or have in mind; anyadjustments they make are strongly in

the Public, Policy Makers, and the Press —is to help you select and communicate quantitative data in ways lay audiences can understand. You will gain the most from this workbook by reviewing its contents in concert with the book . Making Data Talk: Communicating Public Health Data to t