Political Communication in Real Time

Theoretical and Applied Research Approaches

Dan Schill, Rita Kirk, Amy E. Jasperson, Dan Schill, Rita Kirk, Amy E. Jasperson

• Publisher: Routledge
• Year: 2016
• eBook ISBN: 9781317363040
• Topic: Politique et relations internationales
• Subtopic: Histoire et théorie politique

Much has been made of the speed and constancy of modern politics. Whether watching cable news, retweeting political posts, or receiving news alerts on our phones, political communication now happens continuously and in real time. Traditional research methods often do not capture this dynamic environment. Early studies that guided the study of political communication took place at a time when transistors and FM radio, television, and widely distributed films technologically changed the way people gained information and developed knowledge of the world around them. Now, the environment has transformed again through digital innovations.

This book provides one of the first systematic assessment of real-time methods used to study the new digital media environment. It features twelve chapters—authored by leading researchers in the field—using continuous or real time response methods to study political communication in various forms. Moreover, the authors explain how viewer attitudes can be measured over time, message effects can be pin-­pointed down to the second of impact, behaviors can be tracked and analyzed unobtrusively, and respondents can naturally respond on their smartphone, tablet, or even console gaming system. Leading practitioners in the field working for CNN, Microsoft, and Twitter show how the approach is being innovatively used in the field.

Political Communication in Real Time is a welcome addition to the growing field of interest in "big data" and continuous response research. This volume will appeal to scholars and practitioners in political science and communication studies wishing to gain new insights into the strengths and limitations of this approach. Political communication is a continuous process, so theories, applications, and cognitive models of such communication require continuous measures and methods.

PART I
Real Time Effects

Measuring Political Message Effects by the Second

1
THE HISTORY, RELIABILITY, VALIDITY, AND UTILITY OF REAL TIME RESPONSE

Dan Schill

In the world of media and politics, we are experiencing a time of tremendous change. The digital and social media ecosystem of 2016 is notably different from the Internet of 2012, which had important differences from the Internet of 2008. For instance, more than one in four registered voters now follow political news on their cell phones and the number of Americans who follow politicians on social media has doubled in the past five years (Smith, 2014). The use of second screens to dual-view television and social media has been widely used by audiences during political debates, live coverage, breaking news, and campaigns (Gil de Zúñiga, Garcia-Perdomo, & McGregor, 2015). Like previous elections, new digital and social platforms have been influential. In the 2016 primary, political advertisements began running on Snapchat, a four-year-old mobile application with 100 million daily users that lets users exchange messages that disappear after 24 hours or less (Viebeck, 2015).

Just as voters and campaigns are adjusting to this new environment, researchers too must adapt and study political communication as it is practiced. As George Washington University professor Dave Karpf (2012) asserted, “Standard practices within the social sciences are not well suited to such a rapidly changing medium” (p. 640). Particularly, two trends in this evolving environment are impacting research techniques: the increasing speed and constancy of political communication and the abundance of digital data and traces.

First, the current interpretive frame of politics is instant and constant. With continuous deadlines, 24-hour news, news and political organizations instantly pushing alerts to mobile devices, instant polls, and minute-by-minute citizen and journalist critique on Twitter and other social media outlets, the information flow never appears to cease. As Andrew Chadwick (2010) noted in his analysis of Britain’s first-ever live televised party leaders debate, these new political information cycles are not simply about an acceleration of pace, “they are characterised by more complex temporal structures, and the idea of a ‘24-hour news cycle’ does not quite capture their multiplicity” (p. 39). Our current use of a common language to describe this multiplicity is simply inadequate.

Second, because much of our political communication and our political behavior occurs over digital and social networks, there is now an abundance of accessible political message data and digital traces of behavior. Scott Keeter, Director of Survey Research for Pew Research, rightly stated, “At no time in history has so much of the public’s discussion been so accessible to a wide audience and available for systematic analysis. We are just at the very beginning in understanding what’s possible” (cited in Bialik, 2012, para. 3). And ubiquitous data collection is just getting started. According to a 2015 report from Intel Security’s McAfee Labs (2015):

Within the next five years, the volume and types of personal information gathered and stored will grow from a person’s name, address, phone number, email address, and some purchasing history to include frequently visited locations, “normal” behaviors, what we eat, watch, and listen to, our weight, blood pressure, prescriptions, sleeping habits, daily schedule, and exercise routine. Sensors will feed information to all sorts of organizations, returning ads, recommendations, and offers with real value. This combined information represents the digital exhaust that will become a mainstay and unavoidable by-product of modern life. (p. 18)

The increased popularity of Internet-connected devices, wearable devices, subscription music services, and smart televisions and cable boxes, combined with the continued popularity of digital communication applications and services will open new research areas.

The purpose of this chapter is to lay the conceptual, theoretical, and methodological groundwork for a series of tools best-termed Real Time Response (RTR), Continuous Response Measurement (CRM), or Moment-to-Moment (MtM) research. The chapter begins with a discussion of the importance and challenge of public opinion research, followed by a brief review of online and hybrid processing theories that undergird this method, and a quick history of the RTR approach and ways it has been used. The chapter concludes with a discussion of the reliability and validity of RTR. The chapter is designed to bring new readers up to speed and includes extensive references to guide additional study.

Public Opinion and the Challenges of Polling

By one count, over the past twenty years, twelve hundred polling organizations conducted nearly thirty-seven thousand polls by making more than three billion phone calls (Lepore, 2015). But why? As Rita Kirk argues in Chapter 3 of this book, media organizations and citizens alike promote polls because they are a quick and dramatic way to interpret a long political campaign. However, in a democracy, surveys also play an important normative role in asserting the public’s opinions on issues of the day. As early pollster Elmo Roper put it, the public-opinion survey is “the greatest contribution to democracy since the introduction of the secret ballot” (cited in Lepore, 2015, para. 16). John Krosnik and Arthur Lupia (2012) agreed when they wrote, “It is critical for a nation that cherishes its democratic legitimacy to seek credible measures of how citizens think, feel, and act in electoral contexts” (p. 9). Most observers agree that the survey is an important means of political participation and there is a close connection between citizen participation and survey research (Verba, 1996).

Because of polling’s prominence, coupled with often-low rates of citizen participation, surveys have become one symbolic embodiment of the American people. And, as W. Lance Bennett (2011) has described, this symbolism may be more of illusory echo of public involvement in a dramatized democracy than actual engagement as polling decisions are driven by implicit journalistic perceptions of power and the agendas of political institutions, questions may be worded to fit the most dramatic and simple news narratives, and interpretations are constrained by problems of sampling, response rates, and levels of participant knowledge and ability. Harvard history professor Jill Lepore (2015) has gone farther, contending, “Public-opinion polling isn’t enhancing political participation. Instead, it’s a form of disenfranchisement” (para. 41).

Interestingly, at the time when polls are wielding their greatest influence over American elections (including determining who got to participate in the 2016 primary debates), those same surveys are faced with many challenges. Although these difficulties are not new, changes in society have led to what Cliff Zukin (2015), past president of the American Associate for Public Opinion Research and Rutgers political science professor, called a “near crisis” of “the increasing unreliability of election and other polling” (p. 2). Keeter of Pew notes that the decrease in response rate is driven by several factors: “People are harder to contact for a survey now than in the past. That’s a consequence of busier lives and greater mobility, but also technology that makes it easier for people to ignore phone calls coming from unknown telephone numbers. The rising rate of outright refusals is likely driven by growing concerns about privacy and confidentiality, as well as perceptions that surveys are burdensome” (cited in Desilver & Keeter, 2015, para. 11).

By using rich data from carefully selected samples, RTR methods (such as dial focus groups, large scale analytics, online panel RTR studies, facial CRM, and real time digital text analysis) are ways to address these challenges of traditional phone polling. We need to supplement the existing, already strong approaches with new tools. As well said by Karpf (2012):

an expanding range of interesting and important questions cannot be answered using the most refined tools in our toolbox. The new media environment demands new techniques. Those techniques carry risks—they have not undergone the years of seasoning and sophistication that dominant methods have. But they also carry the promise of expanding the scope of our inquiry and applying intellectual rigor to topics of broad social significance. (p. 641)

Theoretical Grounding: Online and Hybrid Processing

RTR is one such new tool. In some ways, RTR is similar to traditional survey and experimental methods in that it asks participants to self-report opinions or tracks observable behaviors. The key distinctions of RTR is that it tracks these opinions, perceptions, and behaviors as they occur and over time. Not only does comparing how an audience processed a message in the first minute, to the second minute, to the third minute allow researchers to identify inflection points of influence, it is more analogous to how individuals actually process media and other messages. In other words, “Communication is a continuous process, [so] cognitive models of the communication process require continuous measures” (Biocca, David, & West, 1994, p. 15).

Work in political cognition and political information processing often differentiates between memory-based models, which suggest evaluations are formed at the time judgment is needed from the attitudes citizens can recall, and online models (e.g., Lodge, Steenbergen, & Brau, 1995), which argue that “citizens spontaneously extract the evaluative implications of political information as soon as they are exposed to it, integrate these implication into an ongoing summary counter or running tally, and then proceed to forget the nongist descriptive details” (Lavine, 2002, p. 227). The most recent work has found this dichotomy to be empirically unfounded and theoretically flawed as online processes rely on memory and vice versa (Taber, 2011). In these hybrid models (e.g., Hastie & Pennington, 1989; Kim & Garrett, 2011; Kim, Taber, & Lodge, 2009; Taber, 2003), online processing and memory processing work together and simultaneously include both recall and running tally impressions.

Real time research is based on these online and hybrid models. As further detailed later in the chapter, political information is incredibly complex and multidimensional, and many current methods rely on overly simplified assumptions of single-candidate, static evaluations. Kim and Garrett (2012) remind us that, “When individuals are making a vote choice, they do compare candidates as they learn about them in real time” (p. 364). Attitude or experience is often measured using a single retrospective rating and potentially introduces retrospective memory biases. Continuous measurement attempts to resolve these weaknesses by looking at many ratings and instantly recording attitudes as they occur.

History of Real Time Response

Real time response is not new and has long been associated with electronic media and political communication (Breazeale, 2000; Lazarsfeld & Stanton, 1944; Levy, 1982; Millard, 1992; Tringali, 2009). In the 1930s, George Gallup used a rudimentary form of continuous audience response to edit Gone with the Wind, and in the 1930s and 1940s, Paul Lazarsfeld and Frank Stanton developed their “Program Analyzer” to test CBS radio programming. From the 1960s through the 1980s, the Hollywood “Preview House” used real-time response to test commercials, films, and television shows.

The traditional RTR method was the in-person dial test focus group. Dial testing has been around since the 1940s to evaluate various media stimuli and is an increasingly popular methodology for academic researchers. Groups typically meet in focus group facilities and view programs, continuously rating what they see and hear second-by-second on a hand-held dial. Viewers turn the dial to the right (clockwise) to indicate favorability to what they see and hear and turn the dial to the left (counter-clockwise) when they dislike the content. A computer reads and averages the audience’s ratings and instantly plots the group reaction on a moving graph that rises and falls in real time as the audience responds favorably or unfavorably.

The Perception Analyzer, one of the leading dial test hardware systems, is actively used by approximately 250 organizations around the world and is used to test messages in the United States, Canada, Mexico, Brazil, United Kingdom, Germany, Romania, Turkey, Kazakhstan, Israel, Australia, Panama, New Zealand, Russia and China. Real-time response is most often used in applied contexts (Allen, 1996; Kolbert, 1992; Maullin & Quirk, 1995; Newsweek, 1996; Sargent & Lindgren, 1995; Schill & Kirk, 2014): companies use them to test the effectiveness of advertisements, television and film studios to evaluate the audience response to new programs and films, product development firms to test the marketability of new products, and political consultants to assess messages in ads, speeches and debates.

Today, real-time response plays a central role in the production and dissemination of nearly all major entertainment programs and advertisements (Tomashoff, 2012). Real time audience response systems are not limited to media testing, however. For example, PBS has used dial systems to maximize fundraising campaigns, keypad systems were used in Alaskan town hall meetings to allow residents to vote on how state oil revenue should be allocated, and legal communication experts have used RTR to refine arguments during mock trials (Breazeale, 2000).

Currently, there are a growing number of studies using various forms of RTR, including in-person dial groups, but also automatic facial analysis, and continuous sentiment analysis of digital trace data. Much of the previous research evaluates the effect of media messages on views. The largest group of RTR studies evaluates TV commercial effects (Aaker, Stayman, & Hagerty, 1986; Bailey, 2015; Baumgartner, Sujan, & Padgett, 1997; Bente, Aelker, & Furtjes, 2009; Boyd & Hughes, 1992; Evans & Hackenbracht, 2014; Madrigal & Bee, 2005, McDuff, El Kaliouby, Cohn, & Picard, 2015; McDuff, El Kaliouby, Senecha, Demirdjian, & Picard, 2014; Teixeria, Picard, & El Kaliouby, 2014; Thornton, 2005; Vanden Abeele & MacLachlan, 1994; Woltman Elpers, Mukherjee, & Hoyer, 2004; Woltman Elpers, Wedel, & Pieters, 2003). Other media effects studies have examined television shows (Baggaley, 1986; Gunter, 1995; Hernandez et al., 2013; Hui, Meyvis, Assael; 2014; Kodra, Senechal, McDuff, & El Kaliouby, 2013), public service announcements and health communication messages (Algie & Rossiter, 2010; Bollinger & Kreuter, 2012; Tedesco & Ivory, 2009; Wang, Vang, Lookadoo, Tchernev, & Cooper, 2014), websites (Altunyurt, 2009), and emotion inducing films (Fredrickson & Kahneman, 1993; Sze, Gyurak, Yuan, & Levenson, 2010; Tsai, Levenson, & Carstensen, 2000). Another group of media effects studies scrutinized the real time perception of music (Brittin, 1996; Geringer, Madsen, & Gregory, 2004; Lane, 2012; Schreiber & Kahneman, 2000; Schu...

Table of contents

• Cover Page
• Half Title Page
• Routledge Studies in Global Information, Politics and Society
• Title Page
• Copyright Page
• Contents
• Lists of Illustrations and Chapter Appendices
• Preface
• Series Editor's Foreword
• Acknowledgments
• About the Editors
• About the Contributors
• Part I Real Time Effects: Measuring Political Message Effects by the Second
• Part II Real Time Effects: Measuring Political Message Effects by the Second
• Part III Partisanship and Polarization: Real Time Selective Information Processing
• Index