Abstract
This report details the steps involved in setting up a polling club as part of related classes in political communication, public policy, and civic engagement among college students. It also examines an extracurricular activity that provides students with the opportunity to assess public opinion on policy matters at the local, state, and national levels. Insights as well as challenges from professors and students involved in such pertinent themes as web analytics, aggregate polling, and the internal and external constraints and biases inherent in such a project will be explored, as well as the need to focus on an integrated strategic communication perspective that bridges the frequent silos of marketing, advertising, public relations, journalism, and political communication.
Keywords
Do you believe that people are truthful when answering public opinion polls?
According to an Arizona Republic poll, 81% of the respondents believe that people are truthful (“Survey Results,” 1994).
Do polls shape public policy? Are polls reflective of the overall population? There is no consensus of opinion on these salient questions, yet, the prevalent perception is that polls are used by opinion leaders to support policy positions, and this translates into polling literacy as an integral part of any collegiate advocacy program.
The power of the press in society is an ongoing salient theme in political communication. This gatekeeping function of the media was a theme of Walter Lippmann’s 1922 classic, Public Opinion, at a time when mass media was in its infancy. By 1963, Bernard Cohen embraced the power of the media in the deliberative process, when he wrote that the media “may not be successful much of the time in telling people what to think, but it is stunningly successful in telling its readers what to think about” (p. 13; Erbring, Goldenberg, & Miller, 1980). As McCombs and Shaw (1972) concluded in their classic 1968 study, the media is significantly influential in setting the agenda; the authors found a significant correlation between what the local and national media chose to cover in news reports and what the electorate later identified as the most important issues in the election. The challenge is how is the public engaged in the necessary deliberation on key issues when there is a dearth of reporting or coverage?
Teaching practical research polling skills through an engaging/interactive format creates a rewarding experience for all involved. A growing number of institutions have initiated this educational process and created their own primary survey research facilities. These institutions include Suffolk University, Quinnipiac University, Monmouth University, Elon University, Sacred Heart University, University of Connecticut, Winthrop University, University of Arkansas, Fairleigh Dickinson University, and University of Virginia, among others.
The purpose of this essay is to document benefits and outline challenges, efforts to ensure that students come away from the activity with valuable hands-on experience in the public polling process, as well as insights on how to be an effective spokesperson in analyzing and describing the results to a target audience as a prerequisite for students to be successful in civic engagement.
There are numerous benefits of such entrepreneurial approaches. A persuasive argument is that institutions of higher learning are ideal sites for such polling centers, to ensure that polling and public opinion research are undergirded with an educational and ethical base and helpful tools for advocates of civic engagement. One need only examine the web pages of local, national, and international media to note the value of such educational polling-based centers to see the importance of this topic.
For example, Monmouth University and George Washington University polls are cited as part of the website on Real Clear Politics. Princeton provides polling for the Boston Globe, and the University of New Hampshire’s polling is affiliated with the Boston Herald. Suffolk University initially was aligned with a local Boston NBC affiliate but has emerged as a public brand of the academic polling society, with interviews on all major news outlets. In 2008, Suffolk University was the only poll in the nation to accurately predict Hillary Clinton’s victory over Barack Obama in New Hampshire (Suffolk University, 2008).
Emerson College, which offered one of the first collaborative media/pollster classes in the United States, has experienced sporadic interest in polling beginning in 1997 and ending around 2006 with two published polls overseen by David Paleologos, the current Suffolk University pollster. The first was published in the New York Times in 1997 regarding the U.S. Senate race in New York (“Emerson College/Suffolk University,” 1998), and another Emerson poll was prominent on the Internet through numerous websites including the Daily Kos (TexasDemocrat06, 2005).
Polling as a part of the curriculum has been a part of several classes at Emerson for the past 3 years, under the tutelage of Spencer Kimball, a Republican political consultant. In the political communication capstone course, students learn first-hand how data are collected and analyzed through automated polling; this polling experience has been one of the highest rated aspects of the course, according to student evaluations. In addition to the student interest, there was also validity in the polling project. In spring semester 2012, the class polled Massachusetts “likely voters” about the Republican presidential primary and accurately predicted the outcome of the Super Tuesday primary.
The latent interest in polling in the curriculum and community surfaced again in fall 2012, with an additional unique component not found at other colleges or universities. At Emerson, the students were asked to write, conduct, and analyze a poll; to be in charge of disseminating the results to the media; and to be responsible for any media interviews and analysis of the poll results. Such involvement requires students to be fully engaged in all aspects of the communication process. It was agreed upon by the proponents that such activity would provide students the opportunity to be involved in the polling process throughout the campaign, which promised to be a challenging and decisive election between President Obama and Governor Romney.
Students were taught by Kimball, and they in turn peer taught each other on the art and science of measuring and gauging public opinion, along with understanding how meaningful data are packaged, communicated, and used for the public to deliberate on salient issues in public policy. The internal polls conducted by the students in the months of the presidential election allowed them to measure, compare, and contrast their results against the polls conducted throughout the campaign as well as the actual outcome of the presidential election.
Achieving these objectives required such activities as choosing and researching general areas and specific questions of policy, writing survey instruments, analyzing data, disseminating that information for public consumption, and subsequently measuring its effects within the public sphere. This hands-on approach, in learning the process and effect of polling through direct experience, created an innovative academic expectation and simultaneously provided faculty and students teachable moments in using real-life communication techniques.
Emerson College Polling Society Methodology and Procedures
The steps involved in such teachable moments include the following: Students choose an area of policy or a political campaign of interest to them. They then conduct research on the issues surrounding the topic(s) they have chosen, examining trends in current polls, debates on the issue, and the micro/macro context of the topic within society.
The first step is the investigative phase. Students write a survey instrument and are able to test different types of questions. For example, they can test the integrity of the sample by implementing the “bogus” candidate technique (adding a bogus candidate to the list asked about), which provides researchers a baseline on the reliability of the data being gathered about the actual candidates (Lavrakas & Merkle, 1990). The sample used for the poll is obtained from a registered voter list rather than the use of random digit dial. The data are then collected using an Interactive Voice Response (IVR) system.
There are distinct advantages and disadvantages to the IVR system. Traditional live survey research is very expensive, a reason that many schools do not consider political polling. Automated systems provide a low-cost alternative to live operators conducting data collection and have grown in popularity and validity, as methodologies employed by researchers such as Scott Rasmussen of Rasmussen Reports and Public Policy Polling advance this option’s presence and credibility in the marketplace. A major drawback to IVR data collection is that automated phone calls are not allowed to call a cell phone (FCC law); only live operators can call cell phones for political opinion surveys. The result is a 5- to 6-point conservative bias in regional areas and low responses among 18- to 29-year-olds (Keeter, Christian, & Dimock, 2010). Students employ a weighting technique of demographic breakdowns to offset the bias discussed above. Frequencies and cross-tabulations using IBM’s Statistical Package for the Social Sciences are used to run the analysis. (Additional details on procedures and methodology are available from the authors upon request.)
The final step in the process is to publish and track the group’s results, which may lead to student engagement. After the students interpret the frequencies and cross-tabs, they write an executive summary, followed by a press release. The students then choose a spokesperson to interpret the results of the poll for the media. Students research names and contact information of reporters who have covered the policy issue or the campaign and make direct contact with the reporters with the polling results, a step that begins to build professional relationships for students among the media. The group also publishes their polls using a third-party press release distribution vendor, and this media auditing is part of the analysis phase of the project.
Over the days and weeks following the release of the poll, students are assigned to monitor the web for any news stories that develop from the press release or other media interviews. Once such media stories are collected, students discuss how the polls were reported and compare notes and examples that provide evidence to support the claims that such coverage was biased or objective in its account of the poll. Therefore, critical thinking skills are developed and practiced in such analytical dialogues on the mediated realities on each poll.
As part of the dissemination of the data in the engagement with the public, students also become familiar with the use of computer-mediated communication including the online press releases mentioned and also the creation and maintenance of the groups’ website (www.emersoncollegepollingsociety.com). In addition, the online strategy includes students initiating both a Twitter account and a Facebook fan page. These social media sites were a product of the class activity near the end of the semester, due to the major focus of the group being devoted to learning the fundamentals of polling research and the dissemination of the information. Only after mastering the basics of polling could students focus on the means by which such information could be distributed and repeated on various diverse communication venues. Within this section of the program, polling students learn the difference between organic growth on SMS and artificial inflation of likes and followers through advertising and public relations. Students also use web analytics, as a part of the polling process, to study the response of campaign-specific ads posted to gauge effectiveness, as well as the number of comments and likes on each posting, which acquaint students firsthand with the dynamics of writing on the Internet.
Analysis of Program
The first proposed poll for the group focused on Massachusetts and the U.S. Senate race between Senator Scott Brown and Elizabeth Warren. The data were collected and ready for publication on August 20, 2012, with the results showing Brown holding a 6-point lead, the identical spread of a poll by Public Policy Polling (2012). Even with this congruity, the ongoing discussion of the propriety of a poll affiliated with the college, given the political affiliation of the professor, resulted in the poll not being released for public consumption. Yet, although internal only, the poll provided evidence to the students that they were mastering the process.
Several meetings with faculty and administration resulted in an agreement that the college would publish the next poll on September 9, which followed the two national conventions. Again, the Emerson students formulated the poll and collected and analyzed it for public release, but the college once again concluded against publishing the poll under Emerson’s moniker. This poll showed Warren narrowing Brown’s edge to just 1 point.
Within the context of ongoing discussions on the propriety of such polls with an Emerson affiliation, a dozen students attended the Polling Society meeting over the course of the next 2 months honed their skills through writing, analyzing, and interpreting results of presidential polls in Ohio, Virginia, Florida, Nevada, California, Massachusetts, Missouri, Arizona, and Colorado. Members of this group analyzed their results on Election Day to compare how accurate they were and what weighting methods they should employ.
Following the November election, the student group, with the approval of the administration, focused on conducting and publishing polls on policy issues. The Emerson College Polling Society (ECPS) published their first poll on November 29, 2012, regarding the handling of the “fiscal cliff,” and a second poll on December 11 on the treatment of soldiers with post-traumatic stress disorder (PTSD).
The dissemination process was the same for each of the polls: A student drafted and vetted a written press release published using E-release services. Any student who wanted to be quoted in the release was allowed to add his or her opinion, as long as it was based on the analysis. Overall, six students elected to be quoted, thereby offering students a practical experience in being a spokesperson.
The first poll on November 29 on the issue of the fiscal cliff was picked up only by a seemingly nonpartisan website, Tech Journal (“Most Americans Want,” 2012). The second poll focused on the return of soldiers from war and the causes and effects of PTSD. Although the student group saw this issue as a very newsworthy topic, particularly around the holidays, it was not picked up by any mainstream news outlets (Veterans Info Site, 2012).
The ECPS group published a third poll on December 20 about a hypothetical match-up between Governor Deval Patrick and Former U.S. Senator Scott Brown, assuming that Senator John Kerry would leave the Senate to become secretary of state. Coincidently, another poll regarding Kerry’s successor was published 4 hours later, with different results, by Mass Inc., a Boston-based research center. The ECPS survey showed Patrick over Brown by 5 points, and Mass Inc. had Brown over Patrick by 6 points, which provided the media with rich material to cherry pick both polls to write their stories and for students to see firsthand the agenda-setting function of the media through framing, all dependent on the respective political leanings/orientation.
For example, The Lowell Sun was the first to publish the results with the headline, “Poll: Patrick Would Beat Brown” (Camire, 2012), citing the Emerson poll. Townhall.com used the headline, “Poll: Scott Brown’s Return Seems Imminent” (Doherty, 2012), citing both polls. Other sites were more objective like Yahoo! News and The Atlantic Wire with the headline, “So Who’s Going to Replace John Kerry for Massachusetts Senate?” (Trotter, 2012).
Such diversity in presentation of the data and results offered students a salient case study of media bias and framing. This poll and the mediated reality offered by the media of the results gave students a rich opportunity to examine press dynamics. Specifically, after this poll and coverage, some students were intrigued to examine if each headline in such coverage fits the information in the story and what slant the news articles featured as relevant cues in preparing the students to write more effective press releases.
Furthermore, the above analysis provided students with specific examples of how media distort information, as well as the bias in what the media select to report from the data offered. For instance, most of the public agree that the policy that deters U.S. service veterans from seeking treatment for PTSD or how people would like to see the fiscal cliff issue addressed should be more important than a hypothetical U.S. Senate race. However, the media gave almost no coverage to the first two polls but almost universal coverage of and interest in the hypothetical “horse race” poll. This type of coverage only further supports the pervasive argument that journalists choose to focus on political horse races rather than on public policy issues.
During the spring semester, the ECPS published two polls on gun policy following the Newtown shootings. There was virtually no coverage of either poll and students used this opportunity of no coverage of a salient issue to consider and ponder the factors that operate within the 24-hour news cycle—what is newsworthy and why? It also prompted further consideration of the role of the media in setting the agenda for public dialogue and how social media could be employed to get the message out to a specific audience. The lack of coverage reinforced the notion that the public had lost interest in the issue.
The media dynamic that is evident regarding coverage of the polls in this study suggests that even with the advancement in new technology in flattening the media hierarchy and providing more people with access, we are still short of what Thomas Jefferson described as civic engagement based on debate and discussion of the “marketplace of ideas.” The evidence in this essay suggests that the media tend to cover and be interested in horse race stories in politics and are less inclined to cover important issues of the day.
The ECPS examined the current race for U.S. Senate in Massachusetts as their last poll for the semester. After some discussion and within the context of their newfound appreciation of the media’s penchant for breaking news, the students decided to be the first to poll the state after the Republican and Democratic primaries. The state primary was held on April 30, 2013, and the poll was conducted May 1 with the results published on May 2, making it the first poll to be published after the primary. Using the above-mentioned procedure of dissemination, the title of the press release was “Emerson College Poll: Markey Starts With 6 Point Advantage Over Gomez in U.S. Senate Race.”
The poll provided another teachable moment of how the news uses polling to legitimize an issue. The Weekly Standard published the first original news story of the poll results (“First Mass. Senate Poll: Single-Digit Race Between Markey, Gomez”) on May 2 at 5:32 p.m. (Warren, 2013). Their story referred to the paid press release that was published on the Wall Street Journal’s website (“Emerson College Poll,” 2013) using the legitimating source model (Jowett & O’Donnell, 1999). From this point, other conservative websites including The Daily Caller, Brietbart, and Red State published stories about the poll results. That evening, the local news outlet New England Cable News ran a lead story about the poll results (“New Poll Shows,” 2013). Once again, students were able to witness the symbiotic relationship between media and pollsters in the political process.
Almost immediately following the Weekly Standard publication of the poll, the Gomez (R) campaign and the Republican National Senatorial Committee published the poll on their websites. In recalling initial fears of a perception of bias considering that Kimball, the teacher, is also a Republican consultant, the question was, Would this question the credibility of the poll results? However, on the following day (May 3), another poll was released by Public Policy Polling, a Democratic polling organization, showing Markey holding a 4-point lead over his Republican rival Gomez. This added to the credibility of the Emerson poll, and almost immediately after the National Journal (Shepard, 2013) released their news story (“In Massachusetts, Kids Poll the Darndest Things”), the group was contacted by The Huffington Post for another story about the organization (“POLLSTER UPDATE—Markey, Gomez Running Close,” 2013). This created the needed credibility and assurance in the source of the poll for other established publications: Bloomberg News used the data for their publication titled, “Gomez Runs as Massachusetts Republican, Donated to Obama” (Linskey, 2013); The Hill published under the title, “Second Poll Shows Slim Markey Lead” (Sink, 2013); The Washington Post used the headline, “Can Lightning Strike Twice in Massachusetts?” (Rubin, 2013); Politico cited the data with their article, “Markey Leads Gomez by 6 Points” (2013); and a few days later, Fox News ran the story, “Polls Show Gomez Closing on Markey Days After Massachusetts Senate Primary” (2013). Currently, both Real Clear Politics (“Massachusetts Senate Special Election,” 2013) and The Huffington Post (“POLLSTER UPDATE,” 2013) have cited the student groups’ poll in their polls section.
Social Networking Sites
As part of the communication strategy of the ECPS, Facebook, Twitter, and Tumbler pages have been created. The latter two are being further developed, with more than 100 followers on Twitter. The Facebook page accrued more than 1,600 followers since January 2013, with the largest demographic being 13- to 17-year-olds. Emerson students were challenged with the task of not only increasing followers but increasing engagement through comments and “liking” of articles. One successful strategy employed by the students was to create polls for Facebook members to vote on, which, on several occasions, garnered hundreds of respondents.
Summary and Next Steps
From an educational perspective, the ECPS provides not only a laboratory for students to use the latest in technology to design, gather, analyze, and disseminate its polling data but also the opportunity to establish meaningful relationships with members of the media and leaders in the community, thereby providing invaluable connections for internships or, more important, job placement after graduation. As technology continues to evolve, the public marketplace of ideas continues to expand actually and virtually. Within this context, a fundamental requisite for students in the educational process is to learn the theoretical, practical, and analytical skills necessary to communicate, to influence, and to be active leaders in this marketplace.
The ECPS project’s primary objective is to provide students with hands-on experience in gauging public opinion. The future affords the opportunity for potential collaboration with other educational institutions and traditional as well as emerging media in the release of polls and focus group findings of relevance to targeted publics. Such synergistic efforts provide the pathways for common ground, thereby nurturing a relationship among the entities with which to share increased civic engagement as a desired outcome. The ECPS offers potential for involvement of students interested not only in electoral politics advocacy and policy making but also in marketing, public relations, sports, theater, and other disciplines assessing public opinion on any given topic. This provides great potential for cross-departmental projects.
From a pedagogical perspective, a major outcome of projects like the Polling Society at Emerson is creation of a dialogue among the student body on issues of the day and a subsequent forum to communicate relevant information on such ideas that advances a major shared objective across academia, business, and government—a greater appreciation of civic engagement.
Epilogue
The ECPS accurately predicted the difference in the U.S. Senate race between Ed Markey and Gabriel Gomez of 10 points (“Massachusetts Senate Special Election,” 2013).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.