Exploring the Contours of Public Opinion: A Conversation with Vincent Price

Dr. Vincent Price is the Associate Provost at the University of Pennsylvania as well as the Steven H. Chaffee Professor of Communication and Political Science. He is widely published in the area of social influence processes, where his work spans a range of topics—from how mass media shapes public opinion to social identification and opinion formation. Dr. Price was editor-in-chief of Public Opinion Quarterly (1997-2001), and his book, Public Opinion, is available in five languages (Sage 1992). We conducted the following interview in October 2008.

What role do you see technology playing in survey research?

We're in a time of tremendous churn in standard practices. The multiplication of interactive technologies has given survey researchers more options than in past. In terms of a single technology that will serve the field, I don't think there's any clear vision into the future. Rather, I think that we're looking at alternative technologies that might be suited to different sorts of research tasks, and there's a lot of searching for new kinds. Because there is no one communication technology with complete penetration in U.S. population, it's complicated the task of drawing probability samples.

Regarding your research on the role of online political conversation and deliberation – what is the difference between the approach that you and Dr. Joseph Cappella are using, as compared to what Drs. Iyengar and Fishkin do via deliberative polling?

The impetus behind the deliberative poll was to create an alternative forum – a deliberative forum in which one could obtain higher quality opinion. The base proposition is essentially that simple surveys produce superficial responses. Therefore, if survey respondents are engaged in a much more elaborate procedure involving briefings – some educational sessions and some deliberative sessions – this process will produce more intelligent, if you will, public opinion. I see it as creating sort of a hydroponic garden, in which you can observe not what public opinion is, but what public opinion might be in this alternative setting.

What we do is somewhat related, but our aims have less to do with advancing deliberation as an alternative to standard surveys than they do with testing the basic propositions behind deliberative theory, which I think are largely assumed in much deliberative polling. We're interested in examining whether people can tolerate disagreement, and the ways in which they respond to counterargument and argument. We ask questions to gauge how differently participants react in various kinds of group settings. In each project, we're not so much interested in presenting what we find as an alternative to what a typical survey would capture. It is simply a more elaborate form of surveying, to plumb the contours and dimensions of public opinion, for a sense of where public opinion might go under certain circumstances. This provides a deeper sense of what the public thinks behind the kinds of questions asked. Groups sometimes generate knowledge that is greater than the sum of members’ individual understandings of an issue.

In 2000, you conducted a complex electronic dialogue project around the Presidential campaign. Could you please describe this effort?

This project involved two baseline surveys conducted with KnowledgePanel, which gathered information such as communication behaviors; interest in the campaign; and a variety of political attitudes and opinions. For a random subset of the panel, we followed the surveys with a series of moderated, hour-long, online text-based discussions. These took place on roughly a monthly basis leading up to the election, after which we ended with a post-election survey for all participants. This was a quasi-experimental study of the impact of these discussions on voters, as they followed the campaign, as well as an experiment with online discussion groups.

You and Dr. Cappella are currently using similar methodology for a project about how people think about genetics and health. What did you learn from the earlier project that you're applying to this one?

From a design standpoint, we have learned that it's quite feasible to manage a large number of these distributed discussions, using a highly standardized script to maintain comparability from one group to the next. We've seen advances in what the technology is able to do, and the ease with which people interact in group conversations. In the current project, our primary interest is in looking at ethical decision-making about genetic testing. We're designing a series of scenarios and asking groups to think through them and respond. This is a way to get a sense of how the public thinks about complex scientific and health related issues, at a time when the science and technology has far outstripped policy makers' ability to think through how these issues might play out for ordinary citizens. In each project, we come with a very experimental mindset and try to deploy these technologies in a way that advances our understanding of a particular aspect of public knowledge and opinion.

What was your interest in text-based discussions, and what did Knowledge Networks contribute?

I was quite interested in the text-based discussions for a variety of reasons. Participants were not meeting face-to-face, and research on group interactions suggests that people's social standing in groups affects their levels of participation. I was interested in the text-based interaction as a leveler, which would permit people freedom and openness to express their views, rather than their being channeled through complex social interactions based around, for example, race, gender and other characteristics. When you remove voice and physical appearance from the equation, the hope is that you'll have livelier, more candid interactions. This essentially was our experience with that project. We chose KnowledgePanel, because we were interested in an online platform that would allow us to hold these text-based discussions; it gave us the opportunity to do this with a high quality sample with the representative characteristics that one would like in a national project of this scope.

What role do you see Knowledge Networks as having played so far, in terms in survey research?

The availability of a representative panel that is Web enabled is a tremendous contribution that Knowledge Networks brought to the survey research community. As an early user of this technology, I was delighted with what we were able to do in the project. We've used that platform to deliver text and video materials in experimental settings. It's also proven useful for gathering open-ended responses from survey respondents. Interactive forums, in terms of small group discussions, have produced a very rich stream of data that is textual in nature, and can be content analyzed.

What did you learn about how discussion shapes public opinion?

We essentially were interested in testing propositions about what ordinary citizens are capable of doing – in particular, how they manage disagreement. A lot of democratic theory is predicated on the ability of people to be responsive to alternative points of view. We purposefully assembled groups as homogeneously liberal, conservative, or mixed across the political spectrum, thereby producing variance in exposure to disagreement. We found that in general, people responded quite positively. For example, there was no indication that people liked the heterogeneous groups any less than like-minded groups. They did find people who did not share their opinions to be slightly less knowledgeable, in terms of their perceptions. At the end of the day, we found evidence that suggested participants' "argument repertoire" did, in fact, expand. People became more conversant – with arguments both in favor and against their positions, regardless of whether they changed their views. The project was designed to test some basic propositions, and the data suggested a favorable portrait of what ordinary citizens are capable of, at least in the context of these text-based conversations.

Can you comment on how you feel consumers' or respondents' attitudes toward research have changed, and how they might have affected or be affecting research in general?

There's a lot that passes for research that isn't research at all, and in the context of a political year, there are attempts to use surveys as a guise for engaging in other kinds of activities, including essentially campaign activities. This undercuts a national public appreciation for research as a valuable social enterprise and makes it more difficult for researchers to gain respondent cooperation. There have been long-term trends suggesting that survey respondents are harder and harder to obtain; more persistent efforts need to be made.

On the other hand, we live in a loquacious society, so many people simply enjoy the topics of research. So, I don't know that it's fair to broadly characterize consumers' attitudes toward research. I think consumers have attitudes toward particular research enterprises – generally positive attitudes toward university research, and more suspicious attitudes about ventures perhaps seen as less noble.

The increase of activities masquerading as research has made it tougher for legitimate enterprises to convince consumers. Frankly, that's where a panel comes in handy; people can join the panel with a degree of confidence about the nature of their interactions with researchers. It does cut through some of that noise.

Is there anything that you feel survey companies could do, to promote a deeper level of engagement with surveys?

One of the things that survey researchers can do is assist the public in identifying strong, productive, useful enterprises from those efforts that are selling or campaigning in the guise of doing research. Professional associations and organizational involvement in those associations are powerful ways for survey researchers to advance their interests; they can engage in public education so that consumers are more savvy. I'm now involved in organizing AAPOR's next annual meeting; our conference theme is "public choices in changing times." We will be engaging directly these questions about changes in media technologies; the uncertainty of standard methodology going forward for the survey research community; and the substantive changes in public affairs that we've seen over the last 10-15 years, as a product of fairly radical changes in communication technologies. I hope that all of the readers of this newsletter think about joining us in Hollywood, Florida in May of 2009.

What are your thoughts about the debate and public discourse around the current election?

It's a fascinating election on multiple counts. Obviously, we have historic election opportunities represented on the tickets, in a variety of dimensions. On a substantive level, even without impending financial collapse and other developments in recent weeks, these election campaigns would shape up as some of the more interesting in many, many years. In terms of public discourse, the Internet adds a cacophony of voices – some of them not particularly helpful, because it creates a kind of circus sideshow that has diverted a lot of attention from both campaigns, and in some cases, from issues at stake. On the other hand, a circus sideshow draws attention, and the Internet is playing out as a very interesting point in election coverage, where traditional media are feeling their way. They no longer have audiences confident in their ability to tell a singular narrative about the election. For example, many national polls now find that perhaps 25% of the U.S. population would say that they can believe most or all of what they read in even high-end newspapers.

That sort of climate of distrust in conventional media, coupled with the availability of all sorts of material on the Web, creates a playground for a lot of mischief. Nevertheless, it is also an opportunity for people who formerly were not particularly engaged with campaigns to become so. It's too soon to tell whether in the end, these campaigns will be judged as high quality; I have some grave concerns about the current level of discourse. I think it has a lot to do with campaign behavior on the part of the parties, and less so with our communication structure.

Photo: ©, Rmarmion - Dreamstime.com