여론조사보도에 대한 제3자효과 검증: 온라인 여론조사를 주목하며

Measuring the Third-Person Effects of Public Opinion Polls: Focusing On Online Polls

이 연구는 사람들이 전통적인 여론조사와 온라인 여론조사를 통해 얻어진 결과를 어떻게 다르게 인식하는가를 제3자효과론 (Third Person Effects)을 중심으로 살펴보았다. 미디어의 영향이 본인보다 다른 사람에게 더 클 것이라는 제 3자효과론은 그 동안 미디어효과 연구분야에서 폭넓게 적용되어 왔다. 하지만 최근 인터넷을 이용한 온라인 여론조사의 광범위한 사용으로 과연 이러한 여론조사에서, 특히 우리가 기존에 사용해왔던 전화를 통한 무작위 표본조사와 대표성이 약한 온라인 여론조사의 차이를 고려할 때, 실질적으로 미디어 이용자 입장에서는 얼마나 그 차이를 인식하는지는 매우 중요한 연구과제라 할 수 있다. 뿐만아니라, 서로 다른 두 가지 유형의 여론조사의 결과가 수용자에게 미치는 영향 측면에서 서로 다른 크기의 제 3자 효과가 나타나는지에 대한 관심도 커질 수밖에 없다. 이를 위해서 서베이와 실험연구 두 방법을 병용하여 이 문제에 대한 좀더 심층적인 분석을 시도하였다.

During the past decades, public opinion polls have become an ubiquitous tool for probing the complexity of people's beliefs and attitudes on a wide variety of issues. Especially since the late 1970s, the use of polls by news organizations has increased dramatically. Along with the proliferation of traditional polls, in the past few years pollsters and news organizations have come to recognize the advantages of online polls. Increasingly there has been more effort to take the pulse of the public through the Internet. With the Internet's rapid growth during the past years, advocates of online polling often emphasize the relative advantages over traditional polls. Researchers from Harris Black International Ltd., for example, argue that "Internet polling is less expensive and faster and offers higher response rates than telephone surveys." Moreover, since many of the newer online polls draw respondents from large databases of registered Internet users, results of online polls have become more balanced. A series of Harris Black online polls conducted during the 1998 gubernatorial and senatorial elections, for example, has accurately projected the winners in 21 of the 22 races it tracked. Many researchers, however, severely criticize online polls for not being representative of the larger population. Despite the often enormous number of participants, Internet users who participate in online polls tend to be younger, better educated and more affluent than the general population. As Traugott pointed out, the people polled in Internet surveys are a "self selected" group, and thus "have volunteered to be part of the test sample, which could mean they are more comfortable with technology, more informed about news and events ... than Americans who aren't online." The fact that users of online polls are self selected and demographically very different from Americans who have no access to the Internet is likely to influence the estimates of what the majority of people think about social or political issues. One of the goals of this study is therefore to analyze whether people perceive traditional and online public opinion polls differently. While most people might not differentiate sufficiently between traditional random sample polls and non representative online polls, some audiences might perceive online polls as more useful and representative. Since most online polls allow some form of direct participation, mostly in the form of an instant vote by mouse click, and often present their findings based on huge numbers of respondents, consumers of these polls might perceive them as more accurate, representative or reliable than traditional random sample polls. If that is true, perceptions of public opinion in society could be significantly distorted for those who rely on or participate in online polls. In addition to investigating how people perceive random sample and online polls, this study focuses on the perceived impact of public opinion polls. Similar to these past studies, which focused on how public opinion polls can influence the perception of mass opinion, this study will analyze how people perceive the effects of polls on themselves and other people. This interest springs from prior studies of the "third person effect," which have found that people often tend to perceive that persuasive communications exert a stronger influence on others than on themselves. While most studies concerned with the political effects of public opinion polls show that exit polls and early reporting of election returns have only weak or no effects on the outcome of election campaigns, some empirical findings suggest that exposure to polls can move people's opinions both toward and away from perceived majority opinion. Thus, if people indeed believe that polls influence others more than themselves, perceptions of majority opinion could be significantly altered because people might anticipate that others will react more strongly to poll results.