著者
金平 茂紀
出版者
日本マス・コミュニケーション学会
雑誌
マス・コミュニケーション研究 (ISSN:13411306)
巻号頁・発行日
no.77, pp.77-104, 2010-07-31

In the US, public opinion polls have steadily become a part of ordinary life. There is a wide variety of implementation systems for opinion polls, and they cover a wide range of issues. Increasingly, we can see the tendency in the US that conductors of opinion polls have been polarized or politically factionalized in parallel with the polarization of media. There is also a great division in the public about how they evaluate opinion polls. We can see some opinion polls in the US actually mislead the public. The fundamental cause of this misleading is the process of converting complicated realities into numerical data. Furthermore, we can point out that methodological and technical factors like bias in the process of sampling and data handling, misrepresentations by the media, and so on. The crux of the problem in the media is that in many cases opinion polls are manipulated to reinforce certain narratives. More importantly, we should think of the practical context in which the opinion poll that is currently being dealt with, will eventually be used. In Japan, we can see clear tendencies of Japanese the media to put a high priority on the approval rating for the Cabinet as the decisive factor when measuring the success of administration. In fact, opinion polls simply represent a particular outcome, and should not become the causes of that outcome. Some people in Japan are pointing out that we are facing the risk of reification of abstract figures by believing opinion polls too blindly. To avoid believing opinion polls blindly, we should carefully consider the limitations of opinion polls and think critically about their effects.