An Empirical Study of Media Framing Effects

研究的目的。我们的研究旨在通过框架效应假说来探讨大众媒体如何与公众舆论互动的问题。框架效应假说指出，新闻媒体对公共问题的描述方式会影响受众对同一问题的理解。基于传播学研究者A.Edelstein、Y.Ito和H.M.Kepplinger提出的问题情境方案，我们构建了一个框架模型，该模型由七个问题情境类别组成，适用于多种问题和情境。这项研究的主题是日本“陷入困境的经济”。2001年5月，我们对年龄从20岁到69岁的东京居民进行了这项按比例选择的调查，并检查了受访者在考虑陷入困境的经济时对每个PS框架的重要性程度。调查前一年，我们还对《朝日新闻》和《读卖新闻》两家主要全国性日报头版出现的经济新闻进行了内容分析。内容分析显示，制度崩溃框架是媒体描述日本经济问题最频繁的背景。另一方面，我们在调查中发现，受访者对媒体上的经济报道越关注，他们就越倾向于认为细分框架很重要，即使在控制了性别、年龄和政党支持之后也是如此。这些发现支持了媒体框架假说。我们还讨论了框架效应和属性议程设置效应之间的关系--这是框架的一个密切相关的概念--并提出了一些支持后一种假设的研究结果。

Purpose of the study. Our study aims to explore the question of how mass media interact with public opinions in terms of the framing effect hypothesis stating that the way the news media frame a public issue influences the way the audience members understand the same issue. Based on the problematic situation scheme advocated by communication researchers, A. Edelstein, Y. Ito, and H. M. Kepplinger, we constructed the frame model that is composed of seven problematic situation (PS) categories and is applicable across many issues and situations.Methods. The main theme of the study was the "troubled economy" in Japan. We conducted the survey of proportionally selected Tokyo residents with ages ranging from 20 to 69 in May 2001, and examined the extent to which the respondents regard each of the PS frames as important when they consider the troubled economy. We also ran the content analysis of economic news stories appearing in the front pages of two major national dailies, the Asahi and the Yomiuri, for one-year period prior to the survey.Results. The content analysis revealed that the institutional breakdown frame was the most frequently used context in which Japanese economic problems were depicted in the press. On the other hand, we found in the survey that the more attention the respondents pay to economic coverage in the press, the more they tend to consider the breakdown frame as important, even after controlling for gender, age, and party support. These findings lend support to the media framing hypothesis. We also discussed the relationship between framing effects and attribute agenda setting effects -- a closely related concept of framing -- and presented some findings in support of the latter hypothesis as well.