This paper quantitatively examines the gap between content identity declared by the creator of a webcomic and content image formed by consumers. We clarify the relationship between the factors that cause the gap and the consumers' evaluation of content. For empirical analysis, data from a largescale survey of a webcomic site was used. The identity and image of webcomics were estimated using categorization theory. The results show that basically, the more similar the content identity and image are, the stronger the preferences are. However, the hypothesis assuming that the highest preference will be obtained when the content identity and image are moderately incongruent is not supported.

While so-called fake news has become a popular domain of research around the globe since the 2016 U.S. presidential election, few studies have been conducted in Japan. This paper focuses on “middle media” outlets, which play an important role in the Japanese online news ecosystem, to analyze the generation process of fake news. Results show that the middle media outlets contributed to producing fake news by combining news stories published by authentic news organizations and unverifiable, negative social media responses to these stories - a newly discovered role of the middle media. This study also revealed that such false content produced by the middle media was eventually picked up by an influential online news portal site through the path that I call the “fake news pipeline.”