Why are there so many extreme opinions on social media?

An empirical analysis comparing Japan, Korea, and the United States

Abstract

This study tests the following hypothesis: "People with extreme opinions tend to express those more on social media, and the opinion distribution on social media differs from that of society." I analyze the impact of extreme opinions on the number of posts on social media, using questionnaire of 5,986 people in Japan, Korea, and the United States and Tobit model. The results show that my hypothesis was supported. Further, the more sensitive the topic, the more the number of social media posts increased depending on opinion extremity. This finding suggests that the distribution of posts on social media has bias compared to the opinion distribution in society.