Abstract: As internet plays an increasingly important role in communication around the world, people spend more and more time on social media. Research have found that digital emotion contagion, polarized emotion in particular, was associated with adverse outcomes, such as spread of fake news, political segregation, etc. Previous research has found that the emotional experience among East Asians is more balanced between and mixed with both positive and negative emotions, compared to westerners (Goetz et al., 2008). And this difference is explained by dialectical thinking, a thinking style that is about change and contradictions and is more prevalent among East Asian cultures than Western cultures (Peng & Nisbett, 1999). In this project, we will examine whether dialectical thinking may also have impacts on the expression and contagion of polarized emotions on social media. In particular, we hypothesize that individuals with a high level of dialectical thinking may experience and have a stronger preference for more balanced and less polarized emotions, and would be less likely to express and share polarized emotions on social media. Below are studies proposed to examine these hypotheses.
Study 1 will examine the content of tweets from American and Japanese user with sentiment analysis (Thelwall, 2017). We expect that tweets posted by American users would show a higher level of emotion polarity than those posted by Japanese users. Furthermore, among the English tweets, emotion polarity of the tweets would be positive associated with the amount of likes and retweets., but this association would not be found among the Japanese tweets, indicating that polarized emotions are more likely to be shared among the non-dialectical culture than the dialectical culture.
Study 2 will examine the expression and contagion of polarized emotions in Study 2a and Study 2b, respectively. In Study 2a, Chinese and American participants will be randomly assigned to read a positive, negative, or mixed emotion eliciting scenario. Then participants will need to post a tweet on Twitter to express their emotional feelings towards the scenario in public, or share their emotion experience with a friend. We expect that Chinese participants, in particular under the public sharing condition, would have a stronger motivation to express contents that will be approved by their culture, and would express less polarized emotions than the American participants. In Study 2b, dialectical thinking will be primed among participants, and participants' attention toward tweets with different emotional contents will be traced by an eye-tracker. We expect that under the dialectical condition, participants would show a stronger preference for more balanced emotions and thus their attention would be directed away from the contents with polarized emotions, indexed by a longer gazing time recorded by the eye-tracker. In addition, participants under the dialectical priming condition would be less likely to like or retweet the tweets with polarized emotional contents.
In Study 3, a field experiment will be conduct to examine whether dialectical thinking priming would reduce the expression and dissemination of polarized emotions in a real-life. In Study 3a, posts with different levels of polarized emotions will be posted in a social media community each day, and participants will be required to select and comment on those posts. We expect that participants with dialectical thinking priming would be less likely to respond to the posts with a higher level of polarized emotion, and these posts would be less likely to be shared. Based on the findings from Study 3a, participants' preferences and strategy on sharing posts will be calculated, and these strategies will be compared with other different simulated sharing strategies in Study 3b, such as, sharing polarized emotions, random sharing. With agent-based modeling, we will examine the consequences of different sharing strategies, in term of group segregation in social networks (Jackson et al., 2017). We expect that polarized emotions sharing would lead to the highest level of group segregation than did other strategies.
Taken together, this project aims to examine how dialectical thinking may influence the processes of digital emotion expression and contagion with big data analysis, survey, and eye-tracking studies (Study 1 and 2). In Study 3, with priming, field experiment, and agent-based modeling, we will further examine whether dialectical thinking can reduce the contagion of polarized emotions on social media and its negative consequences. Theoretically, this project highlights the how psychological factors, such as culture values, emotion processes, may affect the emotions and behavioral on social media. Practically, this project will shed light on developing intervention that can counteract the polarized emotion expression and contagion in social media and their negative consequences.

Key words: digital emotion contagion, emotion expression, polarized emotion, culture, dialectical thinking