It is not what you say but how you say it.

 In the world of digital journalism, this common aphorism expands to multimedia strategies involving platforms, photo choices, interactivity levels and all other components and possibilities leading to the successfulness of a news report or a social campaign. Moreover, as a new channel of media spread, this successfulness in social media is transferred to the level of motivation of content sharing among the media users. 

My research seeks to understand both “how you say it” —  message processing and design in the context of online journalism and health communication —  and what influences people’s sharing intention of media content in social media, with a special focus on the role of emotion.

My research would primarily benefit both news writing for professional journalists and message design for social campaigns when facing the choice to present and organize information. I am working on answering the important questions such as what is the right approach and style in terms of analytical versus emotional on different topics, what specific emotion should be targeted in a given situation?  How do emotions and other factors shape the virality of media content?  By answering these questions, my work facilitates the design of effective messages and also makes the message influential by increasing message exposure via sharing.  



online news sharing

During the missing Malaysia Airlines flight MH 370 crisis, working together with Dr. Jueman Zhang and Dr. Ross Buck,  I examined the effects of analytical versus emotional content, modalities, and interface cues on intention to retweet news posts and video about the missing Malaysia Airlines flight on Weibo, the Chinese Twitter-like microblogging platform. We found that news posts with analytical content were likely to be retweeted than news posts with content that led to sad emotions. The study also investigated how the weight and order of analytical and emotional content in a video affected attention, arousal, and intention to retweet the video. Videos with content that was largely analytical and slightly emotional attracted more attention than videos with content that was largely emotional and slightly analytical. The former were less arousing than the latter. This work offered insights on different strategies for news agencies when selecting content to create posts on social media. Moreover, if the essential goal for the content creators is to have the most widely shared content on social media, then an analytical posts with neutral valence are considered more valuable and people are motivated more to share them. This work is forthcoming as a book chapter in Social Media and Crisis Communication.

Example Stimulus. Footage from CNN. Analytical video with analytical opening content. 

Example Stimulus: Emotional Post with call to action cue.

Example Stimulus: Emotional Post with call to action cue.

Example Stimulus:   Post with call to action cue.

Example Stimulus:  Post with call to action cue.

Message processing and design

narrative versus statistical evidence

Photo Source: www.

Photo Source: www.

Used self-reports and physiological measures — heart rate (HR) and skin conductance level (SCL) —  my  colleague (Dr. Jueman Zhang, Dr. Makana Chock and Dr. Gina Chen) and I examined the effects of novelty appeals, sexual appeals, narrative versus statistical evidence, and viewer’s sex on cognitive and emotional processing of HIV/AIDS PSAs among heterosexually active single college students. 

We found that novelty or sexual appeals affected self-reported attention and cognitive effort (as measured by HR) differently. Moreover, high novelty or sexual appeals led to greater self-reported arousal, but not greater SCL. Therefore, self-reported and physiological responses provide information that is available and unavailable to the participant’s conscious awareness, respectively. Interesting interactions were found in this study, such that narrative evidence accentuated the effects of novelty appeals on cognitive effort but effects of high sexual appeal were more obvious in statistical than in narrative PSAs. The findings about the interaction effects provided implications about how to use message appeals and message evidence together to design more effective HIV/AIDS PSAs. If the goal of an HIV/AIDS PSA is to increase a viewer’s cognitive effort, the results of the current study endorses the use of novelty appeals with narrative evidence or the use of sexual appeals with statistical evidence. This work is forthcoming in Health Communication.




Working together with Dr. Carolyn Lin, we produced four PSAs guided by both protection motivation theory and message frame construct to investigate how the text and visual content manipulated through STI prevention public service announcements (PSAs) may influence the persuasiveness of these PSAs and protective behavioral intention of Chinese college students.

A nested design is employed to test four different combinations of message appeals and message frames.  The two “emotional appeal” conditions include an emotion-oriented video matched with either a gain-framed or loss-framed message.  The two “informational appeal” conditions include an infographic-based video matched with either a gain-framed or loss-framed message.  


Drawing Credit: Jieting Chen

Social EMotion

From NCA presentation 2013.

From NCA presentation 2013.

Working together with Dr. Ross Buck, we tested the ecological systems view on eight primary social emotions emerging naturally from universal ecological contingencies involving social comparisons of relative success or failure (Buck, Emotion: A Biosocial Synthesis, 2014). These form four twins (in English): pride/arrogance, guilt/shame, envy/jealousy, and pity/scorn. They are displayed by dominance-submission behaviors.

This implies two hypotheses about the cultural universality of social emotions: the hypothesis of universal labeling that there should be words describing the eight fundamental interpersonal contingency unions, and the hypothesis of universal dynamics that these words should be interrelated similarly; in all nations, cultures, languages, and historical times.

These were tested in samples from four cultures-languages: America, Japan, and Uyghur and Han samples in China. Judges responded in their native language to scenarios involving social comparison with another person who achieved relative success or failure by own efforts or by chance. Judges indicated how they would feel about that person and how that person in turn would feel toward them, using the eight primary social emotions.

Results indicated that words corresponding to the primary social emotions existed in each culture, and that generally they were interrelated similarly as expected. For example, across the four groups twins were positively related (Average r = .708), as were mirror converse emotions that are hypothesized by the model (Average r = .756: e.g., if P is proud/arrogant, the Comparison Other will tend to feel envy/jealousy).

International Summer School in Affective Sciences 2014 -Emotion meets action

2014 International Summer School in Affective Sciences, Switzerland.

In summer 2014, I was selected to participate in the International summer school in Affective Sciences in Switzerland. Various ways in which emotion and action can be related were discussed among experts in psychology, neuroscience, and philosophy. During the summer school, student participants worked in teams on interdisciplinary research projects that try to build models integrating insights from lectures on both emotion research and action research.


  • The role of emotions as reasons for action and decision making
  • Overlap between emotion and action in fiction and music
  • The role of action plays in emotion theories
  • The role of emotion plays in action theories