CM – A study of Japanese and US Twitter users reveals why emotional posts are more likely to go viral

Click here to log in with

or

Forgot Password?

Learn more

September 8, 2021

by Vignesh Ramachandran, Stanford University

It happens all too often: You log on to social media for a fun break but are upset and angry an hour later. Why does this keep happening?

googletag.cmd.push (function () {googletag.display (‘div-gpt-ad-1449240174198-2’);});

New research from Stanford University psychology professors Jeanne Tsai and Brian Knutson and their team suggests that a key piece of the puzzle is culture. In a new study published September 7 in the Journal of Personality and Social Psychology, the team evaluated the emotional content, known as « sentiment », of Twitter posts by Japanese and US users and found that Users are more likely to be influenced by the contributions of others if the contributions violate rather than support their cultural values.

Although social media spans the whole world, current studies focus disproportionately on users in the USA and not on users from non-Western countries Contexts.

« To me, it just felt like a piece was missing if you don’t take culture into account, » said lead author and PhD student Tiffany Hsu. “Culture is such an important aspect of social media and connecting people.” For example, in the US, most people say they like to feel excited, happy, and positive. However, previous research has found that US social media users are most affected by posts that express anger, anger, and other negative emotions. What can explain this paradox?

To find out, the Stanford researchers reached out to social media users in a non-Western country, particularly Japan. Research has shown that Japanese people generally value other affects (emotions) that are less aroused (e.g. calmness, equanimity) than people in the United States.

To test whether Japanese users are also more interested in social -Media posts that violate their values ​​are influenced, the team developed a sentiment analysis algorithm that can be applied to both American and Japanese content.

It was no easy task to get this technology from scratch To build on: The team used machine learning based on manually coded labels from 3,481 Japanese tweets to get them working. The program was based on SentiStrength, which rated short English texts for positivity and negativity and emotional intensity / arousal. The researchers developed a similar algorithm for Japanese texts.

The researchers found that social media users in the US generally had more positive content (e.g. exam week, I’m kind of free 😒), in line with their respective cultural values . These findings spanned a variety of topics, including personal issues, entertainment, and politics.

« Social media users appeared to be left to their own devices to produce culturally valuable content, » said Knutson, professor of psychology at the School of Humanities and Sciences (H&S).

« Users appeared to be most influenced by other people’s contributions when those contributions contained feelings that violated their cultural values, » said Tsai, Yumi and Yasunori Kaneko Family University Fellow in Undergraduate Education at H&S and director of the Stanford Culture and Emotion Lab. “So in the US, users are most affected by negative high arousal content such as anger and disgust, but in Japan users are most affected by positive high arousal conditions such as excitement.”

The team has this phenomenon referred to as « affective hijacking », where exposure to content that violates users’ cultural values ​​temporarily redirects their own content.

The source of the affectively disruptive content is not yet clear, but the fact that it often does Accompanied by propaganda or misinformation suggests it could have come from nefarious actors or even algorithms, the researchers said. Regardless of the source, the fact that the disruptive content can be measured means that tools may be developed to help users identify the disruptive affective content. « Self-control is great, but limited, » Knutson said . “Is there any software we could develop for people – just like a spam filter for email – that could help them apply affective filters to their social media? We are interested in giving people more freedom of choice. ”

This research led the team to work on developing tools that allow social media users to identify posts that“ hijack their emotional experiences ”  » can. Perhaps this could ultimately reduce the time users spend in emotional states they do not appreciate, while at the same time minimizing misinformation.

Use this form if you have identified a typographical error, inaccuracy, or if you would like to submit a change request to the contents of this page.
For general inquiries, please use our contact form.
For general feedback, use the public comments section below (please follow the guidelines).

Your feedback is important to us. Due to the high volume of messages, however, we cannot guarantee individual responses.

Your email address will only be used to let the recipient know who sent the email. Neither your address nor that of the recipient will be used for any other purpose.
The information you entered will appear in your email message and will not be stored in any form by Phys.org.

Receive weekly and / or daily updates in your inbox.
You can unsubscribe at any time and we will never pass your data on to third parties.

This website uses cookies to make navigation easier, to analyze your use of our services, to collect data for the personalization of advertisements and to provide content from third-party providers.
By using our website, you confirm that you have read and understood our privacy policy
and terms of use.