Influencers' Reposts and Viral Diffusion: Prestige Bias in Online Communities

Niitsuma Takuro,Mitsuo Yoshida,Hideaki Tamori,Yo Nakawake

2024-11-08

Abstract:Cultural evolution theory suggests that prestige bias (whereby individuals preferentially learn from prestigious figures) has played a key role in human ecological success. However, its impact within online environments remains unclear, particularly regarding whether reposts by prestigious individuals amplify diffusion more effectively than reposts by non-influential users. Here, we analyzed over 55 million posts and 520 million reposts on Twitter (currently X) to examine whether users with high influence scores (hg-index) more effectively amplified the reach of others' content. Our findings indicate that posts shared by influencers were more likely to be further shared compared to those shared by non-influencers. This effect persisted over time, especially in viral posts. Moreover, a small group of highly influential users accounted for approximately half of the information flow within repost cascades. These findings demonstrate a prestige bias in information diffusion within digital society, suggesting that cognitive biases shape content spread through reposting.

Social and Information Networks,Computers and Society

What problem does this paper attempt to address?

The problems that this paper attempts to solve are: **In online communities, can influencers expand information dissemination more effectively by reposting others' content? And is this phenomenon affected by prestige bias?**

Specifically, the researchers want to explore the following aspects:

1. **Can influencers disseminate content generated by others more effectively?** Previous studies have mainly focused on how influencers generate and disseminate original content, but have not fully explored their performance when reposting others' content. This paper aims to verify whether these users still have strong diffusion capabilities when reposting others' content.

2. **The existence and influence of prestige bias**: Prestige bias refers to the tendency of individuals to learn information more from socially successful or prestigious figures. This paper studies whether this bias also exists in the digital environment, especially in the process of information dissemination on social media.

3. **The role of influencers in information dissemination**: The researchers hope to understand the role of influencers in the entire information dissemination chain, especially how they affect the early - stage information diffusion, by analyzing a large number of Twitter posts and their reposting situations.

4. **Differences in dissemination efficiency among different types of users**: By introducing the concept of "repost cascade probability (CRP)", the researchers quantify and compare the efficiency of users at different influence levels in information dissemination to evaluate the specific manifestations of prestige bias.

### Research methods

To verify the above problems, the researchers adopted the following methods:

- **Data collection**: Analyzed more than 55 million posts and 520 million reposts, covering data of 14.9 million unique users.
- **User classification**: According to the users' `hg - index` (a measure of the users' ability to generate and disseminate original content), the users are divided into six influence levels.
- **Concept introduction**:
- **Repost cascade**: Describes the dissemination chain formed by information through multiple reposts.
- **Secondary spread**: Refers to the phenomenon that after a user reposts others' content, this content is further reposted by other users.
- **Virtual timeline**: Simulates the posts that users see on the timeline and their subsequent behaviors.
- **Repost cascade probability (CRP)**: Measures the probability that the reposted content will continue to be shared.

### Main findings

The research results show that:

- **High - influence users still show higher CRP values when reposting others' content**, especially when the content is very popular (≥1000 reposts).
- **A small number of high - influence users have a significant impact on information dissemination**: Although they account for only 1% of the total number of users, these users contribute approximately half of the reposts and views.
- **Prestige bias does indeed exist**: The content reposted by influencers is more likely to be further disseminated, indicating that prestige bias also plays a role in the digital environment.

### Conclusion

This study shows that **prestige bias does exist in online information dissemination**, and high - influence users not only perform well in generating and disseminating original content, but also have significant advantages when reposting others' content. This provides a new perspective for understanding the information dissemination mechanism and has important implications for the design and management of social platforms.