Anti-establishment sentiment on TikTok: Implications for understanding influence(rs) and expertise on social media
This study utilizes a computational approach to analyze TikTok posts in finance, wellness, and conspiracy theory domains, finding that while anti-establishment sentiment is most prevalent in conspiracy content, engagement patterns and platform incentives suggest it may still influence user behavior across these topical areas.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine the internet as a giant, bustling town square. In this town, there are "officials" (like the government, universities, and doctors) who usually run things. But lately, a lot of people are starting to distrust these officials. They are beginning to believe that the "officials" are corrupt elites working against the common person.
This paper is like a team of detectives (researchers from the University of Michigan) who went into a specific part of this town square called TikTok to see how much of this "distrust talk" is actually happening. They wanted to know: Is this distrust everywhere, or is it only in certain neighborhoods? And does the town square itself encourage people to say these things?
Here is the breakdown of their investigation using simple analogies:
1. The Three Neighborhoods They Visited
The researchers didn't look at the whole town; they focused on three specific "neighborhoods" (topics) to see if the "distrust talk" (which they call Anti-Establishment Sentiment or AES) was common there:
- The Conspiracy Neighborhood: This is where people talk about secret plots, flat earth theories, and hidden agendas. The researchers expected this place to be full of distrust talk.
- The Finance Neighborhood: This is where people give advice on stocks, crypto, and saving money. They wondered if people here were complaining that the "banking system" is rigged.
- The Wellness Neighborhood: This is where people talk about health, yoga, and natural remedies. They wondered if people here were attacking doctors and big pharmaceutical companies.
2. The Detective Work (How They Did It)
To figure out what was being said, the researchers couldn't read every single video (there were nearly 27,000 of them!). So, they used a two-step process:
- Human Training: They hired people to read the videos and comments and learn exactly what counts as "distrust talk." They taught them to spot when someone says, "The government is lying to us," or "Doctors just want our money."
- Robot Assistants: Once the humans were trained, they taught a computer program (an AI) to do the rest of the reading. The AI learned to spot these patterns in the remaining thousands of videos.
3. What They Found (The Results)
The "Conspiracy" Neighborhood was the Hotspot
Just like they expected, the Conspiracy neighborhood was full of distrust talk. About 45% of the videos there were attacking the "establishment." It was the main theme of the conversation.
The "Finance" and "Wellness" Neighborhoods were Surprisingly Quiet
This was the big surprise. The researchers thought people might be complaining a lot about banks or doctors in these areas. Instead, they found:
- Only about 4% of Finance videos contained this distrust talk.
- Only about 1% of Wellness videos contained it.
- Most of the time, people in these neighborhoods were just giving normal advice, not attacking institutions.
The "For You" Page (The Random Walk)
The researchers also created fake accounts to just scroll through TikTok's main feed (the "For You" page) without searching for anything specific. They found that if you just casually scroll, you will almost never see this distrust talk (less than 1% of the time). You mostly see it if you actively go looking for it or if you are already interested in conspiracy theories.
4. How People Reacted (The Engagement)
The researchers also checked how people reacted to these videos:
- Likes: Surprisingly, videos with distrust talk got fewer likes than normal videos in all three neighborhoods.
- Comments and Shares: However, in the Finance and Wellness neighborhoods, people were more likely to comment and share the distrust videos. It seems like these videos spark more debate or make people want to pass them along, even if they don't "like" them.
- Agreement: In the Conspiracy neighborhood, the people commenting mostly agreed with the distrustful videos. They weren't arguing against the conspiracy theories; they were cheering them on.
5. The "Voice" of the Distrust
The researchers also looked at how these creators spoke. They found a specific "style" used by people spreading distrust:
- Us vs. Them: They used language that grouped "us" (the regular people) against "them" (the elites). They used more words like "we" and "us" and fewer words like "I."
- Confidence: They spoke with a very loud, confident tone, acting like they were the ultimate authority, even more so than the experts.
- Morality and Fear: They often used religious language or talked about death and danger to make their point feel urgent and moral.
The Bottom Line
The main takeaway is that while "distrust of the system" is a big topic in the news, it isn't actually dominating TikTok's everyday content.
- If you are looking for conspiracy theories, you will find a lot of it.
- If you are looking for finance or health advice, you will mostly find normal advice, with only a tiny bit of distrust mixed in.
- If you just scroll randomly, you are very unlikely to see it at all.
The paper suggests that while social media can reward this kind of content, it doesn't mean it's everywhere. It depends heavily on what you are interested in and what you choose to watch.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.