An Analysis of Omicron Tweets: 30% are Skeptical of the Medical Establishment

Jefferson Lee
Jan 21, 2022
An Analysis of Omicron Tweets: 30% are Skeptical of the Medical Establishment

Social media is often criticized for spreading anti-vaccination posts. For example:

But how common are these types of tweets really?

And with Omicron's arrival, what sentiments do Twitter users have around COVID-19 now? Are they scared of Omicron's high infectiousness? Jaded by the possibility that COVID is here to stay? Skeptical of the disease and distrustful of the news around it?

In order to answer these questions, we created a dataset of 2000 recent Omicron tweets and labeled the emotions behind them.

In short:

  • 30% of Omicron tweets are skeptical of the medical consensus that vaccines and masks are effective at minimizing the spread of coronavirus.
  • 35% of Omicron tweets that mention vaccines are anti-vaccine. If we limit to tweets from medical workers, this number drops to 11%.
  • 71% of Omicron tweets express anger, fear, or skepticism. 27% express happiness or compassion. 14% express indifference.

For more details, read below!

Methodology

We first filtered for tweets from Individuals (as opposed to Organizations), and categorized them into four types:

  • Skeptical tweets that explicitly denied that COVID is real or that vaccines or masks are effective
  • Positive tweets that expressed happiness, relief, compassion, or lack of concern about the new variant
  • Negative tweets that expressed fear, hopelessness, or anger
  • Objective tweets with purely factual content (school closure notices, case count statistics, etc.)

Medical Skepticism and Distrust

The most common emotion in Omicron tweets is skepticism of the official medical consensus. Tweets like the one below accounted for 30% of tweets with any emotional content:

Pro- and Anti-Vaccination Views

10% of tweets explicitly mentioned vaccination. How did people feel about the vaccines?

~65% were pro-vaccine. ~35%, however, were anti-vaccine and distrustful of the medical establishment.

January Blues

Anger and Fear are the dominant emotions among Twitter users who do believe in COVID.

Anger was almost 3x as common as compassion:

Fear was 2.5x more common than happiness or relief:

But Twitter isn't all doom and gloom. 10% of users shared happy news like the tweet below:

What’s up, Doc?

How do medical workers feel compared to the general public? We also categorized whether tweets came from users who identified as doctors or nurses. 

Just like the rest of us, healthcare workers are mostly afraid and angry, and many of their tweets contained personal stories:

There’s one way that doctors are unique, though: they really encourage vaccination. Medical professionals were three times more likely to tweet their opinion about vaccination, which was overwhelmingly in favor at 89%.

While some anti-vax messages did come from self-identified medical professionals:

They were outnumbered by pleas for people to get vaccinated 9-to-1:

Until next week, stay safe and stay curious. And if you're interested in our full dataset of Omicron reactions, shoot us an email at team@surgehq.ai


Interested in running your own sentiment analysis? Surge AI is a data labeling workforce and platform that provides world-class data to top AI companies and researchers. Fill out our short form and get $50 of free labeling credits today.

surge ai logo

Data Labeling 2.0 for Rich, Creative AI

Superintelligent AI, meet your human teachers. Our data labeling platform is designed from the ground up to train the next generation of AI — whether it’s systems that can code in Python, summarize poetry, or detect the subtleties of toxic speech. Use our powerful data labeling workforce and tools to build the rich, human-powered datasets you need today.

Jefferson Lee

Jefferson Lee

Jefferson leads Surge AI's data labeling and content moderation products — whether it's helping customers evaluate large language models, moderate content, or train Spam and Hate Speech classifiers. He was previously an early engineer on Airbnb's Trust and Safety ML team, and studied computer science at Harvard.

Data Labeling for the
Richness of AI

Build human-powered datasets using our global labeling workforce and platform.