Over the course of the Pandemic, social media sleuths, epidemiologists and health nerds alike noticed an interesting trend in the review section for Yankee candles on Amazon.

There was a spike in carbon dioxide cases when there was a lot of negative reviews.

fresh wave of bad reviews for yankee candles pic.twitter.com/1mlandB78I

— drewtoothpaste (@drewtoothpaste) December 21, 2021

Losing your sense of smell is a common symptom of an infections. People started to wonder if the reviews themselves were a good indicator of a surge in the virus.

The lack of official data tracking infections across the U.S. heading into another winter has brought that theory under scrutiny.

How a review became a warning sign

The Yankee Candle theory was first discovered by an associate professor of political science at Northeastern University.

It wouldn't be hard to find out if the link was real. He wanted to create a model to test the predictions he made using social media data.

It's easy to do. "Maybe I'll try to get some Amazon reviews and see what the trends are, instead of just cutting and paste a few reviews that mention a lack of smell."

The relationship was clear and the cases followed a similar pattern to the reviews.

Here's a plot of the "no smell" complaints for the top three Yankee Candles on Amazon. pic.twitter.com/EFUsGil5k4

— Nick Beauchamp (@nick_beauchamp) December 22, 2021

He had to add more data to find a definitive answer to the findings in December 2021. He wrote a paper in January and submitted it to a journal in June.

"It's a very small paper, but it's caught a lot of people's attention because it's trying to do slightly more carefully something that a lot of people have been noticing," he said.

If there was a surge in COVID cases, there would be an increase in the negative reviews. Is it possible that it would work the other way?

I was trying to find out if we could use the reviews to predict guilt cases. We found that up through December of 2021. You can't really do better than using the past COVID cases to predict future COVID cases.

Something happened after that. After adding more data to his model, he discovered that the relationship between the reviews and COVID rates had changed again.

The rise in negative reviews may be an early warning sign.

Maybe it's due to the lack of measurement of COVID, or worse, something else is changing. I assume the reviews weren't changing a lot.

When users notice a surge in carbon dioxide cases, the study has evolved into their own meta-data sets, gaining popularity again.

Some researchers refer to these trends as "digital breadcrumbs," because they can give insight into a person's real life circumstances.

Even with all of his controls, Beauchamp is still skeptical about the study.

Why some believe the official data is a "big mess"

As President Joe Biden declared the Pandemic "over", the quality of COVID tracking has become a cause for concern for Beauchamp and other experts.

Traditional data sources are not getting better. The CDC is reducing the amount of data it collects. Everyone is measuring themselves less often. They are less likely to report these things to the government.

He said the Yankee Candle reviews were an example of how many people were still invested in tracking carbon dioxide.

Those of us who still care about and worry about the Pandemic, and don't think that it's over, are grasping for other sources of data that can be used to track new waves and that sort of thing.

The evolving nature of the virus has made it difficult to determine the most efficient ways of collecting and analyzing COVID data, especially three years into the Pandemic.

Every case that we were documenting mattered a lot. We were trying to find out what to do with the data.

New issues like reinfections and how to document them came up as time went on. The reduction of testing is one of the hurdles. People who choose to test at home often don't report their results to public health departments.

Tracking some key sources even if they are less robust than in the past has proved to be an effective strategy at this point.

If there was no recent change in the amount of testing available, then observing trends in reported cases was the best method.

The most relevant question I get asked, as a doctor or epidemiologist, is, "What is the risk of me doing X activity to contract SARS-coV-2?" He said that the data is a big mess and that you can mostly answer that based on what's happening around you.

Wastewater data could be useful if it was not very precise in measuring case numbers.

A combination of data sources could help experts and regular people make the best decisions for themselves.

People are weighing the risks and benefits with limited data. He said that you can triangulate a lot of things to get an assessment of where we are with new variant.

Including the Yankee candle data in the mix is something to think about.

Public health uses these types of things more for research. He doesn't believe candle reviews will change the public health strategy.

It could be a sign that there is more data online that could be useful. If there is, he is all for it.

He said it was better to join together in a movement. I'm happy to be part of that.