One of the more popular and well disseminated bits of COVID-related anecdata, data collected via anecdote,

rather than a more full-on and complete scientific investigative process,

involves an American candle company called Yankee Candle,

which makes, among other things, highly fragrant decorative candles that you would

likely be acutely aware of if you ever got within a few hundred feet of one of their stores or

kiosks at a mall, back when people went to malls. From cinnamon to lilac to lavender and clean

cotton, Yankee candle is probably best known for their distinct lack of subtlety in their fragrances, which are potent and long-lasting, which is what a lot of their customers appreciate about their products.

And this is why a science illustrator and cartoonist named Terry Nelson thought

this company would be a useful source of anecdata for an emerging question about how COVID

spread and could potentially be detected using our ability, or lack thereof, to smell things.

So back in November of 2020, she scoured online reviews of Yankee candle products and found that there was a noticeable influx in negative reviews,

noting that the candles these review-leaving customers had purchased didn't have any fragrance.

A research assistant with the Harvard Study of Adult Development at Brynmar College, named

Kate Petrova, saw this tweet, and decided to take it a step further.

She scraped about 20,000 reviews of the most popular scented and unscented candles on Amazon, meaning she used code to

collect these reviews and aggregate them in such a way that they could be parsed for data,

which is a bit more formal and organized than the approach taken by Nelson, and she found via

that scraped data that before 2020, the reviews of these top-ranking

scented candles stayed pretty consistently in the four to four and a half stars range,

which is pretty good.

But from January 2020 onward, they dropped to an average of around three and a quarter

stars, substantially worse.

...