COVID-19 and Bayes’ Theorem
Within the past two years, the COVID-19 pandemic spread globally in an exponential manner. The overwhelming cases prompted action to be taken place in order to help people be aware of their potential of getting infected or already having the virus. Showing symptoms of the virus can be highly indicative of being infected but there is still a probability that you do not have the virus. This uncertainty can be measured using statistical models like Bayes’ Theorem. Bayesian models helped people be more conscious of their probabilities of having the virus through the likelihood of symptoms and more aware of the accuracy of the COVID tests. Statisticians, researchers, and epidemiologists from all over the world came together and created bayesian strategies to help with predicting the accuracy of the COVID tests. They wanted to be sure to take into account false positives and false negatives when test results are distributed to people.
The models these teams came up with took into consideration randomness which stemmed from the priors (assumptions or beliefs before evidence is presented). They found that the randomness or the random predictors had a strong connection to the tail events of the situation which means they had more to do with the more rare events. It was highlighted throughout the majority of the article that humans use Bayes’ theorem unconsciously, but they just do not have concrete evidence to actually calculate probabilities. Humans use Bayes’ theorem in predicting things based on their observations. There may be things that seem logical to us humans, but when you actually do the Bayes’ calculations, you can be shocked at certain results. Things that you might have thought were highly unlikely can be the complete opposite, using evidence and calculations. This one formula is very powerful in the sense that it can help humans comprehend things that may seem improbable. Something as big as the novel coronavirus has a lot of fuzzy areas, including the testing of one’s COVID status, yet this one theorem makes it slightly comprehensible.
https://www.nytimes.com/2020/08/04/science/coronavirus-bayes-statistics-math.html