Testing a limited number of random people, we can reach conclusions within a certain probability.
Statistical analysis is a funny thing. It turns out that you don’t need a lot of data to reach a conclusion. At the start of the play “Rosencrantz and Guildentstern are Dead,” the two main characters, bit players in Shakespeare’s Hamlet, are flipping a coin. The coin has come up heads 92 times in a row. The action starts when the coin finally comes up tails. Mathematically, if you flip a coin, the chances it will be heads or tails is 50 percent. If you flip the coin again, though, the odds that it will be heads the second time drop to only 25 percent.
Why is this important? Well, in reaching decisions about a characteristic in a group, like exposure to COVID-19, at some point, by testing a limited number of random people, we can reach conclusions within a certain probability.
In statistics, we call these confidence levels. These levels are most often given as 95 percent or 99 percent. Now, here is the interesting part. Whether our sample size is the one million people in a particular city or the 365 million people living in the United States, the sample size to determine a characteristic does not change a lot. Remember my coin example?
Based on statistics, and assuming people have similar characteristics, if I sample 16,369 people in a group of a million people, I can determine within 99 percent probability the percentage of people with COVID-19. If I increase the sample size to 365 million people, the sample size only increases to 16,640.
Instead of only sampling people who are sick and may have COVID-19, if we test random samples, both nationally and locally, we can make better decisions and cut through the fear.