Want to know how many people have been infected? Test randomly

So how many people do you need to randomly test to get data that can accurately describe the whole U.S.? Fortunately, the mathematics behind this question have long been worked out, and the number is probably smaller than you might think.

Presidential approval polls often sample roughly 1,000 people. This produces a margin of error of approximately 3%, meaning that random chance could make the results off by up to 3%.

A margin of error of 3% may be fine for estimating presidential approval, but it is probably not accurate enough for the coronavirus pandemic. If 10,000 individuals in the U.S. were tested for the virus, the margin of error for the virus’s infection rate becomes 1%. In practice, these margins of error are conservative. Actual margins of error from a random sample of 10,000 individuals will probably be much smaller and likely accurate enough to start giving public health officials useful information about the total number of infected and case fatality rates for those who have the coronavirus.