Hacker News new | past | comments | ask | show | jobs | submit login

Not necessarily.

(1) It's important to make sure the numbers you're using are directly comparable -- e.g., if #confirmed is the count of unique individuals who are confirmed to have the virus and #tested is the number of tests administered rather than the number of people who have been tested then you won't get the expected results.

(2) There can be confounding variables like choosing to test people with milder symptoms.

(3) False positive/negative rates can be a factor (though they shouldn't be in a halfway decent test). If your false negative rate is x and your false positive rate is greater than 1-x then your positive test result rate will fall as your true positive rate rises.

(4) Etc. The problem is that the numbers you're measuring (reported test results) aren't the numbers you care about (true infection rates). They might align, and they will mostly align with good tests and unbiased data that are correctly and honestly reported, but they don't have to.

Caveat: I haven't been closely following covid news and have no idea how much any of those potential discrepancies might apply IRL.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: