> Rapid tests are right 80% of the time. Two rapid tests back to back are right 96% of the time.
I doubt the false negative % for two tests on the same person at roughly the same time are uncorrelated. (For two reasons: It seems logical that they would depend on viral load, and/or other non-random factors; And if they were independent, we would be able to achieve arbitrarily high specificity by taking and processing multiple samples -- just three would get you >99% -- which doesn't seem to be happening.) So the false negative rate for two tests might be much closer to 20% than 4%.
> That means after 2 tests there's a 96% chance you don't have COVID, and a 4% chance you do.
This part is definitely not accurate. The real probability will depend on base rates and facts about the individual case.
I doubt the false negative % for two tests on the same person at roughly the same time are uncorrelated. (For two reasons: It seems logical that they would depend on viral load, and/or other non-random factors; And if they were independent, we would be able to achieve arbitrarily high specificity by taking and processing multiple samples -- just three would get you >99% -- which doesn't seem to be happening.) So the false negative rate for two tests might be much closer to 20% than 4%.
> That means after 2 tests there's a 96% chance you don't have COVID, and a 4% chance you do.
This part is definitely not accurate. The real probability will depend on base rates and facts about the individual case.