[burger]

Assuming that NYC's test has accuracy similar to the Stanford test, false positives should not be a major issue for the NYC and NYC-area results.  It could matter for the statewide results.

But we'll need more information to know how good the sampling was.  I saw that they sampled at grocery store and big box store parking lots.  That's potentially more random than calling people or facebook ads, but the devil is always in the details.

For future studies like this, I think one way around the self-selection problem would be to pay people to participate.  If you give everyone an incentive to give blood, maybe you can reduce motivation to get tested as a factor.  I bet China could do a great study--just have the blood extraction people show up at everyone's door with a few police officers.  100% participation guaranteed!

[burger]

Assuming that NYC's test has accuracy similar to the Stanford test, false positives should not be a major issue for the NYC and NYC-area results.  It could matter for the statewide results.

But we'll need more information to know how good the sampling was.  I saw that they sampled at grocery store and big box store parking lots.  That's potentially more random than calling people or facebook ads, but the devil is always in the details.

For future studies like this, I think one way around the self-selection problem would be to pay people to participate.  If you give everyone an incentive to give blood, maybe you can reduce motivation to get tested as a factor.  I bet China could do a great study--just have the blood extraction people show up at everyone's door with a few police officers.  100% participation guaranteed!