2 ms·
It's not a random sample; it's a convenience sample of people who responded to Facebook ads.
by mlyle 7y ago
It's not a random sample; it's a convenience sample of people who responded to Facebook ads.
- Medicalidiot 7y agoWhat about Facebook ads doesn't mean that they're an adequate sample? Is there a confounding factor that you have in mind?
- mlyle 7y agoMany confounding factors. Socioeconomic factors that can't be fixed with stratified sampling (access to transportation, usage of Facebook, etc.. conversely the willingness of a $10 amazon gift card to motivate people to participate will vary). People who are concerned about possible past exposure to COVID-19 may have a different probability to respond. Of course, the bigger issue is that we can't rule out that the antibody tests have a false positive rate that can explain the whole result. We need serology tests in places with a higher positive rate to know. Basically, 50 positive results out of 3330 is a really small signal. Just a small false positive rate or sampling issue could explain all the positive results.