the result too sensitive to selection bias.
不知道这批受试者当中多大比例有症状。也不知测试假阳性多高。The sampling may not be random and
所有跟帖:
•
1)kit sensitivity is about at 60% or 91%, specificity is 99.5%.
-三丝-
♂
(0 bytes)
()
04/18/2020 postreply
05:27:40
•
假阴性的会多 (80%)。sampling is biased towards facebook 白人女子组。
-三丝-
♂
(0 bytes)
()
04/18/2020 postreply
05:30:30