Hello Hope,

Quote Originally Posted by Hope123 View Post
Coming from a different point of view, I like Bradford Hill criteria. Koch is good for labwork whereas Bradford Hill looks more at populations. Here's a quick link:

http://www.southalabama.edu/coe/bset...20criteria.pdf
 

Good set indeed! (And yes, statistical research of populations is different from lab search into - more or less - established natural kinds, whether copper or mice. The difference is that a natural kind is characterized by a definite set of properties that permit generalizations - "all copper...", "all mice.." etc. - whereas populations may include various kinds of things.)

Thanks for posting it.

Maarten.