Random Subject Samples, Not So Random

Article points out that subject samples in most behavioral and social science research studies are not truly randomly chosen as would ideally be called for. What isn't mentioned is that subject samples in biological and medical studies (which also generalize results to entire populations) are also not random, and thusly flawed (indeed, the entire notion of "randomness" is almost undefinable, even in math, let alone in other sciences):

http://tinyurl.com/yhas5ar