"Ross J. Micheals" wrote: > > Is there a statistical definition of "representative sampling?" That is, > is there a way to show that a sampling strategy will yield (within some > tolerance) a sample that is "representative" of the population? For > example, given a sampling strategy with replacement, clearly, if I > always choose the same member/person in my population, then this is > *not* a representative sample (I always get the same person!) What sort > of sampling schemes *are* "representative" (if this is even a meaningful > concept, statistically), and how can one show this? > > References or thoughts? Thanks in advance for your help. > > Sent via Deja.com http://www.deja.com/ > Before you buy.
I don't know if there really is such a formal, technical term as representative sample. What matters for a statistician (insofar as generalizability is concerned) is whether the elements of the sample were obtained by using a probability-based method that can be described fully.