> David Lieb wrote: "How many of us were tested? Not me. How many of us
> KNOW anyone who took this test? Not me. How many of us think that these
> results reflect the bias of the people asking the questions? Insert
> Horschak sound track here."
> You would think so, but the science of statistical methodology can now
> enable such polls to be conducted with relatively small samples. In and
> of
> itself, large groups don't add any more validity to it.
Horse hockey. Technically, you are correct, because there was no validity to
start with. You know as well as I do that the motto of these "statisticians"
is 'tell me what results you want and we will get you the numbers to back it
up'. The "relatively small samples" simply means that it doesn't cost as
much.
_______________________________________________
Edit your replies
http://www.team.net/archive
http://autox.team.net/mailman/listinfo/spridgets
|