Quote:
Originally Posted by Low C Sharp
Surely the data show a pattern. If there's really no relationship between PBK chapters and some social GLOs, then you wouldn't expect to see such massive variation among the GLOs in one conference. You'd expect to see a large number of groups around the mean and a few outliers at the top and bottom of the scale. But you don't see that -- not even close. In fact, there's a statistically remarkable clustering of a few groups that have huge PBK overlap. There's clearly a non-random association between (for example) KKG and PBK, and that is of interest even when the data themselves say nothing about the cause. That's where the chat comes in.
|
And I don't deny the fact that this might be interesting statistically, but my point still stands; and to my original point I will add that the concept of "elite" is relative.
Threaded Mode
