Having invited this discussion, I feel entitled to call a closure, though
this may be overruled by the list-owner (Ray Thomas) whose powers are
defined in Mailbase's acceptable use policy document:
http://mailbase.ac.uk/docs/aup.html
I do not regret inviting William Chambers to explain his method, though I
will now disappoint him by saying that, having looked carefully at his
several messages, and now understanding what he means by polarisation, I
stick absolutely to the position that *you cannot infer causality from
correlation*. His subsequent messages make it clear that Bill is, in
fact, deriving any notion of causality from a theoretical or chronological
perspective. The example in his first message (his "manifold") represents
a complete factorial experiment, so again implies design or manipulation.
Continued insistence on inferring causal relationships from observed
correlations is not sophistry, but is pure distortion of defined concepts.
Bill forwarded to the list an abusive message sent individually to both
of us by Eric Brunner. I have replied (twice) to Brunner explaining why
I found his message offensive and unjustified by any action on my part.
He seems incapable of forming a private apology, so I will here invite a
public apology and withdrawal via radstats.
I have contributed to email discussions for over ten years, and am an
active member of several discussion lists. In view of the current
promotion of a general discussion list for statisticians (on statistics,
statistical methods and professional interests), it is most important to
reassure those users of email who "fear the deluge", that incidents of
personal abuse are rare, and that they will be stamped on as soon as they
manifest.
R. Allan Reese
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|