[Martin Taylor 940829 10:40]
Rick Marken (940826.0745) and Rick Marken (940825.1440)
Yesterday, I posted comments on Avery's comp.ai.phil and Martin's Degrees of
Freedom posts. I got an ACK that it had been sent to the listserver but I
have not seen it appear at my other address or on NewsNet.
Well, it got here. Comments below. But I have recently been observing the
same phenomenon. In the past, it was touch and go whether my posting or
the acknowledgment got here first, whereas recently it has often taken hours
or even days between the acknowledgment and the posting.
ยทยทยท
==================
Yep. And at the moment, in my opinion (and, apparently in Bill Powers'
opinion too) the number of input df that can be controlled simulataneously
is no greater than the number of output df available to control them.
Well, at least you've got that far. Congratulations. This is the starting
point for the whole argument about alerting. To repeat: The number of
input df that can be controlled SIMULTANEOUSLY is no greater than the
number of output df available to control them. It's a straightforward
and basic principle, that should not have caused any problem. Now carry
the thought forward to its implications (hint: it may help to go back to
previous postings over the last year or two, and more particularly the
last month or three).
Have you seen any of this "passive statistical analysis" taking place? Could
you describe what kind of statistical analysis is being done and explain how
we could all observe it too?
My first thought is to say that this is an irrelevant question, because if
there were none, the problem would be worse by several orders of magnitude.
I estimated about 4 orders of magnitude to be generous, but allowed Bill P.
another two to make sure I wasn't cheating. But if you prefer to solve the
problem on the assumption that there is none, be my guest.
But then I thought: "that's not fair." Of course you can see it, at least
some of it, physiologically if not psychologically. The data flow is reduced
by a factor of 100 between the retinal receptors and the optic nerve, so
two orders of magnitude are accounted for right there. I don't know whether
there are other places where such dramatic reductions can be observed, but
since we know how to do a least a little better in computer image processing,
I assumed that nature has found a way to do better than we know how, to
reduce the magnitude of the problem. But there remain several orders of
magnitude difference between available input df and available output df,
whereas, as you accept:
The number of input df that can be controlled SIMULTANEOUSLY is no greater
than the number of output df available to control them.
Martin