----- Original Message -----
From: "Roy Carr-Hill" <[log in to unmask]>
To: "Paul Spicker" <[log in to unmask]>
Sent: Thursday, November 11, 2004 3:08 PM
Subject: Re: Statistical certainty
> RE: the Law argument.
>
> Of course we don't want people convicted on the probability of being black
in a white area (for example). But I can't see anything particularly
unjust about using probability arguments so long as they are used correctly
(and of course we have well-known examples where they were not). In fact,
I think it is a plausible account of what goes on in an adverserial
courtroom - like those in the UK - that both sides try and build up a
picture of the situation from selective use of or distrotions of the
evidence that their (defence or prosection) account is 'beyond reasonable
doubt' where they precisely mean a situation that is very unlikely to occur
by chance.
>
> So, a blanket statment that probability arguments would be inherently
unjust is a commentary on the whole system that I might agree with but
perhaps not others. Surely, given what we know about people's - including
statisticians - difficulty of comprehending large numbers, it could only be
useful if there were more precision where that is possible (probably only in
a limited number of cases).
>
> As I am emailing from an Internet cafe on Hanoi using yahoo and can't
easily access my University address (where I am registered with RadStats),
could you please possibly forward this to the List
> Roy Carr-Hill
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
*******************************************************
|