I agree with John Logson that the application of statistics is amoral.    

But by carelessly suggesting that modelling could could be used to allocate
blame John proposed an immoral application, ie inconsistent with morality.

He actually says things like "The responsibility for the casualties in this
case is in some way shared, even if the family may not realise it."   The
implication of the situation he hypothecates is that a family by being
Iraqi, by being in the wrong place, by being attached to the wrong man, are
in part responsible for their own destruction.   The objectionable and
immoral word here is 'responsible'.   

Responsibility lies with the agencies that created the armed conflict
situation.  So a little bit of history is unavoidable.  In this case it is
necessary to go back only as far as 2003.  (Whether it was 'wise' to go to
war then in a different matter. There is an abundance of evidence pointing
to the folly of this enterprise.  But, as John says, this raises issues
which are not the  fundamental concerns of this list).

The major point is that statistical modelling is amoral.   Modelling can be
used to imply for example that the poor are responsible for their own
poverty and is so used.   

John says "Taking Ray's complaint to its logical conclusion (ie don't use
statistical tools where moral judgements are involved)" again shows woolly
thinking or woolly expression of thought.   The statististician who created
the model should take responsibility for moral judgments made on the  basis
of the model created.   For the creator to shelter behind the model
proclaiming that 'the model shows who is to blame' is, I contend, an immoral
position.

John's basic confusion here is to fail to distinguish between descriptive
and inferential statistics.   The survey published by the Lancet belongs to
descriptive statistics.   It is to be hoped that it does have a political
impact.   But the use of a model to allocate blame for deaths in the
conflict is tendentious to say the least.  It distracts attention from the
main consequences of the conflict and from the causes of the conflict.
Many believe, including intermittently (as Washington allows) Tony Blair,
that dealing with causes is a necessary part of finding a solution.

So don't let statistics detract from finding solutions.

Another bizarre part of Johns message is to say that I have proclaimed the
"uselessness of history".   If I have ever given that impression I would
like to know how.    Can anyone quote chapter and verse?


Ray Thomas, 35 Passmore, Tinkers Bridge, Milton Keynes MK6 3DY



-----Original Message-----
From: email list for Radical Statistics [mailto:[log in to unmask]] On
Behalf Of John Logsdon
Sent: Saturday, November 11, 2006 12:54 PM
To: [log in to unmask]
Subject: Re: STATISTICAL AND OTHER KINDS OF EVIDENCE


Recognising Ray's quote from my own posts let me take an issue.

Nowhere in my post was I saying that any military action was justified.  
I was saying that military action was in part a consequence of previous
actions.  I was asking whether statistical tools could be used to elucidate.
The conclusions from any such further analysis could have 'favoured' either
'side'.  I was taking a strictly neutral viewpoint - models and their
application are amoral rather than immoral.  Should science have created the
atomic bomb?

Ray has inferred something that I believe is not in my wording and certainly
not in my intention.  He may regard my words as mumbo-jumbo but spinning my
text into that which was not in it just clouds the issue.  I have my beliefs
and judgement on the Iraq issue which may or may not coincide with Ray's.  

Taking Ray's complaint to its logical conclusion (ie don't use statistical
tools where moral judgements are involved), as the Lancet work has political
consequences, it should not have been done.  But I think it has been very
informative and was trying to make a point that the work - which is a
serious attempt to answer an important question - had some restrictions.  As
following posts demonstrated, deeper analysis is not immediately possible.

Clearly I have failed to communicate my neutrality to at least one list
member but I plead for recognition of the facts and not to confuse these
with prejudice.  Whether or not the Iraq campaign was a legal war, who is to
blame for deaths or non-deaths or other considerations are not the
fundamental concerns of this list.

Ray's comment on the uselessness of history is also unfair and in contrast
to his use of the word 'unprecedented'.  We cannot know at this time what
will happen in the future.  We cannot judge whether the facts as known at
that time were presented properly by politicians or journalists.  And we
could not know such things in 2003.  Undoubtedly there was advice that
terrorism would increase - by the military as well as by others.  It may be
that the invasion was in such terms ill-advised or ill-concluded.  Or
history may, in 50 or 100 years time, conclude otherwise.

We do not know whether this politician or that politician, this journalist
or that journalist, was misrepresenting the truth as they knew it or
presenting it to the best of their ability.  Accusing people of lying with
no facts to back them up is clouding judgement with emotion.  We can only
use the best tools we have to elicit conclusions but we do not know - even
in the physical sciences - whether the model or the underlying assumptions
are correct. This is where statistics can help but it is not a perfect tool.


I think this thread has gone rather off-topic and I will not post on it
again.

Best wishes

John

John Logsdon                               "Try to make things as simple
Quantex Research Ltd, Manchester UK         as possible but not simpler"
[log in to unmask]              [log in to unmask]
+44(0)161 445 4951/G:+44(0)7717758675       www.quantex-research.com


On Sat, 11 Nov 2006, ray thomas wrote:

>  I've just been reading Francis Wheen's 'How Mumbo-Jumbo Conquered the 
> World' and was reminded of some of the messages that appeared on this 
> list. For  example:
> 
> "a 'military' action (I use the word as widely as possible to 
> encompass all
> camps) generally has a precursor or some intelligence that makes it 
> pre-emptive."
> 
> This bit of mumbo-jumbo says that military action is generally 
> justified. The function of the use of mumbo-jumbo is to avoid 
> consideration of the nature of the justification.  The context of this 
> comment is the attribution of "blame" in statistics relating to deaths 
> in Iraq.  It was suggested that modelling should be used to estimated 
> the number of deaths that can be blamed on coalition forces, or blamed 
> on insurgents and their families.
> 
> I suggest that this kind of application would bring statistics and
> statisticians into disrepute.    The model is immoral.    The model
assumes
> that military action is justified.      The model makes people killing
each
> other a normal and rational activity.     
> 
> Is it appropriate that statisticians and members of Radstats should be 
> involved in that kind of application?
> 
> Ray Thomas
> 
> 
> ******************************************************
> 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]
> Disclaimer: The messages sent to this list are the views of the sender
> and cannot be assumed to be representative of the range of views held by
subscribers to the Radical Statistics Group. To find out more about Radical
Statistics and its aims and activities and read current and past issues of
our newsletter you are invited to visit our web site www.radstats.org.uk.
> *******************************************************
> 

******************************************************
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]
Disclaimer: The messages sent to this list are the views of the sender and
cannot be assumed to be representative of the range of views held by
subscribers to the Radical Statistics Group. To find out more about Radical
Statistics and its aims and activities and read current and past issues of
our newsletter you are invited to visit our web site www.radstats.org.uk.
*******************************************************

******************************************************
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]
Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk.
*******************************************************