The key issue appears to be the statement in the paper 'MCDA distinguishes
between facts and value judgments about the facts'. The process that was
performed here appears to depend on 'expert opinion' alone, at every stage
of the process - deciding what are the relevant outcomes, weighting them
for their relative importance, and also rating different drugs with regard
to these outcomes. It is appropriate for the first two of these steps to
be driven by elicited opinion. But the third step should if at all
possible be closer to being directly data-driven, not just rely on the
opinions of experts who have seen the evidence. It is at this point in
the process that their arguments appear to be vulnerable. I realise that
it would be naive to imagine that doing all this in an objective way would
be as straightforward as conducting a Cochrane review on a well-defined
issue. That process is now regarded as reasonably straightforward only
because it has been done so many times, and experience has been gained.
But surely we now have a better idea than this of how to make that third
step data-driven.
Robert G. Newcombe PhD CStat FFPH
Professor of Medical Statistics
Department of Primary Care and Public Health
Centre for Health Sciences Research
Cardiff University
4th floor, Neuadd Meirionnydd
Heath Park, Cardiff CF14 4YS
Tel: (+44) 29 2068 7260
Fax: (+44) 29 2068 7236
Home page https://sites.google.com/site/robertgnewcombe/home
From: "Ameen J (AT)" <[log in to unmask]>
To: [log in to unmask]
Date: 04/11/2010 21:36
The strength of this paper is in its simplicity, not to forget what
Einstein had said "Keep it Simple (KIS)".That is what modelling is mainly
about. We are yet to learn about the complexity of our surroundings and
the creation of miniture, simple and easy to understand models that can
help us better understand these complex phenomena are all welcome.
However, my comments (Ameen 1995) on a paper by Professor Chatfield are
still valid. Models are subjective and some models are better than some
others. There is no "true" model! It is not clear what the authors mean
by stating "accuracy of individuals judgments..."! in their use of a
version of Visual Analogue Scale (VAS) or a ratio scale as the authors
call it. The benefits of the use of such a scale are stated in a paper
(Coll and Ameen 2004) while the use of a four scale scoring limits its
statistical use and induces biasedness as comparison with normality
becomes difficult. In terms of bisedness, I would have liked to read much
more about the nature and the format of the pannel to ensure unbisedness.
Another point that I would like to make is the georeferencing of the
experiment and the findings which would, to a degree, devalue the
universal scientific content of the paper. For example, the economic
impact in pricing, costs and social acceptance of drugs.
Best wishes
Jamal
Coll AM, Ameen JRM & Mead D (2004). Postoperative pain assessment tools
in day surgery: literature review. Journal of Advanced Nursing. 46(2),
124-133;
Ameen JRM (1995) Comments on Model Uncertainty, Data Mining and
Statistical Inference. J. R. Statist. Soc. A. 453
Prof Jamal Ameen
Senior Advisor, Ministry of Planning
KRG, Erbil
Tel: +496(750) 428 84 82
Email: [log in to unmask]
&
Faculty of Advanced Technology
University of Glamorgan, UK
http://staff.glam.ac.uk/users/3851-jrmameen
________________________________________
From: A UK-based worldwide e-mail broadcast system mailing list
[[log in to unmask]] on behalf of Maria Viskaduraki
[[log in to unmask]]
Sent: Thursday, November 04, 2010 1:32 PM
To: [log in to unmask]
I did read the paper. The moral for me: Choose a controversial subject and
you can get through to a top journal.
The art of statistics is a separate story.
Maria Viskaduraki
Biostatistician
Birmingham
-----Original Message-----
From: A UK-based worldwide e-mail broadcast system mailing list [
mailto:[log in to unmask]] On Behalf Of Tony Greenfield
Sent: 04 November 2010 12:18
To: [log in to unmask]
Subject: nutt
to those allstatters who have read David Nutt's paper in the current
Lancet:
I cannot believe that this was peer reviewed.
what to do you think?
Tony Greenfield
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|
