Alasdair,

I agree with what you say. Basically, we are here to work together under the
currently accepted system -there is yet no other.

Have a happy holiday (again) everyone,
Rui

----- Original Message -----
From: "Alasdair Turner" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Saturday, December 13, 2003 1:52 PM
Subject: Re: Not so highly cited -- correction


> Rui,
>
> Perhaps my invitation to "conclude as you will" was a little rash.  I
> humbly conclude as follows:
>
> First, let us remember that this is not an argument against high profile
> researchers, but against the cult of citation counts.
>
> Secondly, note that I fully accept that high citation and high quality are
> often correlated, and that research into citation counts can be very
> revealing.  For example, see CASA's work on the geography of highly cited
> authors.
>
> What my little look at citation counts has shown is that on a per paper
> basis, a junior author's papers are cited only marginally less often than
a
> 'highly cited' author's.
>
> What does this mean?  Well, it doesn't mean I'm saying that the junior
> author is a fantastic researcher, nor that the highly cited author does
not
> deserve his excellent reputation.
>
> In fact, I am saying the opposite: it means that naively counting
citations
> may well be fundamentally flawed.  If we treat the junior author as
> representative of researchers in general (a gross generalisation, I
admit),
> all one would have to do to become highly cited is not to improve one's
> research, but to improve one's paper output.  Of course, I understand
there
> are problems with this statement.  For it to be true, one would have to
> write publishable papers, and submit them where they would be cited, but
as
> Sanjay and others have pointed out, with a system where journals of
> everything and anything exist, perhaps this is not as hard as it might
> seem.
>
> The initial research has led me to a hypothesis: once the scaling
> properties of citation counts have been taken into account, all authors
> will, per paper, have very similar impact factors.
>
> [As an aside, I am sure research on scaling properties of citation counts
> has been conducted, perhaps you could let me know of any papers that you
> know about?]
>
> Of course, some researchers will be exceptions to such a scheme; there
will
> be those who consistently have highly cited papers.  I humbly submit that,
> if we are to compare researchers, an impact factor is a better method than
> raw counts, because it prevents competition in terms of high output, and
> instead concentrates our minds on achieving maximum exposure for each
piece
> of research we publish.
>
> This does not mean that I support league tables of citations in any form.
> In fact, I believe that papers should be written because we feel that we
> have something important to communicate, not because they might be well
> cited.  On this point, it is interesting that you mention my own output.
I
> do not wish to make any claim for their merit or otherwise, and instead
> humbly submit that if people wish to assess my work, they read it for
> themselves.  I would however point out that if you were to repeat the
> experiment using my citation counts, you would find it supports the
> hypothesis above.
>
> Returning to the main drive of my conclusion: this initial result has
shown
> the RAE would be ill-advised to change its current method of peer review
to
> one based on citation counts.  Further research is necessary in order to
> confirm my findings.
>
> Alasdair
>
>
> Rui Carvalho wrote:
> >
> > Hi Alasdair,
> >
> > Scientists have many qualities. But lack of humbleness is not one of
> > those -specially when you still have along way to go...
> >
> > Attached, you will find the paper that brought me to UCL. This appeared
as a
> > review to a paper by Dirk Helbing in the same Nature issue. One year
later,
> > The New Scientist published a review on Space Syntax -and these guys
were
> > not reading architectural books. I will also let you conclude as you
will.
> >
> > By the way, Nature has an impact factor of 30 ... You may find it a good
> > exercise to add up the total impact factor of your publications and find
out
> > where you stand -I was not highly impressed when I did it myself from
your
> > web page.
> >
> > For those of you who do not know about it, we use impact factors
(together
> > with citation counts) to rank ourselves, eventually conquer the respect
of
> > our peers, and the British Government uses them to evaluate us (RAE).
The
> > method has its problems, but drawing general conclusions from an
individual
> > case can be very perverse.
> >
> > Have a good Xmas break everyone.
> >
> > -Rui
> >
> > ----- Original Message -----
> > From: "Alasdair Turner" <[log in to unmask]>
> > To: <[log in to unmask]>
> > Sent: Wednesday, December 10, 2003 4:55 PM
> > Subject: Not so highly cited -- correction
> >
> > > You will note that these figures are still of the same order as the
lesser
> > > known researcher (indeed, the logarithmic metric is the same).
Conclude
> > as
> > > you will.
> > >
> >
>
  --------------------------------------------------------------------------
-
> >                  Name: MBatty.pdf
> >    MBatty.pdf    Type: AcroRd32 File (application/pdf)
> >              Encoding: base64
>
> --
> Alasdair Turner
> Lecturer in Architectural Computing
> Bartlett School of Graduate Studies                    tel +44 20 7679
1806
> UCL  Gower Street  London  WC1E 6BT                    fax +44 20 7813
2843
>
> This email and any attached files are confidential and copyright
protected.
> If you are not the addressee, any dissemination of this communication is
> strictly prohibited. Unless otherwise expressly agreed in writing, nothing
> stated in this communication shall be legally binding.
>