Read before you cite!

Mon Dec 16 14:47:18 EST 2002

Johannes has illustrated a point that I have tried to emphasise before,
namely that proper informetric/scientometric research does not end with the
fitting of mathematical models to data sets. Complete analysis also requires
coss-examination of the data, particularly when there are "outliers" or
other anomolous data points which need to be explained rather than
dismissed.

I confess to being a mathematical/statistical modeller and am grateful that
there are those who are willing to tease out the individual details of data
points rather than be contented with the modeller's broad-brush
descriptions. Hopefully between us we will gain better understanding of the
real underlying processes.

Quentin Burrell

-----Original Message-----
From: ASIS Special Interest Group on Metrics
[mailto:SIGMETRICS at LISTSERV.UTK.EDU]On Behalf Of Johannes Stegmann
Sent: 16 December 2002 15:46
To: SIGMETRICS at LISTSERV.UTK.EDU
Subject: Re: [SIGMETRICS] Read before you cite!

Initially, I did not really know how to deal with the paper by Simkin and
Roychowdhury
(http://arxiv.org/abs/cond-mat/0212043) but now having seen the paper and
its statements mentioned and discussed in the most recent issue of Nature
(Ball, P., 2002: Paper trail reveals references go unread by citing
authors, Nature 420, 594) I cannot overcome the temptation to put my oar in.

The bottom line of the Simkin-Roychowdhury paper - that authors often cite
without having read the referenced papers - might certainly be true but
some of their fashionable equations probably need a change (a wonderful
opportunity for inveterate statisticians - which I'm not - to spend
Christmas holidays): the authors did not mention the possibility of ISI's
database input errors. I looked at "the most popular misprint in a page
number" which is obviously the 78 times cited reference string  KOSTERLITZ
JM,1973,V6,P118 (according to a search in SCISEARCH hosted at DIMDI). I
identified five of the citing papers published in journals the full texts
of which are electronically available for our institution. Two of the five
used the wrong reference string, but three cited the Kosterlitz paper
correctly. The correctly citing papers and the respective Kosterlitz
reference are given below:

Leonel, S.A, Coura, P.Z, Costa B.V, Pires, A.S.T., 2002: Monte Carlo study
of the Ni(C5H5N)(2)Ni(CN)(4)-2d compound.
Solid State Communications, 123, 201-204.
Cites the Kosterlitz paper in correct form:  J.M. Kosterlitz, D.J.
Thouless, J. Phys. C 6 (1973) 1181.

Stephens, G.J., Hu. B.L., 2001: Notes on black hole phase transitions.
International Journal of Theoretical Physics, 40 (12), 2183-2200.
Cites the Kosterlitz paper in correct form: Kosterlitz, J.M. and Thouless,
D.J.(1973). Journal of Physics C 6, 1181.

Borisov, A.B, Kiseliev, V.V., 1998: Vortex dipoles on a soliton lattice
background: Solution of the boundary-value problem by inverse spectral
transform. Physica D, 111, 96-128.
Cites the Kosterlitz paper in correct form: J.M. Kosterlitz, D.J. Thouless,
J. Phys. C 6 (1973) 1181.

Note that although these papers cite correctly, the Kosterlitz reference
string has been assigned to them in "misprinted" form in the SCISEARCH
database.

So, much to think for people to re-formulate formulas and database input
procedures(?).

Have a nice Christmas! Happy New Year to all!

Johannes Stegmann

-------------------------------------------------------
Dr. Johannes Stegmann                   Univ. Hospital Benjamin Franklin
Free University Berlin                  Medical Library
johannes.stegmann at medizin.fu-berlin.de  Hindenburgdamm 30
Tel.: +49 30 8445 2035                  12203 Berlin
Fax:  +49 30 8445 4454                  Germany
  Homepage:  http://www.medizin.fu-berlin.de/medbib/home.html

***********************
This e-mail has been scanned by the Manxnet Mail Guard anti-virus system.
http://www.manx.net/solutions/mailguardhome.asp
***********-***********