Aksnes DW "Characteristics of highly cited papers" Research Evaluation 12(3):159-170 December 2003

[Eugene Garfield]

E-Mail:
Dag W. Aksnes : Dag.W.Aksnes at nifu.no

FULL TEXT AVAILABLE  AT :

http://english.nifustep.no/norsk/publikasjoner/characteristics_of_highly_cited_papers


Title:     Characteristics of highly cited papers

Author(s): Aksnes DW

Source:    RESEARCH EVALUATION 12 (3): 159-170 DEC 2003

Abstract:
Highly cited articles are very different from 'ordinary' cited articles.
Typically, they are authored by a large number of scientists, often
involving international collaboration. The majority of the papers represent
regular journal articles (81%), although review articles (12%) are
over-represented compared to the national average. The citation curves of
highly cited papers follow a typical pattern of rise and decline. However,
different types of citation curves can be identified, reflecting possible
differences in the cognitive function of the articles. Highly cited papers
typically obtain citations from a large number of different journals and
from papers representing both close and remote fields. However, this pattern
is not very different from the average distribution for all papers. We
discuss how the findings can be explained by introducing a conceptual
distinction between quality dynamics and visibility dynamics.

EXCERPT OF PAPER :

Highly cited papers are mainly present in high-impact journals.  It is a
statistical  truism that highly cited papers tend to be published in
high-impact journals (journals with high average citation rates).  In order
to analyse to what extent this was the case in our sample, we collected the
impact factors of the present journals, using the Journal Performance
Indicator database (JPI).  All except three papers were published in
journals to which an impact factor could be assigned.  We then compared the
impat factor with the average for the field to which the journal was
assigned using the NSI-database.  In both cases we used a running five-year
window.  Using this method an index number was calculated for all
publications.  In order to analyse how the distribution of the highly cited
papers differed from the average, we made similar calculations for the
overall national publication set.

The results are shown in Figure 1.  As we can see, the highly cited papers
have a very different distribution from what is the normal Norwegian
distribution.  20% of the highly cited papers appear in journals with a very
high relative journal impact factor (that is, in journals in which the
impact factor is more than four times higher than the average citation.

Similarly, 56% of the highly cited papers appear in journals with a high
relative impact factor (1.5 - 4 times the average).  In comparison, only 15%
of the total population of Norwegian papers appear in this journal category.
 Only 9% of the highly cited articles were published in journals with an
impact factor below the field average.  In contrast, 67% of the Norwegian
article production appears in this category of journals cited below average.

Although these results may not seem unexpected, one should notice that 9% of
the highly cited papers appear in poorly cited journals.  Thus, in order to
become highly cited it is not a necessary condition to be printed in
high-impact journals or journals of high prestige.  This and the skewness in
the citation distribution that can be found within most journals indicate
that contents are important determinants in citation rates.  The journal
address represents a contributory factor, but here there is also a choice
effect.  When scientists think they have something important on their hands
(which will qualify for high citation) they will submit the paper to a
high-status journal (which usually has high -impact factors).  So there is
also a content-status effect derived from author's choices.


Addresses: Aksnes DW (reprint author), Norwegian Inst Studies Res & Higher
Educ, NIFU, Hegdehaugsveien 31, Oslo, NO-0352 Norway
Norwegian Inst Studies Res & Higher Educ, NIFU, Oslo, NO-0352 Norway

E-mail Addresses: Dag.W.Aksnes at nifu.no

Publisher: BEECH TREE PUBLISHING, 10 WATFORD CLOSE,, GUILDFORD GU1 2EP,
SURREY, ENGLAND
Subject Category: INFORMATION SCIENCE & LIBRARY SCIENCE
IDS Number: 779LY

ISSN: 0958-2029

CITED REFERENCES :
   *EUR COMM, 2001, KEY FIG 2001 SPEC ED
   AKSNES DW, 2000, SCIENTOMETRICS, V49, P7
   AKSNES DW, 2003, SCIENTOMETRICS, V56, P235
   AKSNES DW, 2004, IN PRESS SCIENTOMETR
   ARUNACHALAM S, 1984, J INFORM SCI, V8, P93
   AVERSA ES, 1985, SCIENTOMETRICS, V7, P383
   CANO V, 1991, SCIENTOMETRICS, V22, P297
   COLE S, 1992, MAKING SCI NATURE SO
   COLE S, 2000, WEB KNOWLEDGE FESTSC
   CRONIN B, 1984, CITATION PROCESS ROL
   GILBERT GN, 1977, SOC STUD SCI, V7, P113
   GLANZEL W, 1992, RES EVALUAT, V2, P135
   GLANZEL W, 1995, RES EVALUAT, V5, P113
   LAWANI SM, 1986, SCIENTOMETRICS, V9, P13
   LINE MB, 1984, J INFORM SCI, V9, P90
   MACROBERTS MH, 1989, J AM SOC INFORM SCI, V40, P342
   MARTIN BR, 1983, RES POLICY, V12, P61
   MELIN G, 1997, COPRODUCTION SCI KNO
   MERTON RK, 1968, SCIENCE, V159, P56
   PETERS HPF, 1994, J AM SOC INFORM SCI, V45, P39
   PORTER AL, 1985, SCIENTOMETRICS, V8, P161
   PRICE DJD, 1965, SCIENCE, V149, P510
   PRICE DJD, 1976, J AM SOC INFORM SCI, V27, P292
   RINIA EJ, 2002, SCIENTOMETRICS, V54, P347
   SEGIEN PO, 1992, J AM SOC INFORM SCI, V43, P628
   SEGLEN PO, 1997, ALLERGY, V52, P1050
   SEGLEN PO, 1997, BRIT MED J, V314, P498
   TIJSSEN RJW, 2002, SCIENTOMETRICS, V54, P381
   VANRAAN AFJ, 2000, WEB KNOWLEDGE FESTSC