French Issue of Scientometrics

[Gretchen Whitney]

An upcoming issue of Scientometrics will focus on research in France.
Below is an English summary of the introduction to the issue.  --gw

---------- Forwarded message ----------
Date: Mon, 19 Jun 2000 11:47:03 +0200
From: de looze Marie-Angèle <delooze at grenoble.inra.fr>

SCIENTOMETRICS RESEARCH IN FRANCE

(INVITED PAPERS)

GUEST EDITOR:  YOSHIKO OKUBO

Observatoire des Sciences et des Techniques
93 rue de Vaugirard, 75006 Paris (France)

AN INTRODUCTION TO SCIENTOMETRICS IN FRANCE

Yoshiko Okubo

Observatoire des Sciences et des Techniques (OST)
93 rue de Vaugirard, 75006 Paris (France)

The collection of papers in this special issue of Scientometrics Research
in France, while not pretending to be exhaustive, represents a broad
compilation of three types of work: (1) new applications of scientometric
indicators and tools for better understanding of science; (2) development
of measuring tools and the use of new data sources; and (3) research on
electronic networks.  An overview of the contents will describe these groups.

The paper by Arvanitis et al. is a unique study of science in Africa.  It
not only sheds light on an undeservedly neglected topic, but it also deals
with the recent evolution of the scientific output of African nations.  The
authors describe the state of scientific capacity in  countries in Africa
by using published production retrieved in the French Pascal database.
Their research reveals the crisis of African research during the current
decade. Africa has lost 20-25% of its relative capacity to contribute to
world science. There has been a great divergence in the evolution of
science in different nations; while several decades ago middle-sized
scientific powers began to establish themselves, in the 1990s marked
reversals in ranking by country took place. Abrupt changes in science
policies and international aid directly affected African capacity to
produce science. Contractualization and an increase in consultancy study
proposals led to a decline in basic science and added to the difficulties
of developing a breeding ground for learned culture. This paper raises
important questions as to the policy of international aid for this continent.

Doré et al. introduce a new application of classical multidimensional
analyses technique, Correspondence Factor Analysis (CFA). CFA is a
high-performance mapping method that achieves appropriate data reduction
and objectifies correlation among variables. For these advantages, this
technique has widely been used in scientometric studies. The authors
present CFA as a useful tool to analyze time-series, an application which
has so far not been sufficiently exploited.  They demonstrate the power of
the tool by depicting evolution of patent activity in the field of
Chemistry and Biochemistry, from the late 1970s to the present day.  They
confirm the general finding that major European countries did not show
increased interest in filing patent applications in the USA in recent
years.  The authors raise some questions as to the competitiveness of the
major European nations and point to possible changes in patent strategies
on a global level.  The authors also describe the changing nature of the
difficulties encountered in obtaining patent databases.  The major handicap
in analyzing trends in science and technology used to be the lack of easy
access to appropriate databases.  The contribution of the Internet and new
means of data acquisition is discussed.

Sigogneau's paper deals with basic issues related to bibliometric
measurement.  She raises questions about what "document types" should be
counted for scientific performance of countries.  In scientific journals
there are diverse types of papers (or document types) each having a
function in the production and dissemination of science.  Out of numerous
document types, it is widely accepted that only articles, notes and reviews
(and sometimes letters) reflect significant scientific research and thus
appropriate for evaluating national performances.  Sigogneau, however,
calls attention to "Proceedings".  In analyzing Proceedings recorded by the
ISI database and published in journals related to Physics, it is
demonstrated that this document type takes 19% of the entire publications
in the Science Citation Index. Moreover, while Letters show a steady
decline, and Notes are tending to disappear, Proceedings are increasing in
recent years.  The inclusion of Proceedings in bibliometrical analysis may
modify performances of countries.  A detailed analysis on Proceedings
presented here (citation, number of pages, trend) opens debate as to the
significance of this document type.

Zitt et al. examine co-publications of five large producers of science
using a classic Probabilistic Affinity Index, adding an additional
perspective to the extensive bibliometric literature on scientific
cooperation.  They show that when scientific size of a nation is
controlled, scientific linkages created by each actor are specific to a
nation and are influenced by historical, cultural and linguistic
proximities, with few changes over a 10-year period.  A lack of specific
affinities among the three largest European countries is also revealed, and
the ensuing discussion raises some questions as to the uncertain process of
Europeanization in science.  The second analysis examines whether
collaborative intensity is affected by the perception of the relative
strengths and weaknesses in particular fields.  The authors come up with
three models of cooperation: 1) "co-option behavior", a prevailing pattern
where two countries collaborate most in their field of common excellence;
2) "solidarity pattern" when the collaboration is in the common field of
weakness of the two, and 3) a "master-pupil" relationship. The Japan-France
relationship seems to be a typical example of such cooperative behavior.

Solari and Magri constructed a system for evaluating scientific journals
based on a new approach to the Science Citation Index - Journal Citation
Report.  The SCI-JCR, which lists Impact Factor, is a well known tool for
ranking, evaluating and comparing journals, but its shortcoming is that it
is not user-friendly.  The authors argue that with the Impact Factor of the
4,500 journals listed in descending order, it is not only difficult to make
comparisons between journals, but the analyst can easily be misled into
oversimplifying the complex reality of the world of journals.  Their
research involved developing a synthetic method to facilitate the use of
the tool.  The journals listed in the JCR were classified into five group
rankings by use of the box plot method.  The resulting system enables more
precise judgements to be made about the journals, while facilitating the
identification of the scientific discipline of a journal  as well as intra-
and inter-disciplinary comparisons of journals.

Gusmao's paper concerns the construction of comprehensive multilateral S&T
cooperation indicators using European research program data.  European
Union research programs play an increasingly important role within the
research and innovation systems of member states today.  The need for
appropriate indicators to monitor and analyze collaborative phenomenon is
pressing.  Gusmao's work is therefore timely and provides an extremely
useful input to policy makers and analysts.  She not only explains the
difficult procedure of homogenizing and processing raw data, but also
provides a long list of questions that can potentially be answered by this
tool.  The usefulness of this type of data processing was tested by the
Ministry of Research in France recently.

Lemarié et al. analyze 228 European Biotechnology SMEs using data from the
Genetic Engineering catalogue.  This data not only provides complementary
information to the classical patent-based analyses, but it also contains
more details on the SMEs.  The authors analyze technologies which SMEs are
developing and markets on which they are focusing in the field of
Biotechnology, in three European countries.  One of their findings is that
specific patterns of specialization by country are not observable.  They
raise questions about the correlation between emerging technologies and the
dynamics of SMEs and argue that the development of the sector and the rapid
evolution of technologies are mainly due to the creation of new SMEs rather
than to the renewal of technology in existing firms.  Also
technologies/market combinations are mapped out.  A clear technology-market
link exists in generic technologies and health care technologies;
environmental markets are "market-based" containing very few technologies;
no links were perceived in emerging technology, it does not yet have a
corresponding market (technology based).  Numerous works on Biotechnology
have been conducted by authors using patents.  This paper adds another view
of activities of firms in this sector.

Larédo and Mustar's article concerns laboratory studies.  They argue that
owing to the fact that scientific activity is no longer an act by
individual scientists, the conditions under which research activities
develop must be analyzed if one is to understand scientific production
activity.  Thus laboratory studies are an important element of the new
sociology of science.  Scientific activities cannot be captured by a single
indicator, but rather a set of  indicators is needed.  The authors  develop
a method which  allows the  characterization of  "activity profiles" of
research laboratories by combining a set of indicators.  They identify five
complementary contexts in which research activities develop:  1) the
production of certified knowledge; 2) education and training; 3)
participation in  the innovation process or competitive research; 4)
participation in the development of public or collective goods; and 5)
participation in public debate on S&T.  Indicators are chosen for each
category of activity and the resulting measurement provides  the productive
pattern of a laboratory.  75 laboratories in France were actually measured
by this method and the feasibility of the method is discussed.

An original paper by Salaün and Lafouge concerns the development of
electronic publishing. It is said that sooner or later electronic networks
will be the main vehicle for the circulation of scientific literature. So
far, investigation has been widely conducted on communication between
researchers or scientific publishing through this medium, but "demand for
articles" has attracted little attention, though this type of analysis
provides a useful view for the organization of electronic publishing.
Their research starts by raising questions such as: is it possible to
characterize demand in scientific articles? Who requests articles and how
frequently? What articles are requested, in which discipline, and from
which journal?  An examination was conducted to see whether there is any
correlation between requests for scientific articles and the number of
citations in the relevant journals.  By analyzing the requests for
scientific articles to document suppliers they came up with a certain
client profile as well as an indication of in which scientific disciplines
demand is concentrated.


To characterize science and technology activities, the actors, the
production, the dynamics, and science's  contribution to society are some
of the major research goals of scientometricians in France.  Another is to
monitor the constantly changing relationships between science, technology,
and markets.  In order to understand the increasingly complex system of
science, more detailed scientometric information is needed, and the need is
making itself increasingly felt as it is in the scientometric  community at
large.  There is an increasing diversification in the types of questions
asked, in particular by science-policy making process.  We are challenged
by internal and external needs for closer observation of the world of
science and technology.  Such demand spurs us to seek appropriate
approaches, to explore diversified information sources, to establish
empirical data and to develop indicators that are more adapted to social
needs.

The  papers comprising this collection are a fair sampling  of the variety
of applications, methods, tools, indicators, databases and information
sources which make up our attempt to meet these social needs.

There are nine papers in this collection, all "invited papers" from the
French scientometric community.  Several articles could not be published in
this collection, since an international jury found their subject matter to
lie outside the traditional core of "scientometrics".  They fit better in
the field of Econometrics, or Infometrics, Technology Watch, etc.  This
fact does however demonstrate that the French scientometric community is
expanding its scope, reaching toward other disciplines.  In this context,
the question is whether Scientometrics should continue to focus on
classical scientometric themes, or whether some efforts should be made to
include research in adjacent disciplines.

Lastly, a legal issue has appeared which may be worth mentioning in order
to describe the present situation of scientometricians:  There are
restrictions on the use of certain data, which are an indispensable
instrument for our activity.  This is a difficult issue which shows the
importance of industrial property not only for French scientometricians,
but also for the wider scientometric community.

I would like to thank Scientometrics for providing French researchers an
opportunity to be represented together in this special issue.  Also we are
indebted to numerous international experts for their cooperation in
evaluating published and non-published articles.